idnits 2.17.1 draft-ietf-detnet-use-cases-07.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- ** The document seems to lack an IANA Considerations section. (See Section 2.2 of https://www.ietf.org/id-info/checklist for how to handle the case when there are no actions for IANA.) Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (March 5, 2016) is 2974 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Unused Reference: 'ACE' is defined on line 2735, but no explicit reference was found in the text == Unused Reference: 'CCAMP' is defined on line 2743, but no explicit reference was found in the text == Unused Reference: 'DICE' is defined on line 2765, but no explicit reference was found in the text == Unused Reference: 'EA12' is defined on line 2768, but no explicit reference was found in the text == Unused Reference: 'HART' is defined on line 2785, but no explicit reference was found in the text == Unused Reference: 'I-D.ietf-6tisch-terminology' is defined on line 2814, but no explicit reference was found in the text == Unused Reference: 'I-D.ietf-ipv6-multilink-subnets' is defined on line 2820, but no explicit reference was found in the text == Unused Reference: 'I-D.ietf-roll-rpl-industrial-applicability' is defined on line 2825, but no explicit reference was found in the text == Unused Reference: 'I-D.thubert-6lowpan-backbone-router' is defined on line 2853, but no explicit reference was found in the text == Unused Reference: 'IEC61850-90-12' is defined on line 2863, but no explicit reference was found in the text == Unused Reference: 'IEEE8021TSN' is defined on line 2926, but no explicit reference was found in the text == Unused Reference: 'IETFDetNet' is defined on line 2932, but no explicit reference was found in the text == Unused Reference: 'RFC2119' is defined on line 2989, but no explicit reference was found in the text == Unused Reference: 'RFC2460' is defined on line 2994, but no explicit reference was found in the text == Unused Reference: 'RFC2474' is defined on line 2998, but no explicit reference was found in the text == Unused Reference: 'RFC3209' is defined on line 3009, but no explicit reference was found in the text == Unused Reference: 'RFC3393' is defined on line 3014, but no explicit reference was found in the text == Unused Reference: 'RFC3444' is defined on line 3019, but no explicit reference was found in the text == Unused Reference: 'RFC3972' is defined on line 3024, but no explicit reference was found in the text == Unused Reference: 'RFC4291' is defined on line 3033, but no explicit reference was found in the text == Unused Reference: 'RFC4903' is defined on line 3042, but no explicit reference was found in the text == Unused Reference: 'RFC4919' is defined on line 3046, but no explicit reference was found in the text == Unused Reference: 'RFC6282' is defined on line 3063, but no explicit reference was found in the text == Unused Reference: 'RFC6775' is defined on line 3081, but no explicit reference was found in the text == Unused Reference: 'TEAS' is defined on line 3108, but no explicit reference was found in the text == Unused Reference: 'UHD-video' is defined on line 3139, but no explicit reference was found in the text == Outdated reference: A later version (-08) exists of draft-finn-detnet-architecture-03 == Outdated reference: A later version (-05) exists of draft-finn-detnet-problem-statement-04 == Outdated reference: A later version (-30) exists of draft-ietf-6tisch-architecture-09 == Outdated reference: A later version (-10) exists of draft-ietf-6tisch-terminology-06 -- Obsolete informational reference (is this intentional?): RFC 2460 (Obsoleted by RFC 8200) Summary: 1 error (**), 0 flaws (~~), 31 warnings (==), 2 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Internet Engineering Task Force E. Grossman, Ed. 3 Internet-Draft DOLBY 4 Intended status: Informational C. Gunther 5 Expires: September 6, 2016 HARMAN 6 P. Thubert 7 P. Wetterwald 8 CISCO 9 J. Raymond 10 HYDRO-QUEBEC 11 J. Korhonen 12 BROADCOM 13 Y. Kaneko 14 Toshiba 15 S. Das 16 Applied Communication Sciences 17 Y. Zha 18 HUAWEI 19 B. Varga 20 J. Farkas 21 Ericsson 22 F. Goetz 23 J. Schmitt 24 Siemens 25 March 5, 2016 27 Deterministic Networking Use Cases 28 draft-ietf-detnet-use-cases-07 30 Abstract 32 This draft documents requirements in several diverse industries to 33 establish multi-hop paths for characterized flows with deterministic 34 properties. In this context deterministic implies that streams can 35 be established which provide guaranteed bandwidth and latency which 36 can be established from either a Layer 2 or Layer 3 (IP) interface, 37 and which can co-exist on an IP network with best-effort traffic. 39 Additional requirements include optional redundant paths, very high 40 reliability paths, time synchronization, and clock distribution. 41 Industries considered include wireless for industrial applications, 42 professional audio, electrical utilities, building automation 43 systems, radio/mobile access networks, automotive, and gaming. 45 For each case, this document will identify the application, identify 46 representative solutions used today, and what new uses an IETF DetNet 47 solution may enable. 49 Status of This Memo 51 This Internet-Draft is submitted in full conformance with the 52 provisions of BCP 78 and BCP 79. 54 Internet-Drafts are working documents of the Internet Engineering 55 Task Force (IETF). Note that other groups may also distribute 56 working documents as Internet-Drafts. The list of current Internet- 57 Drafts is at http://datatracker.ietf.org/drafts/current/. 59 Internet-Drafts are draft documents valid for a maximum of six months 60 and may be updated, replaced, or obsoleted by other documents at any 61 time. It is inappropriate to use Internet-Drafts as reference 62 material or to cite them other than as "work in progress." 64 This Internet-Draft will expire on September 6, 2016. 66 Copyright Notice 68 Copyright (c) 2016 IETF Trust and the persons identified as the 69 document authors. All rights reserved. 71 This document is subject to BCP 78 and the IETF Trust's Legal 72 Provisions Relating to IETF Documents 73 (http://trustee.ietf.org/license-info) in effect on the date of 74 publication of this document. Please review these documents 75 carefully, as they describe your rights and restrictions with respect 76 to this document. Code Components extracted from this document must 77 include Simplified BSD License text as described in Section 4.e of 78 the Trust Legal Provisions and are provided without warranty as 79 described in the Simplified BSD License. 81 Table of Contents 83 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 5 84 2. Pro Audio and Video . . . . . . . . . . . . . . . . . . . . . 5 85 2.1. Use Case Description . . . . . . . . . . . . . . . . . . 5 86 2.1.1. Uninterrupted Stream Playback . . . . . . . . . . . . 6 87 2.1.2. Synchronized Stream Playback . . . . . . . . . . . . 7 88 2.1.3. Sound Reinforcement . . . . . . . . . . . . . . . . . 7 89 2.1.4. Deterministic Time to Establish Streaming . . . . . . 8 90 2.1.5. Secure Transmission . . . . . . . . . . . . . . . . . 8 91 2.1.5.1. Safety . . . . . . . . . . . . . . . . . . . . . 8 92 2.1.5.2. Digital Rights Management (DRM) . . . . . . . . . 8 93 2.2. Pro Audio Today . . . . . . . . . . . . . . . . . . . . . 9 94 2.3. Pro Audio Future . . . . . . . . . . . . . . . . . . . . 9 95 2.3.1. Layer 3 Interconnecting Layer 2 Islands . . . . . . . 9 96 2.3.2. High Reliability Stream Paths . . . . . . . . . . . . 9 97 2.3.3. Link Aggregation . . . . . . . . . . . . . . . . . . 10 98 2.3.4. Integration of Reserved Streams into IT Networks . . 10 99 2.3.5. Use of Unused Reservations by Best-Effort Traffic . . 10 100 2.3.6. Traffic Segregation . . . . . . . . . . . . . . . . . 10 101 2.3.6.1. Packet Forwarding Rules, VLANs and Subnets . . . 11 102 2.3.6.2. Multicast Addressing (IPv4 and IPv6) . . . . . . 11 103 2.3.7. Latency Optimization by a Central Controller . . . . 11 104 2.3.8. Reduced Device Cost Due To Reduced Buffer Memory . . 12 105 2.4. Pro Audio Asks . . . . . . . . . . . . . . . . . . . . . 12 106 3. Electrical Utilities . . . . . . . . . . . . . . . . . . . . 12 107 3.1. Use Case Description . . . . . . . . . . . . . . . . . . 12 108 3.1.1. Transmission Use Cases . . . . . . . . . . . . . . . 13 109 3.1.1.1. Protection . . . . . . . . . . . . . . . . . . . 13 110 3.1.1.2. Intra-Substation Process Bus Communications . . . 18 111 3.1.1.3. Wide Area Monitoring and Control Systems . . . . 19 112 3.1.1.4. IEC 61850 WAN engineering guidelines requirement 113 classification . . . . . . . . . . . . . . . . . 20 114 3.1.2. Generation Use Case . . . . . . . . . . . . . . . . . 21 115 3.1.3. Distribution use case . . . . . . . . . . . . . . . . 22 116 3.1.3.1. Fault Location Isolation and Service Restoration 117 (FLISR) . . . . . . . . . . . . . . . . . . . . . 22 118 3.2. Electrical Utilities Today . . . . . . . . . . . . . . . 23 119 3.2.1. Security Current Practices and Limitations . . . . . 23 120 3.3. Electrical Utilities Future . . . . . . . . . . . . . . . 25 121 3.3.1. Migration to Packet-Switched Network . . . . . . . . 25 122 3.3.2. Telecommunications Trends . . . . . . . . . . . . . . 26 123 3.3.2.1. General Telecommunications Requirements . . . . . 26 124 3.3.2.2. Specific Network topologies of Smart Grid 125 Applications . . . . . . . . . . . . . . . . . . 27 126 3.3.2.3. Precision Time Protocol . . . . . . . . . . . . . 28 127 3.3.3. Security Trends in Utility Networks . . . . . . . . . 29 128 3.4. Electrical Utilities Asks . . . . . . . . . . . . . . . . 31 129 4. Building Automation Systems . . . . . . . . . . . . . . . . . 31 130 4.1. Use Case Description . . . . . . . . . . . . . . . . . . 31 131 4.2. Building Automation Systems Today . . . . . . . . . . . . 31 132 4.2.1. BAS Architecture . . . . . . . . . . . . . . . . . . 32 133 4.2.2. BAS Deployment Model . . . . . . . . . . . . . . . . 33 134 4.2.3. Use Cases for Field Networks . . . . . . . . . . . . 35 135 4.2.3.1. Environmental Monitoring . . . . . . . . . . . . 35 136 4.2.3.2. Fire Detection . . . . . . . . . . . . . . . . . 35 137 4.2.3.3. Feedback Control . . . . . . . . . . . . . . . . 36 138 4.2.4. Security Considerations . . . . . . . . . . . . . . . 36 139 4.3. BAS Future . . . . . . . . . . . . . . . . . . . . . . . 36 140 4.4. BAS Asks . . . . . . . . . . . . . . . . . . . . . . . . 37 141 5. Wireless for Industrial . . . . . . . . . . . . . . . . . . . 37 142 5.1. Use Case Description . . . . . . . . . . . . . . . . . . 37 143 5.1.1. Network Convergence using 6TiSCH . . . . . . . . . . 38 144 5.1.2. Common Protocol Development for 6TiSCH . . . . . . . 38 146 5.2. Wireless Industrial Today . . . . . . . . . . . . . . . . 39 147 5.3. Wireless Industrial Future . . . . . . . . . . . . . . . 39 148 5.3.1. Unified Wireless Network and Management . . . . . . . 39 149 5.3.1.1. PCE and 6TiSCH ARQ Retries . . . . . . . . . . . 41 150 5.3.2. Schedule Management by a PCE . . . . . . . . . . . . 42 151 5.3.2.1. PCE Commands and 6TiSCH CoAP Requests . . . . . . 42 152 5.3.2.2. 6TiSCH IP Interface . . . . . . . . . . . . . . . 43 153 5.3.3. 6TiSCH Security Considerations . . . . . . . . . . . 43 154 5.4. Wireless Industrial Asks . . . . . . . . . . . . . . . . 44 155 6. Cellular Radio Use Cases . . . . . . . . . . . . . . . . . . 44 156 6.1. Use Case Description . . . . . . . . . . . . . . . . . . 44 157 6.1.1. Network Architecture . . . . . . . . . . . . . . . . 44 158 6.1.2. Time Synchronization Requirements . . . . . . . . . . 45 159 6.1.3. Time-Sensitive Stream Requirements . . . . . . . . . 47 160 6.1.4. Security Considerations . . . . . . . . . . . . . . . 47 161 6.2. Cellular Radio Networks Today . . . . . . . . . . . . . . 48 162 6.3. Cellular Radio Networks Future . . . . . . . . . . . . . 48 163 6.4. Cellular Radio Networks Asks . . . . . . . . . . . . . . 50 164 7. Cellular Coordinated Multipoint Processing (CoMP) . . . . . . 50 165 7.1. Use Case Description . . . . . . . . . . . . . . . . . . 50 166 7.1.1. CoMP Architecture . . . . . . . . . . . . . . . . . . 51 167 7.1.2. Delay Sensitivity in CoMP . . . . . . . . . . . . . . 52 168 7.2. CoMP Today . . . . . . . . . . . . . . . . . . . . . . . 52 169 7.3. CoMP Future . . . . . . . . . . . . . . . . . . . . . . . 52 170 7.3.1. Mobile Industry Overall Goals . . . . . . . . . . . . 52 171 7.3.2. CoMP Infrastructure Goals . . . . . . . . . . . . . . 53 172 7.4. CoMP Asks . . . . . . . . . . . . . . . . . . . . . . . . 53 173 8. Industrial M2M . . . . . . . . . . . . . . . . . . . . . . . 54 174 8.1. Use Case Description . . . . . . . . . . . . . . . . . . 54 175 8.2. Industrial M2M Communication Today . . . . . . . . . . . 55 176 8.2.1. Transport Parameters . . . . . . . . . . . . . . . . 55 177 8.2.2. Stream Creation and Destruction . . . . . . . . . . . 56 178 8.3. Industrial M2M Future . . . . . . . . . . . . . . . . . . 56 179 8.4. Industrial M2M Asks . . . . . . . . . . . . . . . . . . . 57 180 9. Internet-based Applications . . . . . . . . . . . . . . . . . 57 181 9.1. Use Case Description . . . . . . . . . . . . . . . . . . 57 182 9.1.1. Media Content Delivery . . . . . . . . . . . . . . . 57 183 9.1.2. Online Gaming . . . . . . . . . . . . . . . . . . . . 57 184 9.1.3. Virtual Reality . . . . . . . . . . . . . . . . . . . 57 185 9.2. Internet-Based Applications Today . . . . . . . . . . . . 58 186 9.3. Internet-Based Applications Future . . . . . . . . . . . 58 187 9.4. Internet-Based Applications Asks . . . . . . . . . . . . 58 188 10. Use Case Common Elements . . . . . . . . . . . . . . . . . . 58 189 11. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 59 190 11.1. Pro Audio . . . . . . . . . . . . . . . . . . . . . . . 59 191 11.2. Utility Telecom . . . . . . . . . . . . . . . . . . . . 60 192 11.3. Building Automation Systems . . . . . . . . . . . . . . 60 193 11.4. Wireless for Industrial . . . . . . . . . . . . . . . . 60 194 11.5. Cellular Radio . . . . . . . . . . . . . . . . . . . . . 60 195 11.6. Industrial M2M . . . . . . . . . . . . . . . . . . . . . 60 196 11.7. Internet Applications and CoMP . . . . . . . . . . . . . 60 197 12. Informative References . . . . . . . . . . . . . . . . . . . 61 198 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 69 200 1. Introduction 202 This draft presents use cases from diverse industries which have in 203 common a need for deterministic streams, but which also differ 204 notably in their network topologies and specific desired behavior. 205 Together, they provide broad industry context for DetNet and a 206 yardstick against which proposed DetNet designs can be measured (to 207 what extent does a proposed design satisfy these various use cases?) 209 For DetNet, use cases explicitly do not define requirements; The 210 DetNet WG will consider the use cases, decide which elements are in 211 scope for DetNet, and the results will be incorporated into future 212 drafts. Similarly, the DetNet use case draft explicitly does not 213 suggest any specific design, architecture or protocols, which will be 214 topics of future drafts. 216 We present for each use case the answers to the following questions: 218 o What is the use case? 220 o How is it addressed today? 222 o How would you like it to be addressed in the future? 224 o What do you want the IETF to deliver? 226 The level of detail in each use case should be sufficient to express 227 the relevant elements of the use case, but not more. 229 At the end we consider the use cases collectively, and examine the 230 most significant goals they have in common. 232 2. Pro Audio and Video 234 2.1. Use Case Description 236 The professional audio and video industry ("ProAV") includes: 238 o Music and film content creation 240 o Broadcast 241 o Cinema 243 o Live sound 245 o Public address, media and emergency systems at large venues 246 (airports, stadiums, churches, theme parks). 248 These industries have already transitioned audio and video signals 249 from analog to digital. However, the digital interconnect systems 250 remain primarily point-to-point with a single (or small number of) 251 signals per link, interconnected with purpose-built hardware. 253 These industries are now transitioning to packet-based infrastructure 254 to reduce cost, increase routing flexibility, and integrate with 255 existing IT infrastructure. 257 Today ProAV applications have no way to establish deterministic 258 streams from a standards-based Layer 3 (IP) interface, which is a 259 fundamental limitation to the use cases described here. Today 260 deterministic streams can be created within standards-based layer 2 261 LANs (e.g. using IEEE 802.1 AVB) however these are not routable via 262 IP and thus are not effective for distribution over wider areas (for 263 example broadcast events that span wide geographical areas). 265 It would be highly desirable if such streams could be routed over the 266 open Internet, however solutions with more limited scope (e.g. 267 enterprise networks) would still provide a substantial improvement. 269 The following sections describe specific ProAV use cases. 271 2.1.1. Uninterrupted Stream Playback 273 Transmitting audio and video streams for live playback is unlike 274 common file transfer because uninterrupted stream playback in the 275 presence of network errors cannot be achieved by re-trying the 276 transmission; by the time the missing or corrupt packet has been 277 identified it is too late to execute a re-try operation. Buffering 278 can be used to provide enough delay to allow time for one or more 279 retries, however this is not an effective solution in applications 280 where large delays (latencies) are not acceptable (as discussed 281 below). 283 Streams with guaranteed bandwidth can eliminate congestion on the 284 network as a cause of transmission errors that would lead to playback 285 interruption. Use of redundant paths can further mitigate 286 transmission errors to provide greater stream reliability. 288 2.1.2. Synchronized Stream Playback 290 Latency in this context is the time between when a signal is 291 initially sent over a stream and when it is received. A common 292 example in ProAV is time-synchronizing audio and video when they take 293 separate paths through the playback system. In this case the latency 294 of both the audio and video streams must be bounded and consistent if 295 the sound is to remain matched to the movement in the video. A 296 common tolerance for audio/video sync is one NTSC video frame (about 297 33ms) and to maintain the audience perception of correct lip sync the 298 latency needs to be consistent within some reasonable tolerance, for 299 example 10%. 301 A common architecture for synchronizing multiple streams that have 302 different paths through the network (and thus potentially different 303 latencies) is to enable measurement of the latency of each path, and 304 have the data sinks (for example speakers) delay (buffer) all packets 305 on all but the slowest path. Each packet of each stream is assigned 306 a presentation time which is based on the longest required delay. 307 This implies that all sinks must maintain a common time reference of 308 sufficient accuracy, which can be achieved by any of various 309 techniques. 311 This type of architecture is commonly implemented using a central 312 controller that determines path delays and arbitrates buffering 313 delays. 315 2.1.3. Sound Reinforcement 317 Consider the latency (delay) from when a person speaks into a 318 microphone to when their voice emerges from the speaker. If this 319 delay is longer than about 10-15 milliseconds it is noticeable and 320 can make a sound reinforcement system unusable (see slide 6 of 321 [SRP_LATENCY]). (If you have ever tried to speak in the presence of 322 a delayed echo of your voice you may know this experience). 324 Note that the 15ms latency bound includes all parts of the signal 325 path, not just the network, so the network latency must be 326 significantly less than 15ms. 328 In some cases local performers must perform in synchrony with a 329 remote broadcast. In such cases the latencies of the broadcast 330 stream and the local performer must be adjusted to match each other, 331 with a worst case of one video frame (33ms for NTSC video). 333 In cases where audio phase is a consideration, for example beam- 334 forming using multiple speakers, latency requirements can be in the 335 10 microsecond range (1 audio sample at 96kHz). 337 2.1.4. Deterministic Time to Establish Streaming 339 Some audio systems installed in public environments (airports, 340 hospitals) have unique requirements with regards to health, safety 341 and fire concerns. One such requirement is a maximum of 3 seconds 342 for a system to respond to an emergency detection and begin sending 343 appropriate warning signals and alarms without human intervention. 344 For this requirement to be met, the system must support a bounded and 345 acceptable time from a notification signal to specific stream 346 establishment. For further details see [ISO7240-16]. 348 Similar requirements apply when the system is restarted after a power 349 cycle, cable re-connection, or system reconfiguration. 351 In many cases such re-establishment of streaming state must be 352 achieved by the peer devices themselves, i.e. without a central 353 controller (since such a controller may only be present during 354 initial network configuration). 356 Video systems introduce related requirements, for example when 357 transitioning from one camera feed (video stream) to another (see 358 [STUDIO_IP] and [ESPN_DC2]). 360 2.1.5. Secure Transmission 362 2.1.5.1. Safety 364 Professional audio systems can include amplifiers that are capable of 365 generating hundreds or thousands of watts of audio power which if 366 used incorrectly can cause hearing damage to those in the vicinity. 367 Apart from the usual care required by the systems operators to 368 prevent such incidents, the network traffic that controls these 369 devices must be secured (as with any sensitive application traffic). 371 2.1.5.2. Digital Rights Management (DRM) 373 Digital Rights Management (DRM) is very important to the audio and 374 video industries. Any time protected content is introduced into a 375 network there are DRM concerns that must be maintained (see 376 [CONTENT_PROTECTION]). Many aspects of DRM are outside the scope of 377 network technology, however there are cases when a secure link 378 supporting authentication and encryption is required by content 379 owners to carry their audio or video content when it is outside their 380 own secure environment (for example see [DCI]). 382 As an example, two techniques are Digital Transmission Content 383 Protection (DTCP) and High-Bandwidth Digital Content Protection 384 (HDCP). HDCP content is not approved for retransmission within any 385 other type of DRM, while DTCP may be retransmitted under HDCP. 386 Therefore if the source of a stream is outside of the network and it 387 uses HDCP protection it is only allowed to be placed on the network 388 with that same HDCP protection. 390 2.2. Pro Audio Today 392 Some proprietary systems have been created which enable deterministic 393 streams at Layer 3 however they are "engineered networks" which 394 require careful configuration to operate, often require that the 395 system be over-provisioned, and it is implied that all devices on the 396 network voluntarily play by the rules of that network. To enable 397 these industries to successfully transition to an interoperable 398 multi-vendor packet-based infrastructure requires effective open 399 standards, and we believe that establishing relevant IETF standards 400 is a crucial factor. 402 2.3. Pro Audio Future 404 2.3.1. Layer 3 Interconnecting Layer 2 Islands 406 It would be valuable to enable IP to connect multiple Layer 2 LANs. 408 As an example, ESPN recently constructed a state-of-the-art 194,000 409 sq ft, $125 million broadcast studio called DC2. The DC2 network is 410 capable of handling 46 Tbps of throughput with 60,000 simultaneous 411 signals. Inside the facility are 1,100 miles of fiber feeding four 412 audio control rooms (see [ESPN_DC2] ). 414 In designing DC2 they replaced as much point-to-point technology as 415 they could with packet-based technology. They constructed seven 416 individual studios using layer 2 LANS (using IEEE 802.1 AVB) that 417 were entirely effective at routing audio within the LANs. However to 418 interconnect these layer 2 LAN islands together they ended up using 419 dedicated paths in a custom SDN (Software Defined Networking) router 420 because there is no standards-based routing solution available. 422 2.3.2. High Reliability Stream Paths 424 On-air and other live media streams are often backed up with 425 redundant links that seamlessly act to deliver the content when the 426 primary link fails for any reason. In point-to-point systems this is 427 provided by an additional point-to-point link; the analogous 428 requirement in a packet-based system is to provide an alternate path 429 through the network such that no individual link can bring down the 430 system. 432 2.3.3. Link Aggregation 434 For transmitting streams that require more bandwidth than a single 435 link in the target network can support, link aggregation is a 436 technique for combining (aggregating) the bandwidth available on 437 multiple physical links to create a single logical link of the 438 required bandwidth. However, if aggregation is to be used, the 439 network controller (or equivalent) must be able to determine the 440 maximum latency of any path through the aggregate link. 442 2.3.4. Integration of Reserved Streams into IT Networks 444 A commonly cited goal of moving to a packet based media 445 infrastructure is that costs can be reduced by using off the shelf, 446 commodity network hardware. In addition, economy of scale can be 447 realized by combining media infrastructure with IT infrastructure. 448 In keeping with these goals, stream reservation technology should be 449 compatible with existing protocols, and not compromise use of the 450 network for best effort (non-time-sensitive) traffic. 452 2.3.5. Use of Unused Reservations by Best-Effort Traffic 454 In cases where stream bandwidth is reserved but not currently used 455 (or is under-utilized) that bandwidth must be available to best- 456 effort (i.e. non-time-sensitive) traffic. For example a single 457 stream may be nailed up (reserved) for specific media content that 458 needs to be presented at different times of the day, ensuring timely 459 delivery of that content, yet in between those times the full 460 bandwidth of the network can be utilized for best-effort tasks such 461 as file transfers. 463 This also addresses a concern of IT network administrators that are 464 considering adding reserved bandwidth traffic to their networks that 465 ("users will reserve large quantities of bandwidth and then never un- 466 reserve it even though they are not using it, and soon the network 467 will have no bandwidth left"). 469 2.3.6. Traffic Segregation 471 Sink devices may be low cost devices with limited processing power. 472 In order to not overwhelm the CPUs in these devices it is important 473 to limit the amount of traffic that these devices must process. 475 As an example, consider the use of individual seat speakers in a 476 cinema. These speakers are typically required to be cost reduced 477 since the quantities in a single theater can reach hundreds of seats. 478 Discovery protocols alone in a one thousand seat theater can generate 479 enough broadcast traffic to overwhelm a low powered CPU. Thus an 480 installation like this will benefit greatly from some type of traffic 481 segregation that can define groups of seats to reduce traffic within 482 each group. All seats in the theater must still be able to 483 communicate with a central controller. 485 There are many techniques that can be used to support this 486 requirement including (but not limited to) the following examples. 488 2.3.6.1. Packet Forwarding Rules, VLANs and Subnets 490 Packet forwarding rules can be used to eliminate some extraneous 491 streaming traffic from reaching potentially low powered sink devices, 492 however there may be other types of broadcast traffic that should be 493 eliminated using other means for example VLANs or IP subnets. 495 2.3.6.2. Multicast Addressing (IPv4 and IPv6) 497 Multicast addressing is commonly used to keep bandwidth utilization 498 of shared links to a minimum. 500 Because of the MAC Address forwarding nature of Layer 2 bridges it is 501 important that a multicast MAC address is only associated with one 502 stream. This will prevent reservations from forwarding packets from 503 one stream down a path that has no interested sinks simply because 504 there is another stream on that same path that shares the same 505 multicast MAC address. 507 Since each multicast MAC Address can represent 32 different IPv4 508 multicast addresses there must be a process put in place to make sure 509 this does not occur. Requiring use of IPv6 address can achieve this, 510 however due to their continued prevalence, solutions that are 511 effective for IPv4 installations are also required. 513 2.3.7. Latency Optimization by a Central Controller 515 A central network controller might also perform optimizations based 516 on the individual path delays, for example sinks that are closer to 517 the source can inform the controller that they can accept greater 518 latency since they will be buffering packets to match presentation 519 times of farther away sinks. The controller might then move a stream 520 reservation on a short path to a longer path in order to free up 521 bandwidth for other critical streams on that short path. See slides 522 3-5 of [SRP_LATENCY]. 524 Additional optimization can be achieved in cases where sinks have 525 differing latency requirements, for example in a live outdoor concert 526 the speaker sinks have stricter latency requirements than the 527 recording hardware sinks. See slide 7 of [SRP_LATENCY]. 529 2.3.8. Reduced Device Cost Due To Reduced Buffer Memory 531 Device cost can be reduced in a system with guaranteed reservations 532 with a small bounded latency due to the reduced requirements for 533 buffering (i.e. memory) on sink devices. For example, a theme park 534 might broadcast a live event across the globe via a layer 3 protocol; 535 in such cases the size of the buffers required is proportional to the 536 latency bounds and jitter caused by delivery, which depends on the 537 worst case segment of the end-to-end network path. For example on 538 todays open internet the latency is typically unacceptable for audio 539 and video streaming without many seconds of buffering. In such 540 scenarios a single gateway device at the local network that receives 541 the feed from the remote site would provide the expensive buffering 542 required to mask the latency and jitter issues associated with long 543 distance delivery. Sink devices in the local location would have no 544 additional buffering requirements, and thus no additional costs, 545 beyond those required for delivery of local content. The sink device 546 would be receiving the identical packets as those sent by the source 547 and would be unaware that there were any latency or jitter issues 548 along the path. 550 2.4. Pro Audio Asks 552 o Layer 3 routing on top of AVB (and/or other high QoS networks) 554 o Content delivery with bounded, lowest possible latency 556 o IntServ and DiffServ integration with AVB (where practical) 558 o Single network for A/V and IT traffic 560 o Standards-based, interoperable, multi-vendor 562 o IT department friendly 564 o Enterprise-wide networks (e.g. size of San Francisco but not the 565 whole Internet (yet...)) 567 3. Electrical Utilities 569 3.1. Use Case Description 571 Many systems that an electrical utility deploys today rely on high 572 availability and deterministic behavior of the underlying networks. 573 Here we present use cases in Transmission, Generation and 574 Distribution, including key timing and reliability metrics. We also 575 discuss security issues and industry trends which affect the 576 architecture of next generation utility networks 578 3.1.1. Transmission Use Cases 580 3.1.1.1. Protection 582 Protection means not only the protection of human operators but also 583 the protection of the electrical equipment and the preservation of 584 the stability and frequency of the grid. If a fault occurs in the 585 transmission or distribution of electricity then severe damage can 586 occur to human operators, electrical equipment and the grid itself, 587 leading to blackouts. 589 Communication links in conjunction with protection relays are used to 590 selectively isolate faults on high voltage lines, transformers, 591 reactors and other important electrical equipment. The role of the 592 teleprotection system is to selectively disconnect a faulty part by 593 transferring command signals within the shortest possible time. 595 3.1.1.1.1. Key Criteria 597 The key criteria for measuring teleprotection performance are command 598 transmission time, dependability and security. These criteria are 599 defined by the IEC standard 60834 as follows: 601 o Transmission time (Speed): The time between the moment where state 602 changes at the transmitter input and the moment of the 603 corresponding change at the receiver output, including propagation 604 delay. Overall operating time for a teleprotection system 605 includes the time for initiating the command at the transmitting 606 end, the propagation delay over the network (including equipments) 607 and the selection and decision time at the receiving end, 608 including any additional delay due to a noisy environment. 610 o Dependability: The ability to issue and receive valid commands in 611 the presence of interference and/or noise, by minimizing the 612 probability of missing command (PMC). Dependability targets are 613 typically set for a specific bit error rate (BER) level. 615 o Security: The ability to prevent false tripping due to a noisy 616 environment, by minimizing the probability of unwanted commands 617 (PUC). Security targets are also set for a specific bit error 618 rate (BER) level. 620 Additional elements of the the teleprotection system that impact its 621 performance include: 623 o Network bandwidth 625 o Failure recovery capacity (aka resiliency) 627 3.1.1.1.2. Fault Detection and Clearance Timing 629 Most power line equipment can tolerate short circuits or faults for 630 up to approximately five power cycles before sustaining irreversible 631 damage or affecting other segments in the network. This translates 632 to total fault clearance time of 100ms. As a safety precaution, 633 however, actual operation time of protection systems is limited to 634 70- 80 percent of this period, including fault recognition time, 635 command transmission time and line breaker switching time. 637 Some system components, such as large electromechanical switches, 638 require particularly long time to operate and take up the majority of 639 the total clearance time, leaving only a 10ms window for the 640 telecommunications part of the protection scheme, independent of the 641 distance to travel. Given the sensitivity of the issue, new networks 642 impose requirements that are even more stringent: IEC standard 61850 643 limits the transfer time for protection messages to 1/4 - 1/2 cycle 644 or 4 - 8ms (for 60Hz lines) for the most critical messages. 646 3.1.1.1.3. Symmetric Channel Delay 648 Teleprotection channels which are differential must be synchronous, 649 which means that any delays on the transmit and receive paths must 650 match each other. Teleprotection systems ideally support zero 651 asymmetric delay; typical legacy relays can tolerate delay 652 discrepancies of up to 750us. 654 Some tools available for lowering delay variation below this 655 threshold are: 657 o For legacy systems using Time Division Multiplexing (TDM), jitter 658 buffers at the multiplexers on each end of the line can be used to 659 offset delay variation by queuing sent and received packets. The 660 length of the queues must balance the need to regulate the rate of 661 transmission with the need to limit overall delay, as larger 662 buffers result in increased latency. 664 o For jitter-prone IP packet networks, traffic management tools can 665 ensure that the teleprotection signals receive the highest 666 transmission priority to minimize jitter. 668 o Standard packet-based synchronization technologies, such as 669 1588-2008 Precision Time Protocol (PTP) and Synchronous Ethernet 670 (Sync-E), can help keep networks stable by maintaining a highly 671 accurate clock source on the various network devices. 673 3.1.1.1.4. Teleprotection Network Requirements (IEC 61850) 675 The following table captures the main network metrics as based on the 676 IEC 61850 standard. 678 +-----------------------------+-------------------------------------+ 679 | Teleprotection Requirement | Attribute | 680 +-----------------------------+-------------------------------------+ 681 | One way maximum delay | 4-10 ms | 682 | Asymetric delay required | Yes | 683 | Maximum jitter | less than 250 us (750 us for legacy | 684 | | IED) | 685 | Topology | Point to point, point to Multi- | 686 | | point | 687 | Availability | 99.9999 | 688 | precise timing required | Yes | 689 | Recovery time on node | less than 50ms - hitless | 690 | failure | | 691 | performance management | Yes, Mandatory | 692 | Redundancy | Yes | 693 | Packet loss | 0.1% to 1% | 694 +-----------------------------+-------------------------------------+ 696 Table 1: Teleprotection network requirements 698 3.1.1.1.5. Inter-Trip Protection scheme 700 "Inter-tripping" is the signal-controlled tripping of a circuit 701 breaker to complete the isolation of a circuit or piece of apparatus 702 in concert with the tripping of other circuit breakers. 704 +--------------------------------+----------------------------------+ 705 | Inter-Trip protection | Attribute | 706 | Requirement | | 707 +--------------------------------+----------------------------------+ 708 | One way maximum delay | 5 ms | 709 | Asymetric delay required | No | 710 | Maximum jitter | Not critical | 711 | Topology | Point to point, point to Multi- | 712 | | point | 713 | Bandwidth | 64 Kbps | 714 | Availability | 99.9999 | 715 | precise timing required | Yes | 716 | Recovery time on node failure | less than 50ms - hitless | 717 | performance management | Yes, Mandatory | 718 | Redundancy | Yes | 719 | Packet loss | 0.1% | 720 +--------------------------------+----------------------------------+ 722 Table 2: Inter-Trip protection network requirements 724 3.1.1.1.6. Current Differential Protection Scheme 726 Current differential protection is commonly used for line protection, 727 and is typical for protecting parallel circuits. At both end of the 728 lines the current is measured by the differential relays, and both 729 relays will trip the circuit breaker if the current going into the 730 line does not equal the current going out of the line. This type of 731 protection scheme assumes some form of communications being present 732 between the relays at both end of the line, to allow both relays to 733 compare measured current values. Line differential protection 734 schemes assume a very low telecommunications delay between both 735 relays, often as low as 5ms. Moreover, as those systems are often 736 not time-synchronized, they also assume symmetric telecommunications 737 paths with constant delay, which allows comparing current measurement 738 values taken at the exact same time. 740 +----------------------------------+--------------------------------+ 741 | Current Differential protection | Attribute | 742 | Requirement | | 743 +----------------------------------+--------------------------------+ 744 | One way maximum delay | 5 ms | 745 | Asymetric delay Required | Yes | 746 | Maximum jitter | less than 250 us (750us for | 747 | | legacy IED) | 748 | Topology | Point to point, point to | 749 | | Multi-point | 750 | Bandwidth | 64 Kbps | 751 | Availability | 99.9999 | 752 | precise timing required | Yes | 753 | Recovery time on node failure | less than 50ms - hitless | 754 | performance management | Yes, Mandatory | 755 | Redundancy | Yes | 756 | Packet loss | 0.1% | 757 +----------------------------------+--------------------------------+ 759 Table 3: Current Differential Protection metrics 761 3.1.1.1.7. Distance Protection Scheme 763 Distance (Impedance Relay) protection scheme is based on voltage and 764 current measurements. The network metrics are similar (but not 765 identical to) Current Differential protection. 767 +-------------------------------+-----------------------------------+ 768 | Distance protection | Attribute | 769 | Requirement | | 770 +-------------------------------+-----------------------------------+ 771 | One way maximum delay | 5 ms | 772 | Asymetric delay Required | No | 773 | Maximum jitter | Not critical | 774 | Topology | Point to point, point to Multi- | 775 | | point | 776 | Bandwidth | 64 Kbps | 777 | Availability | 99.9999 | 778 | precise timing required | Yes | 779 | Recovery time on node failure | less than 50ms - hitless | 780 | performance management | Yes, Mandatory | 781 | Redundancy | Yes | 782 | Packet loss | 0.1% | 783 +-------------------------------+-----------------------------------+ 785 Table 4: Distance Protection requirements 787 3.1.1.1.8. Inter-Substation Protection Signaling 789 This use case describes the exchange of Sampled Value and/or GOOSE 790 (Generic Object Oriented Substation Events) message between 791 Intelligent Electronic Devices (IED) in two substations for 792 protection and tripping coordination. The two IEDs are in a master- 793 slave mode. 795 The Current Transformer or Voltage Transformer (CT/VT) in one 796 substation sends the sampled analog voltage or current value to the 797 Merging Unit (MU) over hard wire. The MU sends the time-synchronized 798 61850-9-2 sampled values to the slave IED. The slave IED forwards 799 the information to the Master IED in the other substation. The 800 master IED makes the determination (for example based on sampled 801 value differentials) to send a trip command to the originating IED. 802 Once the slave IED/Relay receives the GOOSE trip for breaker 803 tripping, it opens the breaker. It then sends a confirmation message 804 back to the master. All data exchanges between IEDs are either 805 through Sampled Value and/or GOOSE messages. 807 +----------------------------------+--------------------------------+ 808 | Inter-Substation protection | Attribute | 809 | Requirement | | 810 +----------------------------------+--------------------------------+ 811 | One way maximum delay | 5 ms | 812 | Asymetric delay Required | No | 813 | Maximum jitter | Not critical | 814 | Topology | Point to point, point to | 815 | | Multi-point | 816 | Bandwidth | 64 Kbps | 817 | Availability | 99.9999 | 818 | precise timing required | Yes | 819 | Recovery time on node failure | less than 50ms - hitless | 820 | performance management | Yes, Mandatory | 821 | Redundancy | Yes | 822 | Packet loss | 1% | 823 +----------------------------------+--------------------------------+ 825 Table 5: Inter-Substation Protection requirements 827 3.1.1.2. Intra-Substation Process Bus Communications 829 This use case describes the data flow from the CT/VT to the IEDs in 830 the substation via the MU. The CT/VT in the substation send the 831 sampled value (analog voltage or current) to the MU over hard wire. 832 The MU sends the time-synchronized 61850-9-2 sampled values to the 833 IEDs in the substation in GOOSE message format. The GPS Master Clock 834 can send 1PPS or IRIG-B format to the MU through a serial port or 835 IEEE 1588 protocol via a network. Process bus communication using 836 61850 simplifies connectivity within the substation and removes the 837 requirement for multiple serial connections and removes the slow 838 serial bus architectures that are typically used. This also ensures 839 increased flexibility and increased speed with the use of multicast 840 messaging between multiple devices. 842 +----------------------------------+--------------------------------+ 843 | Intra-Substation protection | Attribute | 844 | Requirement | | 845 +----------------------------------+--------------------------------+ 846 | One way maximum delay | 5 ms | 847 | Asymetric delay Required | No | 848 | Maximum jitter | Not critical | 849 | Topology | Point to point, point to | 850 | | Multi-point | 851 | Bandwidth | 64 Kbps | 852 | Availability | 99.9999 | 853 | precise timing required | Yes | 854 | Recovery time on Node failure | less than 50ms - hitless | 855 | performance management | Yes, Mandatory | 856 | Redundancy | Yes - No | 857 | Packet loss | 0.1% | 858 +----------------------------------+--------------------------------+ 860 Table 6: Intra-Substation Protection requirements 862 3.1.1.3. Wide Area Monitoring and Control Systems 864 The application of synchrophasor measurement data from Phasor 865 Measurement Units (PMU) to Wide Area Monitoring and Control Systems 866 promises to provide important new capabilities for improving system 867 stability. Access to PMU data enables more timely situational 868 awareness over larger portions of the grid than what has been 869 possible historically with normal SCADA (Supervisory Control and Data 870 Acquisition) data. Handling the volume and real-time nature of 871 synchrophasor data presents unique challenges for existing 872 application architectures. Wide Area management System (WAMS) makes 873 it possible for the condition of the bulk power system to be observed 874 and understood in real-time so that protective, preventative, or 875 corrective action can be taken. Because of the very high sampling 876 rate of measurements and the strict requirement for time 877 synchronization of the samples, WAMS has stringent telecommunications 878 requirements in an IP network that are captured in the following 879 table: 881 +----------------------+--------------------------------------------+ 882 | WAMS Requirement | Attribute | 883 +----------------------+--------------------------------------------+ 884 | One way maximum | 50 ms | 885 | delay | | 886 | Asymetric delay | No | 887 | Required | | 888 | Maximum jitter | Not critical | 889 | Topology | Point to point, point to Multi-point, | 890 | | Multi-point to Multi-point | 891 | Bandwidth | 100 Kbps | 892 | Availability | 99.9999 | 893 | precise timing | Yes | 894 | required | | 895 | Recovery time on | less than 50ms - hitless | 896 | Node failure | | 897 | performance | Yes, Mandatory | 898 | management | | 899 | Redundancy | Yes | 900 | Packet loss | 1% | 901 +----------------------+--------------------------------------------+ 903 Table 7: WAMS Special Communication Requirements 905 3.1.1.4. IEC 61850 WAN engineering guidelines requirement 906 classification 908 The IEC (International Electrotechnical Commission) has recently 909 published a Technical Report which offers guidelines on how to define 910 and deploy Wide Area Networks for the interconnections of electric 911 substations, generation plants and SCADA operation centers. The IEC 912 61850-90-12 is providing a classification of WAN communication 913 requirements into 4 classes. Table 8 summarizes these requirements: 915 +----------------+------------+------------+------------+-----------+ 916 | WAN | Class WA | Class WB | Class WC | Class WD | 917 | Requirement | | | | | 918 +----------------+------------+------------+------------+-----------+ 919 | Application | EHV (Extra | HV (High | MV (Medium | General | 920 | field | High | Voltage) | Voltage) | purpose | 921 | | Voltage) | | | | 922 | Latency | 5 ms | 10 ms | 100 ms | > 100 ms | 923 | Jitter | 10 us | 100 us | 1 ms | 10 ms | 924 | Latency | 100 us | 1 ms | 10 ms | 100 ms | 925 | Asymetry | | | | | 926 | Time Accuracy | 1 us | 10 us | 100 us | 10 to 100 | 927 | | | | | ms | 928 | Bit Error rate | 10-7 to | 10-5 to | 10-3 | | 929 | | 10-6 | 10-4 | | | 930 | Unavailability | 10-7 to | 10-5 to | 10-3 | | 931 | | 10-6 | 10-4 | | | 932 | Recovery delay | Zero | 50 ms | 5 s | 50 s | 933 | Cyber security | extremely | High | Medium | Medium | 934 | | high | | | | 935 +----------------+------------+------------+------------+-----------+ 937 Table 8: 61850-90-12 Communication Requirements; Courtesy of IEC 939 3.1.2. Generation Use Case 941 The electrical power generation frequency should be maintained within 942 a very narrow band. Deviations from the acceptable frequency range 943 are detected and the required signals are sent to the power plants 944 for frequency regulation. 946 Automatic generation control (AGC) is a system for adjusting the 947 power output of generators at different power plants, in response to 948 changes in the load. 950 +---------------------------------------------------+---------------+ 951 | FCAG (Frequency Control Automatic Generation) | Attribute | 952 | Requirement | | 953 +---------------------------------------------------+---------------+ 954 | One way maximum delay | 500 ms | 955 | Asymetric delay Required | No | 956 | Maximum jitter | Not critical | 957 | Topology | Point to | 958 | | point | 959 | Bandwidth | 20 Kbps | 960 | Availability | 99.999 | 961 | precise timing required | Yes | 962 | Recovery time on Node failure | N/A | 963 | performance management | Yes, | 964 | | Mandatory | 965 | Redundancy | Yes | 966 | Packet loss | 1% | 967 +---------------------------------------------------+---------------+ 969 Table 9: FCAG Communication Requirements 971 3.1.3. Distribution use case 973 3.1.3.1. Fault Location Isolation and Service Restoration (FLISR) 975 Fault Location, Isolation, and Service Restoration (FLISR) refers to 976 the ability to automatically locate the fault, isolate the fault, and 977 restore service in the distribution network. This will likely be the 978 first widespread application of distributed intelligence in the grid. 980 Static power switch status (open/closed) in the network dictates the 981 power flow to secondary substations. Reconfiguring the network in 982 the event of a fault is typically done manually on site to energize/ 983 de-energize alternate paths. Automating the operation of substation 984 switchgear allows the flow of power to be altered automatically under 985 fault conditions. 987 FLISR can be managed centrally from a Distribution Management System 988 (DMS) or executed locally through distributed control via intelligent 989 switches and fault sensors. 991 +----------------------+--------------------------------------------+ 992 | FLISR Requirement | Attribute | 993 +----------------------+--------------------------------------------+ 994 | One way maximum | 80 ms | 995 | delay | | 996 | Asymetric delay | No | 997 | Required | | 998 | Maximum jitter | 40 ms | 999 | Topology | Point to point, point to Multi-point, | 1000 | | Multi-point to Multi-point | 1001 | Bandwidth | 64 Kbps | 1002 | Availability | 99.9999 | 1003 | precise timing | Yes | 1004 | required | | 1005 | Recovery time on | Depends on customer impact | 1006 | Node failure | | 1007 | performance | Yes, Mandatory | 1008 | management | | 1009 | Redundancy | Yes | 1010 | Packet loss | 0.1% | 1011 +----------------------+--------------------------------------------+ 1013 Table 10: FLISR Communication Requirements 1015 3.2. Electrical Utilities Today 1017 Many utilities still rely on complex environments formed of multiple 1018 application-specific proprietary networks, including TDM networks. 1020 In this kind of environment there is no mixing of OT and IT 1021 applications on the same network, and information is siloed between 1022 operational areas. 1024 Specific calibration of the full chain is required, which is costly. 1026 This kind of environment prevents utility operations from realizing 1027 the operational efficiency benefits, visibility, and functional 1028 integration of operational information across grid applications and 1029 data networks. 1031 In addition, there are many security-related issues as discussed in 1032 the following section. 1034 3.2.1. Security Current Practices and Limitations 1036 Grid monitoring and control devices are already targets for cyber 1037 attacks, and legacy telecommunications protocols have many intrinsic 1038 network-related vulnerabilities. For example, DNP3, Modbus, 1039 PROFIBUS/PROFINET, and other protocols are designed around a common 1040 paradigm of request and respond. Each protocol is designed for a 1041 master device such as an HMI (Human Machine Interface) system to send 1042 commands to subordinate slave devices to retrieve data (reading 1043 inputs) or control (writing to outputs). Because many of these 1044 protocols lack authentication, encryption, or other basic security 1045 measures, they are prone to network-based attacks, allowing a 1046 malicious actor or attacker to utilize the request-and-respond system 1047 as a mechanism for command-and-control like functionality. Specific 1048 security concerns common to most industrial control, including 1049 utility telecommunication protocols include the following: 1051 o Network or transport errors (e.g. malformed packets or excessive 1052 latency) can cause protocol failure. 1054 o Protocol commands may be available that are capable of forcing 1055 slave devices into inoperable states, including powering-off 1056 devices, forcing them into a listen-only state, disabling 1057 alarming. 1059 o Protocol commands may be available that are capable of restarting 1060 communications and otherwise interrupting processes. 1062 o Protocol commands may be available that are capable of clearing, 1063 erasing, or resetting diagnostic information such as counters and 1064 diagnostic registers. 1066 o Protocol commands may be available that are capable of requesting 1067 sensitive information about the controllers, their configurations, 1068 or other need-to-know information. 1070 o Most protocols are application layer protocols transported over 1071 TCP; therefore it is easy to transport commands over non-standard 1072 ports or inject commands into authorized traffic flows. 1074 o Protocol commands may be available that are capable of 1075 broadcasting messages to many devices at once (i.e. a potential 1076 DoS). 1078 o Protocol commands may be available to query the device network to 1079 obtain defined points and their values (i.e. a configuration 1080 scan). 1082 o Protocol commands may be available that will list all available 1083 function codes (i.e. a function scan). 1085 These inherent vulnerabilities, along with increasing connectivity 1086 between IT an OT networks, make network-based attacks very feasible. 1088 Simple injection of malicious protocol commands provides control over 1089 the target process. Altering legitimate protocol traffic can also 1090 alter information about a process and disrupt the legitimate controls 1091 that are in place over that process. A man-in-the-middle attack 1092 could provide both control over a process and misrepresentation of 1093 data back to operator consoles. 1095 3.3. Electrical Utilities Future 1097 The business and technology trends that are sweeping the utility 1098 industry will drastically transform the utility business from the way 1099 it has been for many decades. At the core of many of these changes 1100 is a drive to modernize the electrical grid with an integrated 1101 telecommunications infrastructure. However, interoperability 1102 concerns, legacy networks, disparate tools, and stringent security 1103 requirements all add complexity to the grid transformation. Given 1104 the range and diversity of the requirements that should be addressed 1105 by the next generation telecommunications infrastructure, utilities 1106 need to adopt a holistic architectural approach to integrate the 1107 electrical grid with digital telecommunications across the entire 1108 power delivery chain. 1110 The key to modernizing grid telecommunications is to provide a 1111 common, adaptable, multi-service network infrastructure for the 1112 entire utility organization. Such a network serves as the platform 1113 for current capabilities while enabling future expansion of the 1114 network to accommodate new applications and services. 1116 To meet this diverse set of requirements, both today and in the 1117 future, the next generation utility telecommunnications network will 1118 be based on open-standards-based IP architecture. An end-to-end IP 1119 architecture takes advantage of nearly three decades of IP technology 1120 development, facilitating interoperability across disparate networks 1121 and devices, as it has been already demonstrated in many mission- 1122 critical and highly secure networks. 1124 IPv6 is seen as a future telecommunications technology for the Smart 1125 Grid; the IEC (International Electrotechnical Commission) and 1126 different National Committees have mandated a specific adhoc group 1127 (AHG8) to define the migration strategy to IPv6 for all the IEC TC57 1128 power automation standards. 1130 3.3.1. Migration to Packet-Switched Network 1132 Throughout the world, utilities are increasingly planning for a 1133 future based on smart grid applications requiring advanced 1134 telecommunications systems. Many of these applications utilize 1135 packet connectivity for communicating information and control signals 1136 across the utility's Wide Area Network (WAN), made possible by 1137 technologies such as multiprotocol label switching (MPLS). The data 1138 that traverses the utility WAN includes: 1140 o Grid monitoring, control, and protection data 1142 o Non-control grid data (e.g. asset data for condition-based 1143 monitoring) 1145 o Physical safety and security data (e.g. voice and video) 1147 o Remote worker access to corporate applications (voice, maps, 1148 schematics, etc.) 1150 o Field area network backhaul for smart metering, and distribution 1151 grid management 1153 o Enterprise traffic (email, collaboration tools, business 1154 applications) 1156 WANs support this wide variety of traffic to and from substations, 1157 the transmission and distribution grid, generation sites, between 1158 control centers, and between work locations and data centers. To 1159 maintain this rapidly expanding set of applications, many utilities 1160 are taking steps to evolve present time-division multiplexing (TDM) 1161 based and frame relay infrastructures to packet systems. Packet- 1162 based networks are designed to provide greater functionalities and 1163 higher levels of service for applications, while continuing to 1164 deliver reliability and deterministic (real-time) traffic support. 1166 3.3.2. Telecommunications Trends 1168 These general telecommunications topics are in addition to the use 1169 cases that have been addressed so far. These include both current 1170 and future telecommunications related topics that should be factored 1171 into the network architecture and design. 1173 3.3.2.1. General Telecommunications Requirements 1175 o IP Connectivity everywhere 1177 o Monitoring services everywhere and from different remote centers 1179 o Move services to a virtual data center 1181 o Unify access to applications / information from the corporate 1182 network 1184 o Unify services 1186 o Unified Communications Solutions 1188 o Mix of fiber and microwave technologies - obsolescence of SONET/ 1189 SDH or TDM 1191 o Standardize grid telecommunications protocol to opened standard to 1192 ensure interoperability 1194 o Reliable Telecommunications for Transmission and Distribution 1195 Substations 1197 o IEEE 1588 time synchronization Client / Server Capabilities 1199 o Integration of Multicast Design 1201 o QoS Requirements Mapping 1203 o Enable Future Network Expansion 1205 o Substation Network Resilience 1207 o Fast Convergence Design 1209 o Scalable Headend Design 1211 o Define Service Level Agreements (SLA) and Enable SLA Monitoring 1213 o Integration of 3G/4G Technologies and future technologies 1215 o Ethernet Connectivity for Station Bus Architecture 1217 o Ethernet Connectivity for Process Bus Architecture 1219 o Protection, teleprotection and PMU (Phaser Measurement Unit) on IP 1221 3.3.2.2. Specific Network topologies of Smart Grid Applications 1223 Utilities often have very large private telecommunications networks. 1224 It covers an entire territory / country. The main purpose of the 1225 network, until now, has been to support transmission network 1226 monitoring, control, and automation, remote control of generation 1227 sites, and providing FCAPS (Fault, Configuration, Accounting, 1228 Performance, Security) services from centralized network operation 1229 centers. 1231 Going forward, one network will support operation and maintenance of 1232 electrical networks (generation, transmission, and distribution), 1233 voice and data services for ten of thousands of employees and for 1234 exchange with neighboring interconnections, and administrative 1235 services. To meet those requirements, utility may deploy several 1236 physical networks leveraging different technologies across the 1237 country: an optical network and a microwave network for instance. 1238 Each protection and automatism system between two points has two 1239 telecommunications circuits, one on each network. Path diversity 1240 between two substations is key. Regardless of the event type 1241 (hurricane, ice storm, etc.), one path shall stay available so the 1242 system can still operate. 1244 In the optical network, signals are transmitted over more than tens 1245 of thousands of circuits using fiber optic links, microwave and 1246 telephone cables. This network is the nervous system of the 1247 utility's power transmission operations. The optical network 1248 represents ten of thousands of km of cable deployed along the power 1249 lines, with individual runs as long as 280 km. 1251 3.3.2.3. Precision Time Protocol 1253 Some utilities do not use GPS clocks in generation substations. One 1254 of the main reasons is that some of the generation plants are 30 to 1255 50 meters deep under ground and the GPS signal can be weak and 1256 unreliable. Instead, atomic clocks are used. Clocks are 1257 synchronized amongst each other. Rubidium clocks provide clock and 1258 1ms timestamps for IRIG-B. 1260 Some companies plan to transition to the Precision Time Protocol 1261 (PTP, [IEEE1588]), distributing the synchronization signal over the 1262 IP/MPLS network. PTP provides a mechanism for synchronizing the 1263 clocks of participating nodes to a high degree of accuracy and 1264 precision. 1266 PTP operates based on the following assumptions: 1268 It is assumed that the network eliminates cyclic forwarding of PTP 1269 messages within each communication path (e.g. by using a spanning 1270 tree protocol). 1272 PTP is tolerant of an occasional missed message, duplicated 1273 message, or message that arrived out of order. However, PTP 1274 assumes that such impairments are relatively rare. 1276 PTP was designed assuming a multicast communication model, however 1277 PTP also supports a unicast communication model as long as the 1278 behavior of the protocol is preserved. 1280 Like all message-based time transfer protocols, PTP time accuracy 1281 is degraded by delay asymmetry in the paths taken by event 1282 messages. Asymmetry is not detectable by PTP, however, if such 1283 delays are known a priori, PTP can correct for asymmetry. 1285 IEC 61850 will recommend the use of the IEEE PTP 1588 Utility Profile 1286 (as defined in [IEC62439-3:2012] Annex B) which offers the support of 1287 redundant attachment of clocks to Parallel Redundancy Protcol (PRP) 1288 and High-availability Seamless Redundancy (HSR) networks. 1290 3.3.3. Security Trends in Utility Networks 1292 Although advanced telecommunications networks can assist in 1293 transforming the energy industry by playing a critical role in 1294 maintaining high levels of reliability, performance, and 1295 manageability, they also introduce the need for an integrated 1296 security infrastructure. Many of the technologies being deployed to 1297 support smart grid projects such as smart meters and sensors can 1298 increase the vulnerability of the grid to attack. Top security 1299 concerns for utilities migrating to an intelligent smart grid 1300 telecommunications platform center on the following trends: 1302 o Integration of distributed energy resources 1304 o Proliferation of digital devices to enable management, automation, 1305 protection, and control 1307 o Regulatory mandates to comply with standards for critical 1308 infrastructure protection 1310 o Migration to new systems for outage management, distribution 1311 automation, condition-based maintenance, load forecasting, and 1312 smart metering 1314 o Demand for new levels of customer service and energy management 1316 This development of a diverse set of networks to support the 1317 integration of microgrids, open-access energy competition, and the 1318 use of network-controlled devices is driving the need for a converged 1319 security infrastructure for all participants in the smart grid, 1320 including utilities, energy service providers, large commercial and 1321 industrial, as well as residential customers. Securing the assets of 1322 electric power delivery systems (from the control center to the 1323 substation, to the feeders and down to customer meters) requires an 1324 end-to-end security infrastructure that protects the myriad of 1325 telecommunications assets used to operate, monitor, and control power 1326 flow and measurement. 1328 "Cyber security" refers to all the security issues in automation and 1329 telecommunications that affect any functions related to the operation 1330 of the electric power systems. Specifically, it involves the 1331 concepts of: 1333 o Integrity : data cannot be altered undetectably 1335 o Authenticity : the telecommunications parties involved must be 1336 validated as genuine 1338 o Authorization : only requests and commands from the authorized 1339 users can be accepted by the system 1341 o Confidentiality : data must not be accessible to any 1342 unauthenticated users 1344 When designing and deploying new smart grid devices and 1345 telecommunications systems, it is imperative to understand the 1346 various impacts of these new components under a variety of attack 1347 situations on the power grid. Consequences of a cyber attack on the 1348 grid telecommunications network can be catastrophic. This is why 1349 security for smart grid is not just an ad hoc feature or product, 1350 it's a complete framework integrating both physical and Cyber 1351 security requirements and covering the entire smart grid networks 1352 from generation to distribution. Security has therefore become one 1353 of the main foundations of the utility telecom network architecture 1354 and must be considered at every layer with a defense-in-depth 1355 approach. Migrating to IP based protocols is key to address these 1356 challenges for two reasons: 1358 o IP enables a rich set of features and capabilities to enhance the 1359 security posture 1361 o IP is based on open standards, which allows interoperability 1362 between different vendors and products, driving down the costs 1363 associated with implementing security solutions in OT networks. 1365 Securing OT (Operation technology) telecommunications over packet- 1366 switched IP networks follow the same principles that are foundational 1367 for securing the IT infrastructure, i.e., consideration must be given 1368 to enforcing electronic access control for both person-to-machine and 1369 machine-to-machine communications, and providing the appropriate 1370 levels of data privacy, device and platform integrity, and threat 1371 detection and mitigation. 1373 3.4. Electrical Utilities Asks 1375 o Mixed L2 and L3 topologies 1377 o Deterministic behavior 1379 o Bounded latency and jitter 1381 o High availability, low recovery time 1383 o Redundancy, low packet loss 1385 o Precise timing 1387 o Centralized computing of deterministic paths 1389 o Distributed configuration may also be useful 1391 4. Building Automation Systems 1393 4.1. Use Case Description 1395 A Building Automation System (BAS) manages equipment and sensors in a 1396 building for improving residents' comfort, reducing energy 1397 consumption, and responding to failures and emergencies. For 1398 example, the BAS measures the temperature of a room using sensors and 1399 then controls the HVAC (heating, ventilating, and air conditioning) 1400 to maintain a set temperature and minimize energy consumption. 1402 A BAS primarily performs the following functions: 1404 o Periodically measures states of devices, for example humidity and 1405 illuminance of rooms, open/close state of doors, FAN speed, etc. 1407 o Stores the measured data. 1409 o Provides the measured data to BAS systems and operators. 1411 o Generates alarms for abnormal state of devices. 1413 o Controls devices (e.g. turn off room lights at 10:00 PM). 1415 4.2. Building Automation Systems Today 1416 4.2.1. BAS Architecture 1418 A typical BAS architecture of today is shown in Figure 1. 1420 +----------------------------+ 1421 | | 1422 | BMS HMI | 1423 | | | | 1424 | +----------------------+ | 1425 | | Management Network | | 1426 | +----------------------+ | 1427 | | | | 1428 | LC LC | 1429 | | | | 1430 | +----------------------+ | 1431 | | Field Network | | 1432 | +----------------------+ | 1433 | | | | | | 1434 | Dev Dev Dev Dev | 1435 | | 1436 +----------------------------+ 1438 BMS := Building Management Server 1439 HMI := Human Machine Interface 1440 LC := Local Controller 1442 Figure 1: BAS architecture 1444 There are typically two layers of network in a BAS. The upper one is 1445 called the Management Network and the lower one is called the Field 1446 Network. In management networks an IP-based communication protocol 1447 is used, while in field networks non-IP based communication protocols 1448 ("field protocols") are mainly used. Field networks have specific 1449 timing requirements, whereas management networks can be best-effort. 1451 A Human Machine Interface (HMI) is typically a desktop PC used by 1452 operators to monitor and display device states, send device control 1453 commands to Local Controllers (LCs), and configure building schedules 1454 (for example "turn off all room lights in the building at 10:00 PM"). 1456 A Building Management Server (BMS) performs the following operations. 1458 o Collect and store device states from LCs at regular intervals. 1460 o Send control values to LCs according to a building schedule. 1462 o Send an alarm signal to operators if it detects abnormal devices 1463 states. 1465 The BMS and HMI communicate with LCs via IP-based "management 1466 protocols" (see standards [bacnetip], [knx]). 1468 A LC is typically a Programmable Logic Controller (PLC) which is 1469 connected to several tens or hundreds of devices using "field 1470 protocols". An LC performs the following kinds of operations: 1472 o Measure device states and provide the information to BMS or HMI. 1474 o Send control values to devices, unilaterally or as part of a 1475 feedback control loop. 1477 There are many field protocols used today; some are standards-based 1478 and others are proprietary (see standards [lontalk], [modbus], 1479 [profibus] and [flnet]). The result is that BASs have multiple MAC/ 1480 PHY modules and interfaces. This makes BASs more expensive, slower 1481 to develop, and can result in "vendor lock-in" with multiple types of 1482 management applications. 1484 4.2.2. BAS Deployment Model 1486 An example BAS for medium or large buildings is shown in Figure 2. 1487 The physical layout spans multiple floors, and there is a monitoring 1488 room where the BAS management entities are located. Each floor will 1489 have one or more LCs depending upon the number of devices connected 1490 to the field network. 1492 +--------------------------------------------------+ 1493 | Floor 3 | 1494 | +----LC~~~~+~~~~~+~~~~~+ | 1495 | | | | | | 1496 | | Dev Dev Dev | 1497 | | | 1498 |--- | ------------------------------------------| 1499 | | Floor 2 | 1500 | +----LC~~~~+~~~~~+~~~~~+ Field Network | 1501 | | | | | | 1502 | | Dev Dev Dev | 1503 | | | 1504 |--- | ------------------------------------------| 1505 | | Floor 1 | 1506 | +----LC~~~~+~~~~~+~~~~~+ +-----------------| 1507 | | | | | | Monitoring Room | 1508 | | Dev Dev Dev | | 1509 | | | BMS HMI | 1510 | | Management Network | | | | 1511 | +--------------------------------+-----+ | 1512 | | | 1513 +--------------------------------------------------+ 1515 Figure 2: BAS Deployment model for Medium/Large Buildings 1517 Each LC is connected to the monitoring room via the Management 1518 network, and the management functions are performed within the 1519 building. In most cases, fast Ethernet (e.g. 100BASE-T) is used for 1520 the management network. Since the management network is non- 1521 realtime, use of Ethernet without quality of service is sufficient 1522 for today's deployment. 1524 In the field network a variety of physical interfaces such as RS232C 1525 and RS485 are used, which have specific timing requirements. Thus if 1526 a field network is to be replaced with an Ethernet or wireless 1527 network, such networks must support time-critical deterministic 1528 flows. 1530 In Figure 3, another deployment model is presented in which the 1531 management system is hosted remotely. This is becoming popular for 1532 small office and residential buildings in which a standalone 1533 monitoring system is not cost-effective. 1535 +---------------+ 1536 | Remote Center | 1537 | | 1538 | BMS HMI | 1539 +------------------------------------+ | | | | 1540 | Floor 2 | | +---+---+ | 1541 | +----LC~~~~+~~~~~+ Field Network| | | | 1542 | | | | | | Router | 1543 | | Dev Dev | +-------|-------+ 1544 | | | | 1545 |--- | ------------------------------| | 1546 | | Floor 1 | | 1547 | +----LC~~~~+~~~~~+ | | 1548 | | | | | | 1549 | | Dev Dev | | 1550 | | | | 1551 | | Management Network | WAN | 1552 | +------------------------Router-------------+ 1553 | | 1554 +------------------------------------+ 1556 Figure 3: Deployment model for Small Buildings 1558 Some interoperability is possible today in the Management Network, 1559 but not in today's field networks due to their non-IP-based design. 1561 4.2.3. Use Cases for Field Networks 1563 Below are use cases for Environmental Monitoring, Fire Detection, and 1564 Feedback Control, and their implications for field network 1565 performance. 1567 4.2.3.1. Environmental Monitoring 1569 The BMS polls each LC at a maximum measurement interval of 100ms (for 1570 example to draw a historical chart of 1 second granularity with a 10x 1571 sampling interval) and then performs the operations as specified by 1572 the operator. Each LC needs to measure each of its several hundred 1573 sensors once per measurement interval. Latency is not critical in 1574 this scenario as long as all sensor values are completed in the 1575 measurement interval. Availability is expected to be 99.999 %. 1577 4.2.3.2. Fire Detection 1579 On detection of a fire, the BMS must stop the HVAC, close the fire 1580 shutters, turn on the fire sprinklers, send an alarm, etc. There are 1581 typically ~10s of sensors per LC that BMS needs to manage. In this 1582 scenario the measurement interval is 10-50ms, the communication delay 1583 is 10ms, and the availability must be 99.9999 %. 1585 4.2.3.3. Feedback Control 1587 BAS systems utilize feedback control in various ways; the most time- 1588 critial is control of DC motors, which require a short feedback 1589 interval (1-5ms) with low communication delay (10ms) and jitter 1590 (1ms). The feedback interval depends on the characteristics of the 1591 device and a target quality of control value. There are typically 1592 ~10s of such devices per LC. 1594 Communication delay is expected to be less than 10 ms, jitter less 1595 than 1 sec while the availability must be 99.9999% . 1597 4.2.4. Security Considerations 1599 When BAS field networks were developed it was assumed that the field 1600 networks would always be physically isolated from external networks 1601 and therefore security was not a concern. In today's world many BASs 1602 are managed remotely and are thus connected to shared IP networks and 1603 so security is definitely a concern, yet security features are not 1604 available in the majority of BAS field network deployments . 1606 The management network, being an IP-based network, has the protocols 1607 available to enable network security, but in practice many BAS 1608 systems do not implement even the available security features such as 1609 device authentication or encryption for data in transit. 1611 4.3. BAS Future 1613 In the future we expect more fine-grained environmental monitoring 1614 and lower energy consumption, which will require more sensors and 1615 devices, thus requiring larger and more complex building networks. 1617 We expect building networks to be connected to or converged with 1618 other networks (Enterprise network, Home network, and Internet). 1620 Therefore better facilities for network management, control, 1621 reliability and security are critical in order to improve resident 1622 and operator convenience and comfort. For example the ability to 1623 monitor and control building devices via the internet would enable 1624 (for example) control of room lights or HVAC from a resident's 1625 desktop PC or phone application. 1627 4.4. BAS Asks 1629 The community would like to see an interoperable protocol 1630 specification that can satisfy the timing, security, availability and 1631 QoS constraints described above, such that the resulting converged 1632 network can replace the disparate field networks. Ideally this 1633 connectivity could extend to the open Internet. 1635 This would imply an architecture that can guarantee 1637 o Low communication delays (from <10ms to 100ms in a network of 1638 several hundred devices) 1640 o Low jitter (< 1 ms) 1642 o Tight feedback intervals (1ms - 10ms) 1644 o High network availability (up to 99.9999% ) 1646 o Availability of network data in disaster scenario 1648 o Authentication between management and field devices (both local 1649 and remote) 1651 o Integrity and data origin authentication of communication data 1652 between field and management devices 1654 o Confidentiality of data when communicated to a remote device 1656 5. Wireless for Industrial 1658 5.1. Use Case Description 1660 Wireless networks are useful for industrial applications, for example 1661 when portable, fast-moving or rotating objects are involved, and for 1662 the resource-constrained devices found in the Internet of Things 1663 (IoT). 1665 Such network-connected sensors, actuators, control loops (etc.) 1666 typically require that the underlying network support real-time 1667 quality of service (QoS), as well as specific classes of other 1668 network properties such as reliability, redundancy, and security. 1670 These networks may also contain very large numbers of devices, for 1671 example for factories, "big data" acquisition, and the IoT. Given 1672 the large numbers of devices installed, and the potential 1673 pervasiveness of the IoT, this is a huge and very cost-sensitive 1674 market. For example, a 1% cost reduction in some areas could save 1675 $100B 1677 5.1.1. Network Convergence using 6TiSCH 1679 Some wireless network technologies support real-time QoS, and are 1680 thus useful for these kinds of networks, but others do not. For 1681 example WiFi is pervasive but does not provide guaranteed timing or 1682 delivery of packets, and thus is not useful in this context. 1684 In this use case we focus on one specific wireless network technology 1685 which does provide the required deterministic QoS, which is "IPv6 1686 over the TSCH mode of IEEE 802.15.4e" (6TiSCH, where TSCH stands for 1687 "Time-Slotted Channel Hopping", see [I-D.ietf-6tisch-architecture], 1688 [IEEE802154], [IEEE802154e], and [RFC7554]). 1690 There are other deterministic wireless busses and networks available 1691 today, however they are imcompatible with each other, and 1692 incompatible with IP traffic (for example [ISA100], [WirelessHART]). 1694 Thus the primary goal of this use case is to apply 6TiSH as a 1695 converged IP- and standards-based wireless network for industrial 1696 applications, i.e. to replace multiple proprietary and/or 1697 incompatible wireless networking and wireless network management 1698 standards. 1700 5.1.2. Common Protocol Development for 6TiSCH 1702 Today there are a number of protocols required by 6TiSCH which are 1703 still in development, and a second intent of this use case is to 1704 highlight the ways in which these "missing" protocols share goals in 1705 common with DetNet. Thus it is possible that some of the protocol 1706 technology developed for DetNet will also be applicable to 6TiSCH. 1708 These protocol goals are identified here, along with their 1709 relationship to DetNet. It is likely that ultimately the resulting 1710 protocols will not be identical, but will share design principles 1711 which contribute to the eficiency of enabling both DetNet and 6TiSCH. 1713 One such commonality is that although at a different time scale, in 1714 both TSN [IEEE802.1TSNTG] and TSCH a packet crosses the network from 1715 node to node follows a precise schedule, as a train that leaves 1716 intermediate stations at precise times along its path. This kind of 1717 operation reduces collisions, saves energy, and enables engineering 1718 the network for deterministic properties. 1720 Another commonality is remote monitoring and scheduling management of 1721 a TSCH network by a Path Computation Element (PCE) and Network 1722 Management Entity (NME). The PCE/NME manage timeslots and device 1723 resources in a manner that minimizes the interaction with and the 1724 load placed on resource-constrained devices. For example, a tiny IoT 1725 device may have just enough buffers to store one or a few IPv6 1726 packets, and will have limited bandwidth between peers such that it 1727 can maintain only a small amount of peer information, and will not be 1728 able to store many packets waiting to be forwarded. It is 1729 advantageous then for it to only be required to carry out the 1730 specific behavior assigned to it by the PCE/NME (as opposed to 1731 maintaining its own IP stack, for example). 1733 6TiSCH depends on [PCE] and [I-D.finn-detnet-architecture], and we 1734 expect that DetNet will maintain consistency with [IEEE802.1TSNTG]. 1736 5.2. Wireless Industrial Today 1738 Today industrial wireless is accomplished using multiple 1739 deterministic wireless networks which are incompatible with each 1740 other and with IP traffic. 1742 6TiSCH is not yet fully specified, so it cannot be used in today's 1743 applications. 1745 5.3. Wireless Industrial Future 1747 5.3.1. Unified Wireless Network and Management 1749 We expect DetNet and 6TiSCH together to enable converged transport of 1750 deterministic and best-effort traffic flows between real-time 1751 industrial devices and wide area networks via IP routing. A high 1752 level view of a basic such network is shown in Figure 4. 1754 ---+-------- ............ ------------ 1755 | External Network | 1756 | +-----+ 1757 +-----+ | NME | 1758 | | LLN Border | | 1759 | | router +-----+ 1760 +-----+ 1761 o o o 1762 o o o o 1763 o o LLN o o o 1764 o o o o 1765 o 1767 Figure 4: Basic 6TiSCH Network 1769 Figure 5 shows a backbone router federating multiple synchronized 1770 6TiSCH subnets into a single subnet connected to the external 1771 network. 1773 ---+-------- ............ ------------ 1774 | External Network | 1775 | +-----+ 1776 | +-----+ | NME | 1777 +-----+ | +-----+ | | 1778 | | Router | | PCE | +-----+ 1779 | | +--| | 1780 +-----+ +-----+ 1781 | | 1782 | Subnet Backbone | 1783 +--------------------+------------------+ 1784 | | | 1785 +-----+ +-----+ +-----+ 1786 | | Backbone | | Backbone | | Backbone 1787 o | | router | | router | | router 1788 +-----+ +-----+ +-----+ 1789 o o o o o 1790 o o o o o o o o o o o 1791 o o o LLN o o o o 1792 o o o o o o o o o o o o 1794 Figure 5: Extended 6TiSCH Network 1796 The backbone router must ensure end-to-end deterministic behavior 1797 between the LLN and the backbone. We would like to see this 1798 accomplished in conformance with the work done in 1799 [I-D.finn-detnet-architecture] with respect to Layer-3 aspects of 1800 deterministic networks that span multiple Layer-2 domains. 1802 The PCE must compute a deterministic path end-to-end across the TSCH 1803 network and IEEE802.1 TSN Ethernet backbone, and DetNet protocols are 1804 expected to enable end-to-end deterministic forwarding. 1806 +-----+ 1807 | IoT | 1808 | G/W | 1809 +-----+ 1810 ^ <---- Elimination 1811 | | 1812 Track branch | | 1813 +-------+ +--------+ Subnet Backbone 1814 | | 1815 +--|--+ +--|--+ 1816 | | | Backbone | | | Backbone 1817 o | | | router | | | router 1818 +--/--+ +--|--+ 1819 o / o o---o----/ o 1820 o o---o--/ o o o o o 1821 o \ / o o LLN o 1822 o v <---- Replication 1823 o 1825 Figure 6: 6TiSCH Network with PRE 1827 5.3.1.1. PCE and 6TiSCH ARQ Retries 1829 6TiSCH uses the IEEE802.15.4 Automatic Repeat-reQuest (ARQ) mechanism 1830 to provide higher reliability of packet delivery. ARQ is related to 1831 packet replication and elimination because there are two independent 1832 paths for packets to arrive at the destination, and if an expected 1833 packed does not arrive on one path then it checks for the packet on 1834 the second path. 1836 Although to date this mechanism is only used by wireless networks, 1837 this may be a technique that would be appropriate for DetNet and so 1838 aspects of the enabling protocol could be co-developed. 1840 For example, in Figure 6, a Track is laid out from a field device in 1841 a 6TiSCH network to an IoT gateway that is located on a IEEE802.1 TSN 1842 backbone. 1844 The Replication function in the field device sends a copy of each 1845 packet over two different branches, and the PCE schedules each hop of 1846 both branches so that the two copies arrive in due time at the 1847 gateway. In case of a loss on one branch, hopefully the other copy 1848 of the packet still arrives within the allocated time. If two copies 1849 make it to the IoT gateway, the Elimination function in the gateway 1850 ignores the extra packet and presents only one copy to upper layers. 1852 At each 6TiSCH hop along the Track, the PCE may schedule more than 1853 one timeSlot for a packet, so as to support Layer-2 retries (ARQ). 1855 In current deployments, a TSCH Track does not necessarily support PRE 1856 but is systematically multi-path. This means that a Track is 1857 scheduled so as to ensure that each hop has at least two forwarding 1858 solutions, and the forwarding decision is to try the preferred one 1859 and use the other in case of Layer-2 transmission failure as detected 1860 by ARQ. 1862 5.3.2. Schedule Management by a PCE 1864 A common feature of 6TiSCH and DetNet is the action of a PCE to 1865 configure paths through the network. Specifically, what is needed is 1866 a protocol and data model that the PCE will use to get/set the 1867 relevant configuration from/to the devices, as well as perform 1868 operations on the devices. We expect that this protocol will be 1869 developed by DetNet with consideration for its reuse by 6TiSCH. The 1870 remainder of this section provides a bit more context from the 6TiSCH 1871 side. 1873 5.3.2.1. PCE Commands and 6TiSCH CoAP Requests 1875 The 6TiSCH device does not expect to place the request for bandwidth 1876 between itself and another device in the network. Rather, an 1877 operation control system invoked through a human interface specifies 1878 the required traffic specification and the end nodes (in terms of 1879 latency and reliability). Based on this information, the PCE must 1880 compute a path between the end nodes and provision the network with 1881 per-flow state that describes the per-hop operation for a given 1882 packet, the corresponding timeslots, and the flow identification that 1883 enables recognizing that a certain packet belongs to a certain path, 1884 etc. 1886 For a static configuration that serves a certain purpose for a long 1887 period of time, it is expected that a node will be provisioned in one 1888 shot with a full schedule, which incorporates the aggregation of its 1889 behavior for multiple paths. 6TiSCH expects that the programing of 1890 the schedule will be done over COAP as discussed in 1891 [I-D.ietf-6tisch-coap]. 1893 6TiSCH expects that the PCE commands will be issued directly as CoAP 1894 requests or be mapped back and forth into CoAP by a gateway function 1895 at the edge of the 6TiSCH network. For instance, it is possible that 1896 a mapping entity on the backbone transforms a non-CoAP protocol such 1897 as PCEP into the RESTful interfaces that the 6TiSCH devices support. 1898 This architecture will be refined to comply with DetNet 1899 [I-D.finn-detnet-architecture] when the work is formalized. Related 1900 information about 6TiSCH can be found at 1901 [I-D.ietf-6tisch-6top-interface] and RPL [RFC6550]. 1903 If it appears that a path through the network does not perform as 1904 expected, a protocol may be used to update the state in the devices, 1905 but in 6TiSCH that flow was not designed and no protocol was selected 1906 and it is expected that DetNet will determine the appropriate end-to- 1907 end protocols to be used in that case. 1909 A "slotFrame" is the base object that the PCE needs to manipulate to 1910 program a schedule into an LLN node ([I-D.ietf-6tisch-architecture]). 1912 The PCE should be able to read energy data from devices, and compute 1913 paths that will implement policies on how energy in devices is 1914 consumed, for instance to ensure that the spent energy does not 1915 exceeded the available energy over a period of time. 1917 6TiSCH devices can discover their neighbors over the radio using a 1918 mechanism such as beacons, but even though the neighbor information 1919 is available in the 6TiSCH interface data model, 6TiSCH does not 1920 describe a protocol to proactively push the neighborhood information 1921 to a PCE. DetNet should define this protocol, and it and should 1922 operate over CoAP. The protocol should be able to carry multiple 1923 metrics, in particular the same metrics as used for RPL operations 1924 [RFC6551] 1926 5.3.2.2. 6TiSCH IP Interface 1928 "6top" ([I-D.wang-6tisch-6top-sublayer]) is a logical link control 1929 sitting between the IP layer and the TSCH MAC layer which provides 1930 the link abstraction that is required for IP operations. The 6top 1931 data model and management interfaces are further discussed in 1932 [I-D.ietf-6tisch-6top-interface] and [I-D.ietf-6tisch-coap]. 1934 An IP packet that is sent along a 6TiSCH path uses the Differentiated 1935 Services Per-Hop-Behavior Group called Deterministic Forwarding, as 1936 described in [I-D.svshah-tsvwg-deterministic-forwarding]. 1938 5.3.3. 6TiSCH Security Considerations 1940 On top of the classical requirements for protection of control 1941 signaling, it must be noted that 6TiSCH networks operate on limited 1942 resources that can be depleted rapidly in a DoS attack on the system, 1943 for instance by placing a rogue device in the network, or by 1944 obtaining management control and setting up unexpected additional 1945 paths. 1947 5.4. Wireless Industrial Asks 1949 6TiSCH depends on DetNet to define: 1951 o Configuration (state) and operations for deterministic paths 1953 o End-to-end protocols for deterministic forwarding (tagging, IP) 1955 o Protocol for packet replication and elimination 1957 o Protocol for packet automatic retries (ARQ) (specific to wireless) 1959 6. Cellular Radio Use Cases 1961 6.1. Use Case Description 1963 This use case describes the application of deterministic networking 1964 in the context of cellular telecom transport networks. Important 1965 elements include time synchronization, clock distribution, and ways 1966 of establishing time-sensitive streams for both Layer-2 and Layer-3 1967 user plane traffic. 1969 6.1.1. Network Architecture 1971 Figure 7 illustrates a typical 3GPP-defined cellular network 1972 architecture, which includes "Fronthaul" and "Midhaul" network 1973 segments. The "Fronthaul" is the network connecting base stations 1974 (baseband processing units) to the remote radio heads (antennas). 1975 The "Midhaul" is the network inter-connecting base stations (or small 1976 cell sites). 1978 In Figure 7 "eNB" ("E-UTRAN Node B") is the hardware that is 1979 connected to the mobile phone network which communicates directly 1980 with mobile handsets ([TS36300]). 1982 Y (remote radio heads (antennas)) 1983 \ 1984 Y__ \.--. .--. +------+ 1985 \_( `. +---+ _(Back`. | 3GPP | 1986 Y------( Front )----|eNB|----( Haul )----| core | 1987 ( ` .Haul ) +---+ ( ` . ) ) | netw | 1988 /`--(___.-' \ `--(___.-' +------+ 1989 Y_/ / \.--. \ 1990 Y_/ _( Mid`. \ 1991 ( Haul ) \ 1992 ( ` . ) ) \ 1993 `--(___.-'\_____+---+ (small cell sites) 1994 \ |SCe|__Y 1995 +---+ +---+ 1996 Y__|eNB|__Y 1997 +---+ 1998 Y_/ \_Y ("local" radios) 2000 Figure 7: Generic 3GPP-based Cellular Network Architecture 2002 The available processing time for Fronthaul networking overhead is 2003 limited to the available time after the baseband processing of the 2004 radio frame has completed. For example in Long Term Evolution (LTE) 2005 radio, processing of a radio frame is allocated 3ms, but typically 2006 the processing completes much earlier (<400us) allowing the remaining 2007 time to be used by the Fronthaul network. This ultimately determines 2008 the distance the remote radio heads can be located from the base 2009 stations (200us equals roughly 40 km of optical fiber-based 2010 transport, thus round trip time is 2*200us = 400us). 2012 The remainder of the "maximum delay budget" is consumed by all nodes 2013 and buffering between the remote radio head and the baseband 2014 processing, plus the distance-incurred delay. 2016 The baseband processing time and the available "delay budget" for the 2017 fronthaul is likely to change in the forthcoming "5G" due to reduced 2018 radio round trip times and other architectural and service 2019 requirements [NGMN]. 2021 6.1.2. Time Synchronization Requirements 2023 Fronthaul time synchronization requirements are given by [TS25104], 2024 [TS36104], [TS36211], and [TS36133]. These can be summarized for the 2025 current 3GPP LTE-based networks as: 2027 Delay Accuracy: 2028 +-8ns (i.e. +-1/32 Tc, where Tc is the UMTS Chip time of 1/3.84 2029 MHz) resulting in a round trip accuracy of +-16ns. The value is 2030 this low to meet the 3GPP Timing Alignment Error (TAE) measurement 2031 requirements. 2033 Packet Delay Variation: 2034 Packet Delay Variation (PDV aka Jitter aka Timing Alignment Error) 2035 is problematic to Fronthaul networks and must be minimized. If 2036 the transport network cannot guarantee low enough PDV then 2037 additional buffering has to be introduced at the edges of the 2038 network to buffer out the jitter. Buffering is not desirable as 2039 it reduces the total available delay budget. 2041 * For multiple input multiple output (MIMO) or TX diversity 2042 transmissions, at each carrier frequency, TAE shall not exceed 2043 65 ns (i.e. 1/4 Tc). 2045 * For intra-band contiguous carrier aggregation, with or without 2046 MIMO or TX diversity, TAE shall not exceed 130 ns (i.e. 1/2 2047 Tc). 2049 * For intra-band non-contiguous carrier aggregation, with or 2050 without MIMO or TX diversity, TAE shall not exceed 260 ns (i.e. 2051 one Tc). 2053 * For inter-band carrier aggregation, with or without MIMO or TX 2054 diversity, TAE shall not exceed 260 ns. 2056 Transport link contribution to radio frequency error: 2057 +-2 PPB. This value is considered to be "available" for the 2058 Fronthaul link out of the total 50 PPB budget reserved for the 2059 radio interface. Note: the reason that the transport link 2060 contributes to radio frequency error is as follows. The current 2061 way of doing Fronthaul is from the radio unit to remote radio head 2062 directly. The remote radio head is essentially a passive device 2063 (without buffering etc.) The transport drives the antenna 2064 directly by feeding it with samples and everything the transport 2065 adds will be introduced to radio as-is. So if the transport 2066 causes additional frequence error that shows immediately on the 2067 radio as well. 2069 The above listed time synchronization requirements are difficult to 2070 meet with point-to-point connected networks, and more difficult when 2071 the network includes multiple hops. It is expected that networks 2072 must include buffering at the ends of the connections as imposed by 2073 the jitter requirements, since trying to meet the jitter requirements 2074 in every intermediate node is likely to be too costly. However, 2075 every measure to reduce jitter and delay on the path makes it easier 2076 to meet the end-to-end requirements. 2078 In order to meet the timing requirements both senders and receivers 2079 must remain time synchronized, demanding very accurate clock 2080 distribution, for example support for IEEE 1588 transparent clocks in 2081 every intermediate node. 2083 In cellular networks from the LTE radio era onward, phase 2084 synchronization is needed in addition to frequency synchronization 2085 ([TS36300], [TS23401]). 2087 6.1.3. Time-Sensitive Stream Requirements 2089 In addition to the time synchronization requirements listed in 2090 Section Section 6.1.2 the Fronthaul networks assume practically 2091 error-free transport. The maximum bit error rate (BER) has been 2092 defined to be 10^-12. When packetized that would imply a packet 2093 error rate (PER) of 2.4*10^-9 (assuming ~300 bytes packets). 2094 Retransmitting lost packets and/or using forward error correction 2095 (FEC) to circumvent bit errors is practically impossible due to the 2096 additional delay incurred. Using redundant streams for better 2097 guarantees for delivery is also practically impossible in many cases 2098 due to high bandwidth requirements of Fronthaul networks. For 2099 instance, current uncompressed CPRI bandwidth expansion ratio is 2100 roughly 20:1 compared to the IP layer user payload it carries. 2101 Protection switching is also a candidate but current technologies for 2102 the path switch are too slow. We do not currently know of a better 2103 solution for this issue. 2105 Fronthaul links are assumed to be symmetric, and all Fronthaul 2106 streams (i.e. those carrying radio data) have equal priority and 2107 cannot delay or pre-empt each other. This implies that the network 2108 must guarantee that each time-sensitive flow meets their schedule. 2110 6.1.4. Security Considerations 2112 Establishing time-sensitive streams in the network entails reserving 2113 networking resources for long periods of time. It is important that 2114 these reservation requests be authenticated to prevent malicious 2115 reservation attempts from hostile nodes (or accidental 2116 misconfiguration). This is particularly important in the case where 2117 the reservation requests span administrative domains. Furthermore, 2118 the reservation information itself should be digitally signed to 2119 reduce the risk of a legitimate node pushing a stale or hostile 2120 configuration into another networking node. 2122 6.2. Cellular Radio Networks Today 2124 Today's Fronthaul networks typically consist of: 2126 o Dedicated point-to-point fiber connection is common 2128 o Proprietary protocols and framings 2130 o Custom equipment and no real networking 2132 Today's Midhaul and Backhaul networks typically consist of: 2134 o Mostly normal IP networks, MPLS-TP, etc. 2136 o Clock distribution and sync using 1588 and SyncE 2138 Telecommunication networks in the cellular domain are already heading 2139 towards transport networks where precise time synchronization support 2140 is one of the basic building blocks. While the transport networks 2141 themselves have practically transitioned to all-IP packet based 2142 networks to meet the bandwidth and cost requirements, highly accurate 2143 clock distribution has become a challenge. 2145 Transport networks in the cellular domain are typically based on Time 2146 Division Multiplexing (TDM-based) and provide frequency 2147 synchronization capabilities as a part of the transport media. 2148 Alternatively other technologies such as Global Positioning System 2149 (GPS) or Synchronous Ethernet (SyncE) are used [SyncE]. 2151 Both Ethernet and IP/MPLS [RFC3031] (and PseudoWires (PWE) [RFC3985] 2152 for legacy transport support) have become popular tools to build and 2153 manage new all-IP Radio Access Networks (RAN) 2154 [I-D.kh-spring-ip-ran-use-case]. Although various timing and 2155 synchronization optimizations have already been proposed and 2156 implemented including 1588 PTP enhancements 2157 [I-D.ietf-tictoc-1588overmpls][I-D.mirsky-mpls-residence-time], these 2158 solution are not necessarily sufficient for the forthcoming RAN 2159 architectures or guarantee the higher time-synchronization 2160 requirements [CPRI]. There are also existing solutions for the TDM 2161 over IP [RFC5087] [RFC4553] or Ethernet transports [RFC5086]. 2163 6.3. Cellular Radio Networks Future 2165 We would like to see the following in future Cellular Radio networks: 2167 o Unified standards-based transport protocols and standard 2168 networking equipment that can make use of underlying deterministic 2169 link-layer services 2171 o Unified and standards-based network management systems and 2172 protocols in all parts of the network (including Fronthaul) 2174 New radio access network deployment models and architectures may 2175 require time sensitive networking services with strict requirements 2176 on other parts of the network that previously were not considered to 2177 be packetized at all. The time and synchronization support are 2178 already topical for Backhaul and Midhaul packet networks [MEF], and 2179 becoming a real issue for Fronthaul networks. Specifically in the 2180 Fronthaul networks the timing and synchronization requirements can be 2181 extreme for packet based technologies, for example, on the order of 2182 sub +-20 ns packet delay variation (PDV) and frequency accuracy of 2183 +0.002 PPM [Fronthaul]. 2185 The actual transport protocols and/or solutions to establish required 2186 transport "circuits" (pinned-down paths) for Fronthaul traffic are 2187 still undefined. Those are likely to include (but are not limited 2188 to) solutions directly over Ethernet, over IP, and MPLS/PseudoWire 2189 transport. 2191 Even the current time-sensitive networking features may not be 2192 sufficient for Fronthaul traffic. Therefore, having specific 2193 profiles that take the requirements of Fronthaul into account is 2194 desirable [IEEE8021CM]. 2196 The really interesting and important existing work for time sensitive 2197 networking has been done for Ethernet [TSNTG], which specifies the 2198 use of IEEE 1588 time precision protocol (PTP) [IEEE1588] in the 2199 context of IEEE 802.1D and IEEE 802.1Q. While IEEE 802.1AS 2200 [IEEE8021AS] specifies a Layer-2 time synchronizing service other 2201 specification, such as IEEE 1722 [IEEE1722] specify Ethernet-based 2202 Layer-2 transport for time-sensitive streams. New promising work 2203 seeks to enable the transport of time-sensitive fronthaul streams in 2204 Ethernet bridged networks [IEEE8021CM]. Similarly to IEEE 1722 there 2205 is an ongoing standardization effort to define Layer-2 transport 2206 encapsulation format for transporting radio over Ethernet (RoE) in 2207 IEEE 1904.3 Task Force [IEEE19043]. 2209 All-IP RANs and various "haul" networks would benefit from time 2210 synchronization and time-sensitive transport services. Although 2211 Ethernet appears to be the unifying technology for the transport 2212 there is still a disconnect providing Layer-3 services. The protocol 2213 stack typically has a number of layers below the Ethernet Layer-2 2214 that shows up to the Layer-3 IP transport. It is not uncommon that 2215 on top of the lowest layer (optical) transport there is the first 2216 layer of Ethernet followed one or more layers of MPLS, PseudoWires 2217 and/or other tunneling protocols finally carrying the Ethernet layer 2218 visible to the user plane IP traffic. While there are existing 2219 technologies, especially in MPLS/PWE space, to establish circuits 2220 through the routed and switched networks, there is a lack of 2221 signaling the time synchronization and time-sensitive stream 2222 requirements/reservations for Layer-3 flows in a way that the entire 2223 transport stack is addressed and the Ethernet layers that needs to be 2224 configured are addressed. 2226 Furthermore, not all "user plane" traffic will be IP. Therefore, the 2227 same solution also must address the use cases where the user plane 2228 traffic is again another layer or Ethernet frames. There is existing 2229 work describing the problem statement 2230 [I-D.finn-detnet-problem-statement] and the architecture 2231 [I-D.finn-detnet-architecture] for deterministic networking (DetNet) 2232 that targets solutions for time-sensitive (IP/transport) streams with 2233 deterministic properties over Ethernet-based switched networks. 2235 6.4. Cellular Radio Networks Asks 2237 A standard for data plane transport specification which is: 2239 o Unified among all *hauls 2241 o Deployed in a highly deterministic network environment 2243 A standard for data flow information models that are: 2245 o Aware of the time sensitivity and constraints of the target 2246 networking environment 2248 o Aware of underlying deterministic networking services (e.g. on the 2249 Ethernet layer) 2251 Mapping the Fronthaul requirements to IETF DetNet 2252 [I-D.finn-detnet-architecture] Section 3 "Providing the DetNet 2253 Quality of Service", the relevant features are: 2255 o Zero congestion loss. 2257 o Pinned-down paths. 2259 7. Cellular Coordinated Multipoint Processing (CoMP) 2261 7.1. Use Case Description 2263 In cellular wireless communication systems, Inter-Site Coordinated 2264 Multipoint Processing (CoMP, see [CoMP]) is a technique implemented 2265 within a cell site which improves system efficiency and user quality 2266 experience by significantly improving throughput in the cell-edge 2267 region (i.e. at the edges of that cell site's radio coverage area). 2268 CoMP techniques depend on deterministic high-reliability 2269 communication between cell sites, however such connections today are 2270 IP-based which in current mobile networks can not meet the QoS 2271 requirements, so CoMP is an emerging technology which can benefit 2272 from DetNet. 2274 Here we consider the JT (Joint Transmit) application for CoMP, which 2275 provides the highest performance gain (compared to other 2276 applications). 2278 7.1.1. CoMP Architecture 2280 +--------------------------+ 2281 | CoMP | 2282 +--+--------------------+--+ 2283 | | 2284 +----------+ +------------+ 2285 | Uplink | | Downlink | 2286 +-----+----+ +--------+---+ 2287 | | 2288 ------------------- ----------------------- 2289 | | | | | | 2290 +---------+ +----+ +-----+ +------------+ +-----+ +-----+ 2291 | Joint | | CS | | DPS | | Joint | | CS/ | | DPS | 2292 |Reception| | | | | |Transmission| | CB | | | 2293 +---------+ +----+ +-----+ +------------+ +-----+ +-----+ 2294 | | 2295 |----------- |------------- 2296 | | | | 2297 +------------+ +---------+ +----------+ +------------+ 2298 | Joint | | Soft | | Coherent | | Non- | 2299 |Equalization| |Combining| | JT | | Coherent JT| 2300 +------------+ +---------+ +----------+ +------------+ 2302 Figure 8: Framework of CoMP Technology 2304 As shown in Figure 8, CoMP reception and transmission is a framework 2305 in which multiple geographically distributed antenna nodes cooperate 2306 to improve the performance of the users served in the common 2307 cooperation area. The design principal of CoMP is to extend the 2308 current single-cell to multi-UE (User Equipment) transmission to a 2309 multi-cell- to-multi-UEs transmission by base station cooperation. 2311 7.1.2. Delay Sensitivity in CoMP 2313 In contrast to the single-cell scenario, CoMP has delay-sensitive 2314 performance parameters, which are "backhaul latency" and "CSI 2315 (Channel State Information) reporting and accuracy". The essential 2316 feature of CoMP is signaling between eNBs, so the backhaul latency is 2317 the dominating limitation of the CoMP performance. Generally, JT can 2318 benefit from coordinated scheduling (either distributed or 2319 centralized) of different cells if the signaling delay between eNBs 2320 is within 4-10ms. This delay requirement is both rigid and absolute 2321 because any uncertainty in delay will degrade the performance 2322 significantly. 2324 7.2. CoMP Today 2326 Due to the strict sensitivity to latency and synchronization, CoMP 2327 between eNB has not been deployed yet. This is because the current 2328 interface path between eNBs cannot meet the delay bound because it is 2329 usually IP-based and passing through multiple network hops (this 2330 interface is called "X2" or "eX2" for "enhanced X2"). Today lack of 2331 absolute delay guarantee on X2/eX2 traffic is the main obstacle to JT 2332 and multi-eNB coordination. 2334 There is still lack of Layer-3 (IP) transport protocol and signaling 2335 that is capable of low latency services; current techniques such as 2336 MPLS and PWE focus on establishing circuits using pre-routed paths 2337 but there is no such signaling for reservation of time-sensitive 2338 stream. 2340 7.3. CoMP Future 2342 7.3.1. Mobile Industry Overall Goals 2344 [METIS] documents the fundamental challenges as well as overall 2345 technical goals of the 5G mobile and wireless system as the starting 2346 point. These future systems should support (at similar cost and 2347 energy consumption levels as today's system): 2349 o 1000 times higher mobile data volume per area 2351 o 10 times to 100 times higher typical user data rate 2353 o 10 times to 100 times higher number of connected devices 2355 o 10 times longer battery life for low power devices 2357 o 5 times reduced End-to-End (E2E) latency 2358 The current LTE networking system has E2E latency less than 20ms 2359 [LTE-Latency] which leads to around 5ms E2E latency for 5G networks. 2360 To fulfill these latency demands at similar cost will be challenging 2361 because as the system also requires 100x bandwidth and 100x connected 2362 devices, simply adding redundant bandwidth provisioning can no longer 2363 be an efficient solution. 2365 In addition to bandwidth provisioning, reserved critical flows should 2366 not be affected by other flows no matter the pressure of the network. 2367 Deterministic networking techniques in both layer-2 and layer-3 using 2368 IETF protocol solutions can be promising to serve these scenarios. 2370 7.3.2. CoMP Infrastructure Goals 2372 Inter-site CoMP is one of the key requirements for 5G and is also a 2373 near-term goal for the current 4.5G network architecture. Assuming 2374 network architecture remains unchanged (i.e. no Fronthaul network and 2375 data flow between eNB is via X2/eX2) we would like to see the 2376 following in the near future: 2378 o Unified protocols and delay-guaranteed forwarding network 2379 equipment that is capable of delivering deterministic latency 2380 services. 2382 o Unified management and protocols which take delay and timing into 2383 account. 2385 o Unified deterministic latency data model and signaling for 2386 resource reservation. 2388 7.4. CoMP Asks 2390 To fully utilize the power of CoMP, it requires: 2392 o Very tight absolute delay bound (100-500us) within 7-10 hops. 2394 o Standardized data plane with highly deterministic networking 2395 capability. 2397 o Standardized control plane to unify backhaul network elements with 2398 time-sensitive stream reservation signaling. 2400 In addition, a standardized deterministic latency data flow model 2401 that includes: 2403 o Network-aware constraints on the networking environment 2404 o Time-aware description of flow characteristics and network 2405 resources, which may not need to be bandwidth based 2407 o Application-aware description of deterministic latency services. 2409 8. Industrial M2M 2411 8.1. Use Case Description 2413 Industrial Automation in general refers to automation of 2414 manufacturing, quality control and material processing. In this 2415 "machine to machine" (M2M) use case we consider machine units in a 2416 plant floor which periodically exchange data with upstream or 2417 downstream machine modules and/or a supervisory controller within a 2418 local area network. 2420 The actors of M2M communication are Programmable Logic Controllers 2421 (PLCs). Communication between PLCs and between PLCs and the 2422 supervisory PLC (S-PLC) is achieved via critical control/data streams 2423 Figure 9. 2425 S (Sensor) 2426 \ +-----+ 2427 PLC__ \.--. .--. ---| MES | 2428 \_( `. _( `./ +-----+ 2429 A------( Local )-------------( L2 ) 2430 ( Net ) ( Net ) +-------+ 2431 /`--(___.-' `--(___.-' ----| S-PLC | 2432 S_/ / PLC .--. / +-------+ 2433 A_/ \_( `. 2434 (Actuator) ( Local ) 2435 ( Net ) 2436 /`--(___.-'\ 2437 / \ A 2438 S A 2440 Figure 9: Current Generic Industrial M2M Network Architecture 2442 This use case focuses on PLC-related communications; communication to 2443 Manufacturing-Execution-Systems (MESs) are not addressed. 2445 This use case covers only critical control/data streams; non-critical 2446 traffic between industrial automation applications (such as 2447 communication of state, configuration, set-up, and database 2448 communication) are adequately served by currently available 2449 prioritizing techniques. Such traffic can use up to 80% of the total 2450 bandwidth required. There is also a subset of non-time-critical 2451 traffic that must be reliable even though it is not time sensitive. 2453 In this use case the primary need for deterministic networking is to 2454 provide end-to-end delivery of M2M messages within specific timing 2455 constraints, for example in closed loop automation control. Today 2456 this level of determinism is provided by proprietary networking 2457 technologies. In addition, standard networking technologies are used 2458 to connect the local network to remote industrial automation sites, 2459 e.g. over an enterprise or metro network which also carries other 2460 types of traffic. Therefore, flows that should be forwarded with 2461 deterministic guarantees need to be sustained regardless of the 2462 amount of other flows in those networks. 2464 8.2. Industrial M2M Communication Today 2466 Today, proprietary networks fulfill the needed timing and 2467 availability for M2M networks. 2469 The network topologies used today by industrial automation are 2470 similar to those used by telecom networks: Daisy Chain, Ring, Hub and 2471 Spoke, and Comb (a subset of Daisy Chain). 2473 PLC-related control/data streams are transmitted periodically and 2474 carry either a pre-configured payload or a payload configured during 2475 runtime. 2477 Some industrial applications require time synchronization at the end 2478 nodes. For such time-coordinated PLCs, accuracy of 1 microsecond is 2479 required. Even in the case of "non-time-coordinated" PLCs time sync 2480 may be needed e.g. for timestamping of sensor data. 2482 Industrial network scenarios require advanced security solutions. 2483 Many of the current industrial production networks are physically 2484 separated. Preventing critical flows from be leaked outside a domain 2485 is handled today by filtering policies that are typically enforced in 2486 firewalls. 2488 8.2.1. Transport Parameters 2490 The Cycle Time defines the frequency of message(s) between industrial 2491 actors. The Cycle Time is application dependent, in the range of 1ms 2492 - 100ms for critical control/data streams. 2494 Because industrial applications assume deterministic transport for 2495 critical Control-Data-Stream parameters (instead of defining latency 2496 and delay variation parameters) it is sufficient to fulfill the upper 2497 bound of latency (maximum latency). The underlying networking 2498 infrastructure must ensure a maximum end-to-end delivery time of 2499 messages in the range of 100 microseconds to 50 milliseconds 2500 depending on the control loop application. 2502 The bandwidth requirements of control/data streams are usually 2503 calculated directly from the bytes-per-cycle parameter of the control 2504 loop. For PLC-to-PLC communication one can expect 2 - 32 streams 2505 with packet size in the range of 100 - 700 bytes. For S-PLC to PLCs 2506 the number of streams is higher - up to 256 streams. Usually no more 2507 than 20% of available bandwidth is used for critical control/data 2508 streams. In today's networks 1Gbps links are commonly used. 2510 Most PLC control loops are rather tolerant of packet loss, however 2511 critical control/data streams accept no more than 1 packet loss per 2512 consecutive communication cycle (i.e. if a packet gets lost in cycle 2513 "n", then the next cycle ("n+1") must be lossless). After two or 2514 more consecutive packet losses the network may be considered to be 2515 "down" by the Application. 2517 As network downtime may impact the whole production system the 2518 required network availability is rather high (99,999%). 2520 Based on the above parameters we expect that some form of redundancy 2521 will be required for M2M communications, however any individual 2522 solution depends on several parameters including cycle time, delivery 2523 time, etc. 2525 8.2.2. Stream Creation and Destruction 2527 In an industrial environment, critical control/data streams are 2528 created rather infrequently, on the order of ~10 times per day / week 2529 / month. Most of these critical control/data streams get created at 2530 machine startup, however flexibility is also needed during runtime, 2531 for example when adding or removing a machine. Going forward as 2532 production systems become more flexible, we expect a significant 2533 increase in the rate at which streams are created, changed and 2534 destroyed. 2536 8.3. Industrial M2M Future 2538 We would like to see a converged IP-standards-based network with 2539 deterministic properties that can satisfy the timing, security and 2540 reliability constraints described above. Today's proprietary 2541 networks could then be interfaced to such a network via gateways or, 2542 in the case of new installations, devices could be connected directly 2543 to the converged network. 2545 8.4. Industrial M2M Asks 2547 o Converged IP-based network 2549 o Deterministic behavior (bounded latency and jitter ) 2551 o High availability (presumably through redundancy) (99.999 %) 2553 o Low message delivery time (100us - 50ms) 2555 o Low packet loss (burstless, 0.1-1 %) 2557 o Precise time synchronization accuracy (1us) 2559 o Security (e.g. prevent critical flows from being leaked between 2560 physically separated networks) 2562 9. Internet-based Applications 2564 9.1. Use Case Description 2566 There are many applications that communicate across the open Internet 2567 that could benefit from guaranteed delivery and bounded latency. The 2568 following are some representative examples. 2570 9.1.1. Media Content Delivery 2572 Media content delivery continues to be an important use of the 2573 Internet, yet users often experience poor quality audio and video due 2574 to the delay and jitter inherent in today's Internet. 2576 9.1.2. Online Gaming 2578 Online gaming is a significant part of the gaming market, however 2579 latency can degrade the end user experience. For example "First 2580 Person Shooter" (FPS) games are highly delay-sensitive. 2582 9.1.3. Virtual Reality 2584 Virtual reality (VR) has many commercial applications including real 2585 estate presentations, remote medical procedures, and so on. Low 2586 latency is critical to interacting with the virtual world because 2587 perceptual delays can cause motion sickness. 2589 9.2. Internet-Based Applications Today 2591 Internet service today is by definition "best effort", with no 2592 guarantees on delivery or bandwidth. 2594 9.3. Internet-Based Applications Future 2596 We imagine an Internet from which we will be able to play a video 2597 without glitches and play games without lag. 2599 For online gaming, the maximum round-trip delay can be 100ms and 2600 stricter for FPS gaming which can be 10-50ms. Transport delay is the 2601 dominate part with a 5-20ms budget. 2603 For VR, 1-10ms maximum delay is needed and total network budget is 2604 1-5ms if doing remote VR. 2606 Flow identification can be used for gaming and VR, i.e. it can 2607 recognize a critical flow and provide appropriate latency bounds. 2609 9.4. Internet-Based Applications Asks 2611 o Unified control and management protocols to handle time-critical 2612 data flow 2614 o Application-aware flow filtering mechanism to recognize the timing 2615 critical flow without doing 5-tuple matching 2617 o Unified control plane to provide low latency service on Layer-3 2618 without changing the data plane 2620 o OAM system and protocols which can help to provide E2E-delay 2621 sensitive service provisioning 2623 10. Use Case Common Elements 2625 Looking at the use cases collectively, the following common desires 2626 for the DetNet-based networks of the future emerge: 2628 o Open standards-based network (replace various proprietary 2629 networks, reduce cost, create multi-vendor market) 2631 o Centrally administered (though such administration may be 2632 distributed for scale and resiliency) 2634 o Integrates L2 (bridged) and L3 (routed) environments (independent 2635 of the Link layer, e.g. can be used with Ethernet, 6TiSCH, etc.) 2637 o Carries both deterministic and best-effort traffic (guaranteed 2638 end-to-end delivery of deterministic flows, deterministic flows 2639 isolated from each other and from best-effort traffic congestion, 2640 unused deterministic BW available to best-effort traffic) 2642 o Ability to add or remove systems from the network with minimal, 2643 bounded service interruption (applications include replacement of 2644 failed devices as well as plug and play) 2646 o Uses standardized data flow information models capable of 2647 expressing deterministic properties (models express device 2648 capabilities, flow properties. Protocols for pushing models from 2649 controller to devices, devices to controller) 2651 o Scalable size (long distances (many km) and short distances 2652 (within a single machine), many hops (radio repeaters, microwave 2653 links, fiber links...) and short hops (single machine)) 2655 o Scalable timing parameters and accuracy (bounded latency, 2656 guaranteed worst case maximum, minimum. Low latency, e.g. control 2657 loops may be less than 1ms, but larger for wide area networks) 2659 o High availability (99.9999 percent up time requested, but may be 2660 up to twelve 9s) 2662 o Reliability, redundancy (lives at stake) 2664 o Security (from failures, attackers, misbehaving devices - 2665 sensitive to both packet content and arrival time) 2667 11. Acknowledgments 2669 11.1. Pro Audio 2671 This section was derived from draft-gunther-detnet-proaudio-req-01. 2673 The editors would like to acknowledge the help of the following 2674 individuals and the companies they represent: 2676 Jeff Koftinoff, Meyer Sound 2678 Jouni Korhonen, Associate Technical Director, Broadcom 2680 Pascal Thubert, CTAO, Cisco 2682 Kieran Tyrrell, Sienda New Media Technologies GmbH 2684 11.2. Utility Telecom 2686 This section was derived from draft-wetterwald-detnet-utilities-reqs- 2687 02. 2689 Faramarz Maghsoodlou, Ph. D. IoT Connected Industries and Energy 2690 Practice Cisco 2692 Pascal Thubert, CTAO Cisco 2694 11.3. Building Automation Systems 2696 This section was derived from draft-bas-usecase-detnet-00. 2698 11.4. Wireless for Industrial 2700 This section was derived from draft-thubert-6tisch-4detnet-01. 2702 This specification derives from the 6TiSCH architecture, which is the 2703 result of multiple interactions, in particular during the 6TiSCH 2704 (bi)Weekly Interim call, relayed through the 6TiSCH mailing list at 2705 the IETF. 2707 The authors wish to thank: Kris Pister, Thomas Watteyne, Xavier 2708 Vilajosana, Qin Wang, Tom Phinney, Robert Assimiti, Michael 2709 Richardson, Zhuo Chen, Malisa Vucinic, Alfredo Grieco, Martin Turon, 2710 Dominique Barthel, Elvis Vogli, Guillaume Gaillard, Herman Storey, 2711 Maria Rita Palattella, Nicola Accettura, Patrick Wetterwald, Pouria 2712 Zand, Raghuram Sudhaakar, and Shitanshu Shah for their participation 2713 and various contributions. 2715 11.5. Cellular Radio 2717 This section was derived from draft-korhonen-detnet-telreq-00. 2719 11.6. Industrial M2M 2721 The authors would like to thank Feng Chen and Marcel Kiessling for 2722 their comments and suggestions. 2724 11.7. Internet Applications and CoMP 2726 This section was derived from draft-zha-detnet-use-case-00. 2728 This document has benefited from reviews, suggestions, comments and 2729 proposed text provided by the following members, listed in 2730 alphabetical order: Jing Huang, Junru Lin, Lehong Niu and Oilver 2731 Huang. 2733 12. Informative References 2735 [ACE] IETF, "Authentication and Authorization for Constrained 2736 Environments", . 2739 [bacnetip] 2740 ASHRAE, "Annex J to ANSI/ASHRAE 135-1995 - BACnet/IP", 2741 January 1999. 2743 [CCAMP] IETF, "Common Control and Measurement Plane", 2744 . 2746 [CoMP] NGMN Alliance, "RAN EVOLUTION PROJECT COMP EVALUATION AND 2747 ENHANCEMENT", NGMN Alliance NGMN_RANEV_D3_CoMP_Evaluation_ 2748 and_Enhancement_v2.0, March 2015, 2749 . 2752 [CONTENT_PROTECTION] 2753 Olsen, D., "1722a Content Protection", 2012, 2754 . 2757 [CPRI] CPRI Cooperation, "Common Public Radio Interface (CPRI); 2758 Interface Specification", CPRI Specification V6.1, July 2759 2014, . 2762 [DCI] Digital Cinema Initiatives, LLC, "DCI Specification, 2763 Version 1.2", 2012, . 2765 [DICE] IETF, "DTLS In Constrained Environments", 2766 . 2768 [EA12] Evans, P. and M. Annunziata, "Industrial Internet: Pushing 2769 the Boundaries of Minds and Machines", November 2012. 2771 [ESPN_DC2] 2772 Daley, D., "ESPN's DC2 Scales AVB Large", 2014, 2773 . 2776 [flnet] Japan Electrical Manufacturers' Association, "JEMA 1479 - 2777 English Edition", September 2012. 2779 [Fronthaul] 2780 Chen, D. and T. Mustala, "Ethernet Fronthaul 2781 Considerations", IEEE 1904.3, February 2015, 2782 . 2785 [HART] www.hartcomm.org, "Highway Addressable remote Transducer, 2786 a group of specifications for industrial process and 2787 control devices administered by the HART Foundation". 2789 [I-D.finn-detnet-architecture] 2790 Finn, N., Thubert, P., and M. Teener, "Deterministic 2791 Networking Architecture", draft-finn-detnet- 2792 architecture-03 (work in progress), March 2016. 2794 [I-D.finn-detnet-problem-statement] 2795 Finn, N. and P. Thubert, "Deterministic Networking Problem 2796 Statement", draft-finn-detnet-problem-statement-04 (work 2797 in progress), October 2015. 2799 [I-D.ietf-6tisch-6top-interface] 2800 Wang, Q. and X. Vilajosana, "6TiSCH Operation Sublayer 2801 (6top) Interface", draft-ietf-6tisch-6top-interface-04 2802 (work in progress), July 2015. 2804 [I-D.ietf-6tisch-architecture] 2805 Thubert, P., "An Architecture for IPv6 over the TSCH mode 2806 of IEEE 802.15.4", draft-ietf-6tisch-architecture-09 (work 2807 in progress), November 2015. 2809 [I-D.ietf-6tisch-coap] 2810 Sudhaakar, R. and P. Zand, "6TiSCH Resource Management and 2811 Interaction using CoAP", draft-ietf-6tisch-coap-03 (work 2812 in progress), March 2015. 2814 [I-D.ietf-6tisch-terminology] 2815 Palattella, M., Thubert, P., Watteyne, T., and Q. Wang, 2816 "Terminology in IPv6 over the TSCH mode of IEEE 2817 802.15.4e", draft-ietf-6tisch-terminology-06 (work in 2818 progress), November 2015. 2820 [I-D.ietf-ipv6-multilink-subnets] 2821 Thaler, D. and C. Huitema, "Multi-link Subnet Support in 2822 IPv6", draft-ietf-ipv6-multilink-subnets-00 (work in 2823 progress), July 2002. 2825 [I-D.ietf-roll-rpl-industrial-applicability] 2826 Phinney, T., Thubert, P., and R. Assimiti, "RPL 2827 applicability in industrial networks", draft-ietf-roll- 2828 rpl-industrial-applicability-02 (work in progress), 2829 October 2013. 2831 [I-D.ietf-tictoc-1588overmpls] 2832 Davari, S., Oren, A., Bhatia, M., Roberts, P., and L. 2833 Montini, "Transporting Timing messages over MPLS 2834 Networks", draft-ietf-tictoc-1588overmpls-07 (work in 2835 progress), October 2015. 2837 [I-D.kh-spring-ip-ran-use-case] 2838 Khasnabish, B., hu, f., and L. Contreras, "Segment Routing 2839 in IP RAN use case", draft-kh-spring-ip-ran-use-case-02 2840 (work in progress), November 2014. 2842 [I-D.mirsky-mpls-residence-time] 2843 Mirsky, G., Ruffini, S., Gray, E., Drake, J., Bryant, S., 2844 and S. Vainshtein, "Residence Time Measurement in MPLS 2845 network", draft-mirsky-mpls-residence-time-07 (work in 2846 progress), July 2015. 2848 [I-D.svshah-tsvwg-deterministic-forwarding] 2849 Shah, S. and P. Thubert, "Deterministic Forwarding PHB", 2850 draft-svshah-tsvwg-deterministic-forwarding-04 (work in 2851 progress), August 2015. 2853 [I-D.thubert-6lowpan-backbone-router] 2854 Thubert, P., "6LoWPAN Backbone Router", draft-thubert- 2855 6lowpan-backbone-router-03 (work in progress), February 2856 2013. 2858 [I-D.wang-6tisch-6top-sublayer] 2859 Wang, Q. and X. Vilajosana, "6TiSCH Operation Sublayer 2860 (6top)", draft-wang-6tisch-6top-sublayer-04 (work in 2861 progress), November 2015. 2863 [IEC61850-90-12] 2864 TC57 WG10, IEC., "IEC 61850-90-12 TR: Communication 2865 networks and systems for power utility automation - Part 2866 90-12: Wide area network engineering guidelines", 2015. 2868 [IEC62439-3:2012] 2869 TC65, IEC., "IEC 62439-3: Industrial communication 2870 networks - High availability automation networks - Part 3: 2871 Parallel Redundancy Protocol (PRP) and High-availability 2872 Seamless Redundancy (HSR)", 2012. 2874 [IEEE1588] 2875 IEEE, "IEEE Standard for a Precision Clock Synchronization 2876 Protocol for Networked Measurement and Control Systems", 2877 IEEE Std 1588-2008, 2008, 2878 . 2881 [IEEE1722] 2882 IEEE, "1722-2011 - IEEE Standard for Layer 2 Transport 2883 Protocol for Time Sensitive Applications in a Bridged 2884 Local Area Network", IEEE Std 1722-2011, 2011, 2885 . 2888 [IEEE19043] 2889 IEEE Standards Association, "IEEE 1904.3 TF", IEEE 1904.3, 2890 2015, . 2892 [IEEE802.1TSNTG] 2893 IEEE Standards Association, "IEEE 802.1 Time-Sensitive 2894 Networks Task Group", March 2013, 2895 . 2897 [IEEE802154] 2898 IEEE standard for Information Technology, "IEEE std. 2899 802.15.4, Part. 15.4: Wireless Medium Access Control (MAC) 2900 and Physical Layer (PHY) Specifications for Low-Rate 2901 Wireless Personal Area Networks". 2903 [IEEE802154e] 2904 IEEE standard for Information Technology, "IEEE standard 2905 for Information Technology, IEEE std. 802.15.4, Part. 2906 15.4: Wireless Medium Access Control (MAC) and Physical 2907 Layer (PHY) Specifications for Low-Rate Wireless Personal 2908 Area Networks, June 2011 as amended by IEEE std. 2909 802.15.4e, Part. 15.4: Low-Rate Wireless Personal Area 2910 Networks (LR-WPANs) Amendment 1: MAC sublayer", April 2911 2012. 2913 [IEEE8021AS] 2914 IEEE, "Timing and Synchronizations (IEEE 802.1AS-2011)", 2915 IEEE 802.1AS-2001, 2011, 2916 . 2919 [IEEE8021CM] 2920 Farkas, J., "Time-Sensitive Networking for Fronthaul", 2921 Unapproved PAR, PAR for a New IEEE Standard; 2922 IEEE P802.1CM, April 2015, 2923 . 2926 [IEEE8021TSN] 2927 IEEE 802.1, "The charter of the TG is to provide the 2928 specifications that will allow time-synchronized low 2929 latency streaming services through 802 networks.", 2016, 2930 . 2932 [IETFDetNet] 2933 IETF, "Charter for IETF DetNet Working Group", 2015, 2934 . 2936 [ISA100] ISA/ANSI, "ISA100, Wireless Systems for Automation", 2937 . 2939 [ISA100.11a] 2940 ISA/ANSI, "Wireless Systems for Industrial Automation: 2941 Process Control and Related Applications - ISA100.11a-2011 2942 - IEC 62734", 2011, . 2945 [ISO7240-16] 2946 ISO, "ISO 7240-16:2007 Fire detection and alarm systems -- 2947 Part 16: Sound system control and indicating equipment", 2948 2007, . 2951 [knx] KNX Association, "ISO/IEC 14543-3 - KNX", November 2006. 2953 [lontalk] ECHELON, "LonTalk(R) Protocol Specification Version 3.0", 2954 1994. 2956 [LTE-Latency] 2957 Johnston, S., "LTE Latency: How does it compare to other 2958 technologies", March 2014, 2959 . 2962 [MEF] MEF, "Mobile Backhaul Phase 2 Amendment 1 -- Small Cells", 2963 MEF 22.1.1, July 2014, 2964 . 2967 [METIS] METIS, "Scenarios, requirements and KPIs for 5G mobile and 2968 wireless system", ICT-317669-METIS/D1.1 ICT- 2969 317669-METIS/D1.1, April 2013, . 2972 [modbus] Modbus Organization, "MODBUS APPLICATION PROTOCOL 2973 SPECIFICATION V1.1b", December 2006. 2975 [net5G] Ericsson, "5G Radio Access, Challenges for 2020 and 2976 Beyond", Ericsson white paper wp-5g, June 2013, 2977 . 2979 [NGMN] NGMN Alliance, "5G White Paper", NGMN 5G White Paper v1.0, 2980 February 2015, . 2983 [PCE] IETF, "Path Computation Element", 2984 . 2986 [profibus] 2987 IEC, "IEC 61158 Type 3 - Profibus DP", January 2001. 2989 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 2990 Requirement Levels", BCP 14, RFC 2119, 2991 DOI 10.17487/RFC2119, March 1997, 2992 . 2994 [RFC2460] Deering, S. and R. Hinden, "Internet Protocol, Version 6 2995 (IPv6) Specification", RFC 2460, DOI 10.17487/RFC2460, 2996 December 1998, . 2998 [RFC2474] Nichols, K., Blake, S., Baker, F., and D. Black, 2999 "Definition of the Differentiated Services Field (DS 3000 Field) in the IPv4 and IPv6 Headers", RFC 2474, 3001 DOI 10.17487/RFC2474, December 1998, 3002 . 3004 [RFC3031] Rosen, E., Viswanathan, A., and R. Callon, "Multiprotocol 3005 Label Switching Architecture", RFC 3031, 3006 DOI 10.17487/RFC3031, January 2001, 3007 . 3009 [RFC3209] Awduche, D., Berger, L., Gan, D., Li, T., Srinivasan, V., 3010 and G. Swallow, "RSVP-TE: Extensions to RSVP for LSP 3011 Tunnels", RFC 3209, DOI 10.17487/RFC3209, December 2001, 3012 . 3014 [RFC3393] Demichelis, C. and P. Chimento, "IP Packet Delay Variation 3015 Metric for IP Performance Metrics (IPPM)", RFC 3393, 3016 DOI 10.17487/RFC3393, November 2002, 3017 . 3019 [RFC3444] Pras, A. and J. Schoenwaelder, "On the Difference between 3020 Information Models and Data Models", RFC 3444, 3021 DOI 10.17487/RFC3444, January 2003, 3022 . 3024 [RFC3972] Aura, T., "Cryptographically Generated Addresses (CGA)", 3025 RFC 3972, DOI 10.17487/RFC3972, March 2005, 3026 . 3028 [RFC3985] Bryant, S., Ed. and P. Pate, Ed., "Pseudo Wire Emulation 3029 Edge-to-Edge (PWE3) Architecture", RFC 3985, 3030 DOI 10.17487/RFC3985, March 2005, 3031 . 3033 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 3034 Architecture", RFC 4291, DOI 10.17487/RFC4291, February 3035 2006, . 3037 [RFC4553] Vainshtein, A., Ed. and YJ. Stein, Ed., "Structure- 3038 Agnostic Time Division Multiplexing (TDM) over Packet 3039 (SAToP)", RFC 4553, DOI 10.17487/RFC4553, June 2006, 3040 . 3042 [RFC4903] Thaler, D., "Multi-Link Subnet Issues", RFC 4903, 3043 DOI 10.17487/RFC4903, June 2007, 3044 . 3046 [RFC4919] Kushalnagar, N., Montenegro, G., and C. Schumacher, "IPv6 3047 over Low-Power Wireless Personal Area Networks (6LoWPANs): 3048 Overview, Assumptions, Problem Statement, and Goals", 3049 RFC 4919, DOI 10.17487/RFC4919, August 2007, 3050 . 3052 [RFC5086] Vainshtein, A., Ed., Sasson, I., Metz, E., Frost, T., and 3053 P. Pate, "Structure-Aware Time Division Multiplexed (TDM) 3054 Circuit Emulation Service over Packet Switched Network 3055 (CESoPSN)", RFC 5086, DOI 10.17487/RFC5086, December 2007, 3056 . 3058 [RFC5087] Stein, Y(J)., Shashoua, R., Insler, R., and M. Anavi, 3059 "Time Division Multiplexing over IP (TDMoIP)", RFC 5087, 3060 DOI 10.17487/RFC5087, December 2007, 3061 . 3063 [RFC6282] Hui, J., Ed. and P. Thubert, "Compression Format for IPv6 3064 Datagrams over IEEE 802.15.4-Based Networks", RFC 6282, 3065 DOI 10.17487/RFC6282, September 2011, 3066 . 3068 [RFC6550] Winter, T., Ed., Thubert, P., Ed., Brandt, A., Hui, J., 3069 Kelsey, R., Levis, P., Pister, K., Struik, R., Vasseur, 3070 JP., and R. Alexander, "RPL: IPv6 Routing Protocol for 3071 Low-Power and Lossy Networks", RFC 6550, 3072 DOI 10.17487/RFC6550, March 2012, 3073 . 3075 [RFC6551] Vasseur, JP., Ed., Kim, M., Ed., Pister, K., Dejean, N., 3076 and D. Barthel, "Routing Metrics Used for Path Calculation 3077 in Low-Power and Lossy Networks", RFC 6551, 3078 DOI 10.17487/RFC6551, March 2012, 3079 . 3081 [RFC6775] Shelby, Z., Ed., Chakrabarti, S., Nordmark, E., and C. 3082 Bormann, "Neighbor Discovery Optimization for IPv6 over 3083 Low-Power Wireless Personal Area Networks (6LoWPANs)", 3084 RFC 6775, DOI 10.17487/RFC6775, November 2012, 3085 . 3087 [RFC7554] Watteyne, T., Ed., Palattella, M., and L. Grieco, "Using 3088 IEEE 802.15.4e Time-Slotted Channel Hopping (TSCH) in the 3089 Internet of Things (IoT): Problem Statement", RFC 7554, 3090 DOI 10.17487/RFC7554, May 2015, 3091 . 3093 [SRP_LATENCY] 3094 Gunther, C., "Specifying SRP Latency", 2014, 3095 . 3098 [STUDIO_IP] 3099 Mace, G., "IP Networked Studio Infrastructure for 3100 Synchronized & Real-Time Multimedia Transmissions", 2007, 3101 . 3104 [SyncE] ITU-T, "G.8261 : Timing and synchronization aspects in 3105 packet networks", Recommendation G.8261, August 2013, 3106 . 3108 [TEAS] IETF, "Traffic Engineering Architecture and Signaling", 3109 . 3111 [TS23401] 3GPP, "General Packet Radio Service (GPRS) enhancements 3112 for Evolved Universal Terrestrial Radio Access Network 3113 (E-UTRAN) access", 3GPP TS 23.401 10.10.0, March 2013. 3115 [TS25104] 3GPP, "Base Station (BS) radio transmission and reception 3116 (FDD)", 3GPP TS 25.104 3.14.0, March 2007. 3118 [TS36104] 3GPP, "Evolved Universal Terrestrial Radio Access 3119 (E-UTRA); Base Station (BS) radio transmission and 3120 reception", 3GPP TS 36.104 10.11.0, July 2013. 3122 [TS36133] 3GPP, "Evolved Universal Terrestrial Radio Access 3123 (E-UTRA); Requirements for support of radio resource 3124 management", 3GPP TS 36.133 12.7.0, April 2015. 3126 [TS36211] 3GPP, "Evolved Universal Terrestrial Radio Access 3127 (E-UTRA); Physical channels and modulation", 3GPP 3128 TS 36.211 10.7.0, March 2013. 3130 [TS36300] 3GPP, "Evolved Universal Terrestrial Radio Access (E-UTRA) 3131 and Evolved Universal Terrestrial Radio Access Network 3132 (E-UTRAN); Overall description; Stage 2", 3GPP TS 36.300 3133 10.11.0, September 2013. 3135 [TSNTG] IEEE Standards Association, "IEEE 802.1 Time-Sensitive 3136 Networks Task Group", 2013, 3137 . 3139 [UHD-video] 3140 Holub, P., "Ultra-High Definition Videos and Their 3141 Applications over the Network", The 7th International 3142 Symposium on VICTORIES Project PetrHolub_presentation, 3143 October 2014, . 3146 [WirelessHART] 3147 www.hartcomm.org, "Industrial Communication Networks - 3148 Wireless Communication Network and Communication Profiles 3149 - WirelessHART - IEC 62591", 2010. 3151 Authors' Addresses 3152 Ethan Grossman (editor) 3153 Dolby Laboratories, Inc. 3154 1275 Market Street 3155 San Francisco, CA 94103 3156 USA 3158 Phone: +1 415 645 4726 3159 Email: ethan.grossman@dolby.com 3160 URI: http://www.dolby.com 3162 Craig Gunther 3163 Harman International 3164 10653 South River Front Parkway 3165 South Jordan, UT 84095 3166 USA 3168 Phone: +1 801 568-7675 3169 Email: craig.gunther@harman.com 3170 URI: http://www.harman.com 3172 Pascal Thubert 3173 Cisco Systems, Inc 3174 Building D 3175 45 Allee des Ormes - BP1200 3176 MOUGINS - Sophia Antipolis 06254 3177 FRANCE 3179 Phone: +33 497 23 26 34 3180 Email: pthubert@cisco.com 3182 Patrick Wetterwald 3183 Cisco Systems 3184 45 Allees des Ormes 3185 Mougins 06250 3186 FRANCE 3188 Phone: +33 4 97 23 26 36 3189 Email: pwetterw@cisco.com 3190 Jean Raymond 3191 Hydro-Quebec 3192 1500 University 3193 Montreal H3A3S7 3194 Canada 3196 Phone: +1 514 840 3000 3197 Email: raymond.jean@hydro.qc.ca 3199 Jouni Korhonen 3200 Broadcom Corporation 3201 3151 Zanker Road 3202 San Jose, CA 95134 3203 USA 3205 Email: jouni.nospam@gmail.com 3207 Yu Kaneko 3208 Toshiba 3209 1 Komukai-Toshiba-cho, Saiwai-ku, Kasasaki-shi 3210 Kanagawa, Japan 3212 Email: yu1.kaneko@toshiba.co.jp 3214 Subir Das 3215 Applied Communication Sciences 3216 150 Mount Airy Road, Basking Ridge 3217 New Jersey, 07920, USA 3219 Email: sdas@appcomsci.com 3221 Yiyong Zha 3222 Huawei Technologies 3224 Email: zhayiyong@huawei.com 3226 Balazs Varga 3227 Ericsson 3228 Konyves Kalman krt. 11/B 3229 Budapest 1097 3230 Hungary 3232 Email: balazs.a.varga@ericsson.com 3233 Janos Farkas 3234 Ericsson 3235 Konyves Kalman krt. 11/B 3236 Budapest 1097 3237 Hungary 3239 Email: janos.farkas@ericsson.com 3241 Franz-Josef Goetz 3242 Siemens 3243 Gleiwitzerstr. 555 3244 Nurnberg 90475 3245 Germany 3247 Email: franz-josef.goetz@siemens.com 3249 Juergen Schmitt 3250 Siemens 3251 Gleiwitzerstr. 555 3252 Nurnberg 90475 3253 Germany 3255 Email: juergen.jues.schmitt@siemens.com