idnits 2.17.1 draft-ietf-detnet-security-05.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 abstract seems to contain references ([ARINC664P7]), which it shouldn't. Please replace those with straight textual mentions of the documents in question. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (August 29, 2019) is 1702 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Missing Reference: 'RFC2475' is mentioned on line 1594, but not defined == Outdated reference: A later version (-06) exists of draft-ietf-detnet-data-plane-framework-01 == Outdated reference: A later version (-07) exists of draft-ietf-detnet-ip-01 == Outdated reference: A later version (-07) exists of draft-ietf-detnet-ip-over-tsn-00 == Outdated reference: A later version (-13) exists of draft-ietf-detnet-mpls-01 Summary: 1 error (**), 0 flaws (~~), 6 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Internet Engineering Task Force T. Mizrahi 3 Internet-Draft HUAWEI 4 Intended status: Informational E. Grossman, Ed. 5 Expires: March 1, 2020 DOLBY 6 A. Hacker 7 MISTIQ 8 S. Das 9 Applied Communication Sciences 10 J. Dowdell 11 Airbus Defence and Space 12 H. Austad 13 Cisco Systems 14 K. Stanton 15 INTEL 16 N. Finn 17 HUAWEI 18 August 29, 2019 20 Deterministic Networking (DetNet) Security Considerations 21 draft-ietf-detnet-security-05 23 Abstract 25 A deterministic network is one that can carry data flows for real- 26 time applications with extremely low data loss rates and bounded 27 latency. Deterministic networks have been successfully deployed in 28 real-time operational technology (OT) applications for some years 29 (for example [ARINC664P7]). However, such networks are typically 30 isolated from external access, and thus the security threat from 31 external attackers is low. IETF Deterministic Networking (DetNet) 32 specifies a set of technologies that enable creation of deterministic 33 networks on IP-based networks of potentially wide area (on the scale 34 of a corporate network) potentially bringing the OT network into 35 contact with Information Technology (IT) traffic and security threats 36 that lie outside of a tightly controlled and bounded area (such as 37 the internals of an aircraft). These DetNet technologies have not 38 previously been deployed together on a wide area IP-based network, 39 and thus can present security considerations that may be new to IP- 40 based wide area network designers. This draft, intended for use by 41 DetNet network designers, provides insight into these security 42 considerations. In addition, this draft collects all security- 43 related statements from the various DetNet drafts (Architecture, Use 44 Cases, etc) into a single location Section 8. 46 Status of This Memo 48 This Internet-Draft is submitted in full conformance with the 49 provisions of BCP 78 and BCP 79. 51 Internet-Drafts are working documents of the Internet Engineering 52 Task Force (IETF). Note that other groups may also distribute 53 working documents as Internet-Drafts. The list of current Internet- 54 Drafts is at https://datatracker.ietf.org/drafts/current/. 56 Internet-Drafts are draft documents valid for a maximum of six months 57 and may be updated, replaced, or obsoleted by other documents at any 58 time. It is inappropriate to use Internet-Drafts as reference 59 material or to cite them other than as "work in progress." 61 This Internet-Draft will expire on March 1, 2020. 63 Copyright Notice 65 Copyright (c) 2019 IETF Trust and the persons identified as the 66 document authors. All rights reserved. 68 This document is subject to BCP 78 and the IETF Trust's Legal 69 Provisions Relating to IETF Documents 70 (https://trustee.ietf.org/license-info) in effect on the date of 71 publication of this document. Please review these documents 72 carefully, as they describe your rights and restrictions with respect 73 to this document. Code Components extracted from this document must 74 include Simplified BSD License text as described in Section 4.e of 75 the Trust Legal Provisions and are provided without warranty as 76 described in the Simplified BSD License. 78 Table of Contents 80 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 5 81 2. Abbreviations . . . . . . . . . . . . . . . . . . . . . . . . 6 82 3. Security Threats . . . . . . . . . . . . . . . . . . . . . . 6 83 3.1. Threat Model . . . . . . . . . . . . . . . . . . . . . . 6 84 3.2. Threat Analysis . . . . . . . . . . . . . . . . . . . . . 7 85 3.2.1. Delay . . . . . . . . . . . . . . . . . . . . . . . . 7 86 3.2.1.1. Delay Attack . . . . . . . . . . . . . . . . . . 7 87 3.2.2. DetNet Flow Modification or Spoofing . . . . . . . . 7 88 3.2.3. Resource Segmentation or Slicing . . . . . . . . . . 8 89 3.2.3.1. Inter-segment Attack . . . . . . . . . . . . . . 8 90 3.2.4. Packet Replication and Elimination . . . . . . . . . 8 91 3.2.4.1. Replication: Increased Attack Surface . . . . . . 8 92 3.2.4.2. Replication-related Header Manipulation . . . . . 8 93 3.2.5. Path Choice . . . . . . . . . . . . . . . . . . . . . 9 94 3.2.5.1. Path Manipulation . . . . . . . . . . . . . . . . 9 95 3.2.5.2. Path Choice: Increased Attack Surface . . . . . . 9 96 3.2.6. Control Plane . . . . . . . . . . . . . . . . . . . . 9 97 3.2.6.1. Control or Signaling Packet Modification . . . . 9 98 3.2.6.2. Control or Signaling Packet Injection . . . . . . 9 99 3.2.7. Scheduling or Shaping . . . . . . . . . . . . . . . . 9 100 3.2.7.1. Reconnaissance . . . . . . . . . . . . . . . . . 9 101 3.2.8. Time Synchronization Mechanisms . . . . . . . . . . . 10 102 3.3. Threat Summary . . . . . . . . . . . . . . . . . . . . . 10 103 4. Security Threat Impacts . . . . . . . . . . . . . . . . . . . 11 104 4.1. Delay-Attacks . . . . . . . . . . . . . . . . . . . . . . 13 105 4.1.1. Data Plane Delay Attacks . . . . . . . . . . . . . . 13 106 4.1.2. Control Plane Delay Attacks . . . . . . . . . . . . . 14 107 4.2. Flow Modification and Spoofing . . . . . . . . . . . . . 14 108 4.2.1. Flow Modification . . . . . . . . . . . . . . . . . . 14 109 4.2.2. Spoofing . . . . . . . . . . . . . . . . . . . . . . 14 110 4.2.2.1. Dataplane Spoofing . . . . . . . . . . . . . . . 14 111 4.2.2.2. Control Plane Spoofing . . . . . . . . . . . . . 15 112 4.3. Segmentation attacks (injection) . . . . . . . . . . . . 15 113 4.3.1. Data Plane Segmentation . . . . . . . . . . . . . . . 15 114 4.3.2. Control Plane segmentation . . . . . . . . . . . . . 15 115 4.4. Replication and Elimination . . . . . . . . . . . . . . . 16 116 4.4.1. Increased Attack Surface . . . . . . . . . . . . . . 16 117 4.4.2. Header Manipulation at Elimination Bridges . . . . . 16 118 4.5. Control or Signaling Packet Modification . . . . . . . . 16 119 4.6. Control or Signaling Packet Injection . . . . . . . . . . 16 120 4.7. Reconnaissance . . . . . . . . . . . . . . . . . . . . . 16 121 4.8. Attacks on Time Sync Mechanisms . . . . . . . . . . . . . 16 122 4.9. Attacks on Path Choice . . . . . . . . . . . . . . . . . 16 123 5. Security Threat Mitigation . . . . . . . . . . . . . . . . . 17 124 5.1. Path Redundancy . . . . . . . . . . . . . . . . . . . . . 17 125 5.2. Integrity Protection . . . . . . . . . . . . . . . . . . 17 126 5.3. DetNet Node Authentication . . . . . . . . . . . . . . . 17 127 5.4. Encryption . . . . . . . . . . . . . . . . . . . . . . . 18 128 5.4.1. Encryption Considerations for DetNet . . . . . . . . 18 129 5.5. Control and Signaling Message Protection . . . . . . . . 19 130 5.6. Dynamic Performance Analytics . . . . . . . . . . . . . . 19 131 5.7. Mitigation Summary . . . . . . . . . . . . . . . . . . . 20 132 6. Association of Attacks to Use Cases . . . . . . . . . . . . . 21 133 6.1. Use Cases by Common Themes . . . . . . . . . . . . . . . 22 134 6.1.1. Network Layer - AVB/TSN Ethernet . . . . . . . . . . 22 135 6.1.2. Central Administration . . . . . . . . . . . . . . . 22 136 6.1.3. Hot Swap . . . . . . . . . . . . . . . . . . . . . . 22 137 6.1.4. Data Flow Information Models . . . . . . . . . . . . 23 138 6.1.5. L2 and L3 Integration . . . . . . . . . . . . . . . . 23 139 6.1.6. End-to-End Delivery . . . . . . . . . . . . . . . . . 24 140 6.1.7. Proprietary Deterministic Ethernet Networks . . . . . 24 141 6.1.8. Replacement for Proprietary Fieldbuses . . . . . . . 24 142 6.1.9. Deterministic vs Best-Effort Traffic . . . . . . . . 25 143 6.1.10. Deterministic Flows . . . . . . . . . . . . . . . . . 25 144 6.1.11. Unused Reserved Bandwidth . . . . . . . . . . . . . . 26 145 6.1.12. Interoperability . . . . . . . . . . . . . . . . . . 26 146 6.1.13. Cost Reductions . . . . . . . . . . . . . . . . . . . 26 147 6.1.14. Insufficiently Secure Devices . . . . . . . . . . . . 27 148 6.1.15. DetNet Network Size . . . . . . . . . . . . . . . . . 27 149 6.1.16. Multiple Hops . . . . . . . . . . . . . . . . . . . . 27 150 6.1.17. Level of Service . . . . . . . . . . . . . . . . . . 28 151 6.1.18. Bounded Latency . . . . . . . . . . . . . . . . . . . 28 152 6.1.19. Low Latency . . . . . . . . . . . . . . . . . . . . . 28 153 6.1.20. Bounded Jitter (Latency Variation) . . . . . . . . . 29 154 6.1.21. Symmetrical Path Delays . . . . . . . . . . . . . . . 29 155 6.1.22. Reliability and Availability . . . . . . . . . . . . 29 156 6.1.23. Redundant Paths . . . . . . . . . . . . . . . . . . . 30 157 6.1.24. Security Measures . . . . . . . . . . . . . . . . . . 30 158 6.2. Attack Types by Use Case Common Theme . . . . . . . . . . 30 159 6.3. Security Considerations for OAM Traffic . . . . . . . . . 33 160 7. DetNet Technology-Specific Threats . . . . . . . . . . . . . 33 161 7.1. IP . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 162 7.2. MPLS . . . . . . . . . . . . . . . . . . . . . . . . . . 34 163 7.3. TSN . . . . . . . . . . . . . . . . . . . . . . . . . . . 34 164 8. Appendix A: DetNet Draft Security-Related Statements . . . . 34 165 8.1. Architecture (draft 8) . . . . . . . . . . . . . . . . . 34 166 8.1.1. Fault Mitigation (sec 4.5) . . . . . . . . . . . . . 34 167 8.1.2. Security Considerations (sec 7) . . . . . . . . . . . 35 168 8.2. Data Plane Alternatives (draft 4) . . . . . . . . . . . . 36 169 8.2.1. Security Considerations (sec 7) . . . . . . . . . . . 36 170 8.3. Problem Statement (draft 5) . . . . . . . . . . . . . . . 36 171 8.3.1. Security Considerations (sec 5) . . . . . . . . . . . 36 172 8.4. Use Cases (draft 11) . . . . . . . . . . . . . . . . . . 37 173 8.4.1. (Utility Networks) Security Current Practices and 174 Limitations (sec 3.2.1) . . . . . . . . . . . . . . . 37 175 8.4.2. (Utility Networks) Security Trends in Utility 176 Networks (sec 3.3.3) . . . . . . . . . . . . . . . . 38 177 8.4.3. (BAS) Security Considerations (sec 4.2.4) . . . . . . 40 178 8.4.4. (6TiSCH) Security Considerations (sec 5.3.3) . . . . 40 179 8.4.5. (Cellular radio) Security Considerations (sec 6.1.5) 41 180 8.4.6. (Industrial M2M) Communication Today (sec 7.2) . . . 41 181 9. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 41 182 10. Security Considerations . . . . . . . . . . . . . . . . . . . 41 183 11. Informative References . . . . . . . . . . . . . . . . . . . 41 184 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 43 186 1. Introduction 188 Security is of particularly high importance in DetNet networks 189 because many of the use cases which are enabled by DetNet [RFC8578] 190 include control of physical devices (power grid components, 191 industrial controls, building controls) which can have high 192 operational costs for failure, and present potentially attractive 193 targets for cyber-attackers. 195 This situation is even more acute given that one of the goals of 196 DetNet is to provide a "converged network", i.e. one that includes 197 both IT traffic and OT traffic, thus exposing potentially sensitive 198 OT devices to attack in ways that were not previously common (usually 199 because they were under a separate control system or otherwise 200 isolated from the IT network). Security considerations for OT 201 networks is not a new area, and there are many OT networks today that 202 are connected to wide area networks or the Internet; this draft 203 focuses on the issues that are specific to the DetNet technologies 204 and use cases. 206 The DetNet technologies include ways to: 208 o Reserve data plane resources for DetNet flows in some or all of 209 the intermediate nodes (e.g. bridges or routers) along the path of 210 the flow 212 o Provide explicit routes for DetNet flows that do not rapidly 213 change with the network topology 215 o Distribute data from DetNet flow packets over time and/or space to 216 ensure delivery of each packet's data' in spite of the loss of a 217 path 219 This draft includes sections on threat modeling and analysis, threat 220 impact and mitigation, and the association of attacks with use cases 221 based on the Use Case Common Themes section of the DetNet Use Cases 222 draft [RFC8578]. 224 This draft also provides context for the DetNet security 225 considerations by collecting into one place Section 8 the various 226 remarks about security from the various DetNet drafts (Use Cases, 227 Architecture, etc). This text is duplicated here primarily because 228 the DetNet working group has elected not to produce a Requirements 229 draft and thus collectively these statements are as close as we have 230 to "DetNet Security Requirements". 232 2. Abbreviations 234 IT Information technology (the application of computers to 235 store, study, retrieve, transmit, and manipulate data or information, 236 often in the context of a business or other enterprise - Wikipedia). 238 OT Operational Technology (the hardware and software 239 dedicated to detecting or causing changes in physical processes 240 through direct monitoring and/or control of physical devices such as 241 valves, pumps, etc. - Wikipedia) 243 MITM Man in the Middle 245 SN Sequence Number 247 STRIDE Addresses risk and severity associated with threat 248 categories: Spoofing identity, Tampering with data, Repudiation, 249 Information disclosure, Denial of service, Elevation of privilege. 251 DREAD Compares and prioritizes risk represented by these threat 252 categories: Damage potential, Reproducibility, Exploitability, how 253 many Affected users, Discoverability. 255 PTP Precision Time Protocol [IEEE1588] 257 3. Security Threats 259 This section presents a threat model, and analyzes the possible 260 threats in a DetNet-enabled network. The threats considered in this 261 section are independent of any specific technologies used to 262 implement the DetNet; Section 7) considers attacks that are 263 associated with the DetNet technologies encompassed by 264 [I-D.ietf-detnet-data-plane-framework]. 266 We distinguish control plane threats from data plane threats. The 267 attack surface may be the same, but the types of attacks as well as 268 the motivation behind them, are different. For example, a delay 269 attack is more relevant to data plane than to control plane. There 270 is also a difference in terms of security solutions: the way you 271 secure the data plane is often different than the way you secure the 272 control plane. 274 3.1. Threat Model 276 The threat model used in this memo is based on the threat model of 277 Section 3.1 of [RFC7384]. This model classifies attackers based on 278 two criteria: 280 o Internal vs. external: internal attackers either have access to a 281 trusted segment of the network or possess the encryption or 282 authentication keys. External attackers, on the other hand, do 283 not have the keys and have access only to the encrypted or 284 authenticated traffic. 286 o Man in the Middle (MITM) vs. packet injector: MITM attackers are 287 located in a position that allows interception and modification of 288 in-flight protocol packets, whereas a traffic injector can only 289 attack by generating protocol packets. 291 Care has also been taken to adhere to Section 5 of [RFC3552], both 292 with respect to what attacks are considered out-of-scope for this 293 document, but also what is considered to be the most common threats 294 (explored furhter in Section 3.2. Most of the direct threats to 295 DetNet are Active attacks, but it is highly suggested that DetNet 296 application developers take appropriate measures to protect the 297 content of the streams from passive attacks. 299 DetNet-Service, one of the service scenarios described in 300 [I-D.varga-detnet-service-model], is the case where a service 301 connects DetNet networking islands, i.e. two or more otherwise 302 independent DetNet network domains are connected via a link that is 303 not intrinsically part of either network. This implies that there 304 could be DetNet traffic flowing over a non-DetNet link, which may 305 provide an attacker with an advantageous opportunity to tamper with 306 DetNet traffic. The security properties of non-DetNet links are 307 outside of the scope of DetNet Security, but it should be noted that 308 use of non-DetNet services to interconnect DetNet networks merits 309 security analysis to ensure the integrity of the DetNet networks 310 involved. 312 3.2. Threat Analysis 314 3.2.1. Delay 316 3.2.1.1. Delay Attack 318 An attacker can maliciously delay DetNet data flow traffic. By 319 delaying the traffic, the attacker can compromise the service of 320 applications that are sensitive to high delays or to high delay 321 variation. 323 3.2.2. DetNet Flow Modification or Spoofing 325 An attacker can modify some header fields of en route packets in a 326 way that causes the DetNet flow identification mechanisms to 327 misclassify the flow. Alternatively, the attacker can inject traffic 328 that is tailored to appear as if it belongs to a legitimate DetNet 329 flow. The potential consequence is that the DetNet flow resource 330 allocation cannot guarantee the performance that is expected when the 331 flow identification works correctly. 333 3.2.3. Resource Segmentation or Slicing 335 3.2.3.1. Inter-segment Attack 337 An attacker can inject traffic, consuming network device resources, 338 thereby affecting DetNet flows. This can be performed using non- 339 DetNet traffic that affects DetNet traffic, or by using DetNet 340 traffic from one DetNet flow that affects traffic from different 341 DetNet flows. 343 3.2.4. Packet Replication and Elimination 345 3.2.4.1. Replication: Increased Attack Surface 347 Redundancy is intended to increase the robustness and survivability 348 of DetNet flows, and replication over multiple paths can potentially 349 mitigate an attack that is limited to a single path. However, the 350 fact that packets are replicated over multiple paths increases the 351 attack surface of the network, i.e., there are more points in the 352 network that may be subject to attacks. 354 3.2.4.2. Replication-related Header Manipulation 356 An attacker can manipulate the replication-related header fields 357 (R-TAG). This capability opens the door for various types of 358 attacks. For example: 360 o Forward both replicas - malicious change of a packet SN (Sequence 361 Number) can cause both replicas of the packet to be forwarded. 362 Note that this attack has a similar outcome to a replay attack. 364 o Eliminate both replicas - SN manipulation can be used to cause 365 both replicas to be eliminated. In this case an attacker that has 366 access to a single path can cause packets from other paths to be 367 dropped, thus compromising some of the advantage of path 368 redundancy. 370 o Flow hijacking - an attacker can hijack a DetNet flow with access 371 to a single path by systematically replacing the SNs on the given 372 path with higher SN values. For example, an attacker can replace 373 every SN value S with a higher value S+C, where C is a constant 374 integer. Thus, the attacker creates a false illusion that the 375 attacked path has the lowest delay, causing all packets from other 376 paths to be eliminated. Once the flow is hijacked the attacker 377 can either replace en route packets with malicious packets, or 378 simply injecting errors, causing the packets to be dropped at 379 their destination. 381 3.2.5. Path Choice 383 3.2.5.1. Path Manipulation 385 An attacker can maliciously change, add, or remove a path, thereby 386 affecting the corresponding DetNet flows that use the path. 388 3.2.5.2. Path Choice: Increased Attack Surface 390 One of the possible consequences of a path manipulation attack is an 391 increased attack surface. Thus, when the attack described in the 392 previous subsection is implemented, it may increase the potential of 393 other attacks to be performed. 395 3.2.6. Control Plane 397 3.2.6.1. Control or Signaling Packet Modification 399 An attacker can maliciously modify en route control packets in order 400 to disrupt or manipulate the DetNet path/resource allocation. 402 3.2.6.2. Control or Signaling Packet Injection 404 An attacker can maliciously inject control packets in order to 405 disrupt or manipulate the DetNet path/resource allocation. 407 3.2.7. Scheduling or Shaping 409 3.2.7.1. Reconnaissance 411 A passive eavesdropper can identify DetNet flows and then gather 412 information about en route DetNet flows, e.g., the number of DetNet 413 flows, their bandwidths, and their schedules. The gathered 414 information can later be used to invoke other attacks on some or all 415 of the flows. 417 Note that in some cases DetNet flows may be identified based on an 418 explicit DetNet header, but in some cases the flow identification may 419 be based on fields from the L3/L4 headers. If L3/L4 headers are 420 involved, for purposes of this draft we assume they are encrypted 421 and/or integrity-protected from external attackers. 423 3.2.8. Time Synchronization Mechanisms 425 An attacker can use any of the attacks described in [RFC7384] to 426 attack the synchronization protocol, thus affecting the DetNet 427 service. 429 3.3. Threat Summary 431 A summary of the attacks that were discussed in this section is 432 presented in Figure 1. For each attack, the table specifies the type 433 of attackers that may invoke the attack. In the context of this 434 summary, the distinction between internal and external attacks is 435 under the assumption that a corresponding security mechanism is being 436 used, and that the corresponding network equipment takes part in this 437 mechanism. 439 +-----------------------------------------+----+----+----+----+ 440 | Attack | Attacker Type | 441 | +---------+---------+ 442 | |Internal |External | 443 | |MITM|Inj.|MITM|Inj.| 444 +-----------------------------------------+----+----+----+----+ 445 |Delay attack | + | | + | | 446 +-----------------------------------------+----+----+----+----+ 447 |DetNet Flow Modification or Spoofing | + | + | | | 448 +-----------------------------------------+----+----+----+----+ 449 |Inter-segment Attack | + | + | | | 450 +-----------------------------------------+----+----+----+----+ 451 |Replication: Increased Attack Surface | + | + | + | + | 452 +-----------------------------------------+----+----+----+----+ 453 |Replication-related Header Manipulation | + | | | | 454 +-----------------------------------------+----+----+----+----+ 455 |Path Manipulation | + | + | | | 456 +-----------------------------------------+----+----+----+----+ 457 |Path Choice: Increased Attack Surface | + | + | + | + | 458 +-----------------------------------------+----+----+----+----+ 459 |Control or Signaling Packet Modification | + | | | | 460 +-----------------------------------------+----+----+----+----+ 461 |Control or Signaling Packet Injection | | + | | | 462 +-----------------------------------------+----+----+----+----+ 463 |Reconnaissance | + | | + | | 464 +-----------------------------------------+----+----+----+----+ 465 |Attacks on Time Sync Mechanisms | + | + | + | + | 466 +-----------------------------------------+----+----+----+----+ 468 Figure 1: Threat Analysis Summary 470 4. Security Threat Impacts 472 This section describes and rates the impact of the attacks described 473 in Section 3. In this section, the impacts as described assume that 474 the associated mitigation is not present or has failed. Mitigations 475 are discussed in Section 5. 477 In computer security, the impact (or consequence) of an incident can 478 be measured in loss of confidentiality, integrity or availability of 479 information. 481 DetNet raises these stakes significantly for OT applications, 482 particularly those which may have been designed to run in an OT-only 483 environment and thus may not have been designed for security in an IT 484 environment with its associated devices, services and protocols. 486 The severity of various components of the impact of a successful 487 vulnerability exploit to use cases by industry is available in more 488 detail in [RFC8578]. Each of the use cases in the DetNet Use Cases 489 draft is represented in the table below, including Pro Audio, 490 Electrical Utilities, Industrial M2M (split into two areas, M2M Data 491 Gathering and M2M Control Loop), and others. 493 Components of Impact (left column) include Criticality of Failure, 494 Effects of Failure, Recovery, and DetNet Functional Dependence. 495 Criticality of failure summarizes the seriousness of the impact. The 496 impact of a resulting failure can affect many different metrics that 497 vary greatly in scope and severity. In order to reduce the number of 498 variables, only the following were included: Financial, Health and 499 Safety, People well being, Affect on a single organization, and 500 affect on multiple organizations. Recovery outlines how long it 501 would take for an affected use case to get back to its pre-failure 502 state (Recovery time objective, RTO), and how much of the original 503 service would be lost in between the time of service failure and 504 recovery to original state (Recovery Point Objective, RPO). DetNet 505 dependence maps how much the following DetNet service objectives 506 contribute to impact of failure: Time dependency, data integrity, 507 source node integrity, availability, latency/jitter. 509 The scale of the Impact mappings is low, medium, and high. In some 510 use cases there may be a multitude of specific applications in which 511 DetNet is used. For simplicity this section attempts to average the 512 varied impacts of different applications. This section does not 513 address the overall risk of a certain impact which would require the 514 likelihood of a failure happening. 516 In practice any such ratings will vary from case to case; the ratings 517 shown here are given as examples. 519 Table, Part One (of Two) 520 +------------------+-----------------------------------------+-----+ 521 | | Pro A | Util | Bldg |Wire- | Cell |M2M |M2M | 522 | | | | | less | |Data |Ctrl | 523 +------------------+-----------------------------------------+-----+ 524 | Criticality | Med | Hi | Low | Med | Med | Med | Med | 525 +------------------+-----------------------------------------+-----+ 526 | Effects 527 +------------------+-----------------------------------------+-----+ 528 | Financial | Med | Hi | Med | Med | Low | Med | Med | 529 +------------------+-----------------------------------------+-----+ 530 | Health/Safety | Med | Hi | Hi | Med | Med | Med | Med | 531 +------------------+-----------------------------------------+-----+ 532 | People WB | Med | Hi | Hi | Low | Hi | Low | Low | 533 +------------------+-----------------------------------------+-----+ 534 | Effect 1 org | Hi | Hi | Med | Hi | Med | Med | Med | 535 +------------------+-----------------------------------------+-----+ 536 | Effect >1 org | Med | Hi | Low | Med | Med | Med | Med | 537 +------------------+-----------------------------------------+-----+ 538 |Recovery 539 +------------------+-----------------------------------------+-----+ 540 | Recov Time Obj | Med | Hi | Med | Hi | Hi | Hi | Hi | 541 +------------------+-----------------------------------------+-----+ 542 | Recov Point Obj | Med | Hi | Low | Med | Low | Hi | Hi | 543 +------------------+-----------------------------------------+-----+ 544 |DetNet Dependence 545 +------------------+-----------------------------------------+-----+ 546 | Time Dependency | Hi | Hi | Low | Hi | Med | Low | Hi | 547 +------------------+-----------------------------------------+-----+ 548 | Latency/Jitter | Hi | Hi | Med | Med | Low | Low | Hi | 549 +------------------+-----------------------------------------+-----+ 550 | Data Integrity | Hi | Hi | Med | Hi | Low | Hi | Low | 551 +------------------+-----------------------------------------+-----+ 552 | Src Node Integ | Hi | Hi | Med | Hi | Med | Hi | Hi | 553 +------------------+-----------------------------------------+-----+ 554 | Availability | Hi | Hi | Med | Hi | Low | Hi | Hi | 555 +------------------+-----------------------------------------+-----+ 557 Table, Part Two (of Two) 558 +------------------+--------------------------+ 559 | | Mining | Block | Network | 560 | | | Chain | Slicing | 561 +------------------+--------------------------+ 562 | Criticality | Hi | Med | Hi | 563 +------------------+--------------------------+ 564 | Effects 565 +------------------+--------------------------+ 566 | Financial | Hi | Hi | Hi | 567 +------------------+--------------------------+ 568 | Health/Safety | Hi | Low | Med | 569 +------------------+--------------------------+ 570 | People WB | Hi | Low | Med | 571 +------------------+--------------------------+ 572 | Effect 1 org | Hi | Hi | Hi | 573 +------------------+--------------------------+ 574 | Effect >1 org | Hi | Low | Hi | 575 +------------------+--------------------------+ 576 |Recovery 577 +------------------+--------------------------+ 578 | Recov Time Obj | Hi | Low | Hi | 579 +------------------+--------------------------+ 580 | Recov Point Obj | Hi | Low | Hi | 581 +------------------+--------------------------+ 582 |DetNet Dependence 583 +------------------+--------------------------+ 584 | Time Dependency | Hi | Low | Hi | 585 +------------------+--------------------------+ 586 | Latency/Jitter | Hi | Low | Hi | 587 +------------------+--------------------------+ 588 | Data Integrity | Hi | Hi | Hi | 589 +------------------+--------------------------+ 590 | Src Node Integ | Hi | Hi | Hi | 591 +------------------+--------------------------+ 592 | Availability | Hi | Hi | Hi | 593 +------------------+--------------------------+ 595 Figure 2: Impact of Attacks by Use Case Industry 597 The rest of this section will cover impact of the different groups in 598 more detail. 600 4.1. Delay-Attacks 602 4.1.1. Data Plane Delay Attacks 604 Severely delayed messages in a DetNet link can result in the same 605 behavior as dropped messages in ordinary networks as the services 606 attached to the stream has strict deterministic requirements. 608 For a single path scenario, disruption is a real possibility, whereas 609 in a multipath scenario, large delays or instabilities in one stream 610 can lead to increased buffer and CPU resources on the elimination 611 bridge. 613 A data-plane delay attack on a system controlling substantial moving 614 devices, for example in industrial automation, can cause physical 615 damage. For example, if the network promises a bounded latency of 616 2ms for a flow, yet the machine receives it with 5ms latency, the 617 machine's control loop can become unstable. 619 4.1.2. Control Plane Delay Attacks 621 In and of itself, this is not directly a threat to the DetNet 622 service, but the effects of delaying control messages can have quite 623 adverse effects later. 625 o Delayed tear-down can lead to resource leakage, which in turn can 626 result in failure to allocate new streams finally giving rise to a 627 denial of service attack. 629 o Failure to deliver, or severely delaying, signalling messages 630 adding an end-point to a multicast-group will prevent the new EP 631 from receiving expected frames thus disrupting expected behavior. 633 o Delaying messages removing an EP from a group can lead to loss of 634 privacy as the EP will continue to receive messages even after it 635 is supposedly removed. 637 4.2. Flow Modification and Spoofing 639 4.2.1. Flow Modification 641 ToDo. 643 4.2.2. Spoofing 645 4.2.2.1. Dataplane Spoofing 647 Spoofing dataplane messages can result in increased resource 648 consumptions on the bridges throughout the network as it will 649 increase buffer usage and CPU utilization. This can lead to resource 650 exhaustion and/or increased delay. 652 If the attacker manages to create valid headers, the false messages 653 can be forwarded through the network, using part of the allocated 654 bandwidth. This in turn can cause legitimate messages to be dropped 655 when the budget has been exhausted. 657 Finally, the endpoint will have to deal with invalid messages being 658 delivered to the endpoint instead of (or in addition to) a valid 659 message. 661 4.2.2.2. Control Plane Spoofing 663 A successful control plane spoofing-attack will potentionally have 664 adverse effects. It can do virtually anything from: 666 o modifying existing streams by changing the available bandwidth 668 o add or remove endpoints from a stream 670 o drop streams completly 672 o falsely create new streams (exhaust the systems resources, or to 673 enable streams outside the Network engineer's control) 675 4.3. Segmentation attacks (injection) 677 4.3.1. Data Plane Segmentation 679 Injection of false messages in a DetNet stream could lead to 680 exhaustion of the available bandwidth for a stream if the bridges 681 accounts false messages to the stream's budget. 683 In a multipath scenario, injected messages will cause increased CPU 684 utilization in elimination bridges. If enough paths are subject to 685 malicious injection, the legitimate messages can be dropped. 686 Likewise it can cause an increase in buffer usage. In total, it will 687 consume more resources in the bridges than normal, giving rise to a 688 resource exhaustion attack on the bridges. 690 If a stream is interrupted, the end application will be affected by 691 what is now a non-deterministic stream. 693 4.3.2. Control Plane segmentation 695 A successful Control Plane segmentation attack control messages to be 696 interpreted by nodes in the network, unbeknownst to the central 697 controller or the network engineer. This has the potential to create 699 o new streams (exhausting resources) 701 o drop existing (denial of service) 703 o add/remove end-stations to a multicast group (loss of privacy) 705 o modify the stream attributes (affecting available bandwidth 707 4.4. Replication and Elimination 709 The Replication and Elimination is relevant only to Data Plane 710 messages as Signalling is not subject to multipath routing. 712 4.4.1. Increased Attack Surface 714 Covered briefly in Section 4.3 716 4.4.2. Header Manipulation at Elimination Bridges 718 Covered briefly in Section 4.3 720 4.5. Control or Signaling Packet Modification 722 ToDo. 724 4.6. Control or Signaling Packet Injection 726 ToDo. 728 4.7. Reconnaissance 730 Of all the attacks, this is one of the most difficult to detect and 731 counter. Often, an attacker will start out by observing the traffic 732 going through the network and use the knowledge gathered in this 733 phase to mount future attacks. 735 The attacker can, at their leisure, observe over time all aspects of 736 the messaging and signalling, learning the intent and purpose of all 737 traffic flows. At some later date, possibly at an important time in 738 an operational context, the attacker can launch a multi-faceted 739 attack, possibly in conjunction with some demand for ransom. 741 The flow-id in the header of the data plane-messages gives an 742 attacker a very reliable identifier for DetNet traffic, and this 743 traffic has a high probability of going to lucrative targets. 745 4.8. Attacks on Time Sync Mechanisms 747 ToDo. 749 4.9. Attacks on Path Choice 751 This is covered in part in Section 4.3, and as with Replication and 752 Elimination (Section 4.4, this is relevant for DataPlane messages. 754 5. Security Threat Mitigation 756 This section describes a set of measures that can be taken to 757 mitigate the attacks described in Section 3. These mitigations 758 should be viewed as a toolset that includes several different and 759 diverse tools. Each application or system will typically use a 760 subset of these tools, based on a system-specific threat analysis. 762 5.1. Path Redundancy 764 Description 766 A DetNet flow that can be forwarded simultaneously over multiple 767 paths. Path replication and elimination 768 [I-D.ietf-detnet-architecture] provides resiliency to dropped or 769 delayed packets. This redundancy improves the robustness to 770 failures and to man-in-the-middle attacks. 772 Related attacks 774 Path redundancy can be used to mitigate various man-in-the-middle 775 attacks, including attacks described in Section 3.2.1, 776 Section 3.2.2, Section 3.2.3, and Section 3.2.8. 778 5.2. Integrity Protection 780 Description 782 An integrity protection mechanism, such as a Hash-based Message 783 Authentication Code (HMAC) can be used to mitigate modification 784 attacks. Integrity protection can be used on the data plane 785 header, to prevent its modification and tampering. Integrity 786 protection in the control plane is discussed in Section 5.5. 788 Related attacks 790 Integrity protection mitigates attacks related to modification and 791 tampering, including the attacks described in Section 3.2.2 and 792 Section 3.2.4. 794 5.3. DetNet Node Authentication 796 Description 798 Source authentication verifies the authenticity of DetNet sources, 799 allowing to mitigate spoofing attacks. Note that while integrity 800 protection (Section 5.2) prevents intermediate nodes from 801 modifying information, authentication verfies the source of the 802 information. 804 Related attacks 806 DetNet node authentication is used to mitigate attacks related to 807 spoofing, including the attacks of Section 3.2.2, and 808 Section 3.2.4. 810 5.4. Encryption 812 Description 814 DetNet flows can be forwarded in encrypted form. 816 Related attacks 818 Encryption can be used to mitigate recon attacks (Section 3.2.7). 819 However, for a DetNet network to give differentiated quality of 820 service on a flow-by-flow basis, the network must be able to 821 identify the flows individually. This implies that in a recon 822 attack the attacker may also be able to track individual flows to 823 learn more about the system. 825 Encryption can also provide traffic origin verification, i.e. to 826 verify that each packet in a DetNet flow is from a trusted source, 827 for example as part of ingress filtering. 829 5.4.1. Encryption Considerations for DetNet 831 Any compute time which is required for encryption and decryption 832 processing ('crypto') must be included in the flow latency 833 calculations. Thus, crypto algorithms used in a DetNet must have 834 bounded worst-case execution times, and these values must be used in 835 the latency calculations. 837 Some crypto algorithms are symmetric in encode/decode time (such as 838 AES) and others are asymmetric (such as public key algorithms). 839 There are advantages and disadvantages to the use of either type in a 840 given DetNet context. 842 Asymmetrical crypto is typically not used in networks on a packet-by- 843 packet basis due to its computational cost. For example, if only 844 endpoint checks or checks at a small number of intermediate points 845 are required, asymmetric crypto can be used to authenticate 846 distribution or exchange of a secret symmetric crypto key; a 847 successful check based on that key will provide traffic origin 848 verification, as long as the key is kept secret by the participants. 849 TLS and IKE (for IPsec) are examples of this for endpoint checks. 851 However, if secret symmetrical keys are used for this purpose the key 852 must be given to all relays, which increases the probability of a 853 secret key being leaked. Also, if any relay is compromised or 854 misbehaving it may inject traffic into the flow. 856 Alternatively, asymmetric crypto can provide traffic origin 857 verification at every intermediate node. For example, a DetNet flow 858 can be associated with an (asymmetric) keypair, such that the private 859 key is available to the source of the flow and the public key is 860 distributed with the flow information, allowing verification at every 861 node for every packet. However, this is more computationally 862 expensive. 864 In either case, origin verification also requires replay detection as 865 part of the security protocol to prevent an attacker from recording 866 and resending traffic, e.g., as a denial of service attack on flow 867 forwarding resources. 869 If crypto keys are to be regenerated over the duration of the flow 870 then the time required to accomplish this must be accounted for in 871 the latency calculations. 873 5.5. Control and Signaling Message Protection 875 Description 877 Control and sigaling messages can be protected using 878 authentication and integrity protection mechanisms. 880 Related attacks 882 These mechanisms can be used to mitigate various attacks on the 883 control plane, as described in Section 3.2.6, Section 3.2.8 and 884 Section 3.2.5. 886 5.6. Dynamic Performance Analytics 888 Description 890 Information about the network performance can be gathered in real- 891 time in order to detect anomalies and unusual behavior that may be 892 the symptom of a security attack. The gathered information can be 893 based, for example, on per-flow counters, bandwidth measurement, 894 and monitoring of packet arrival times. Unusual behavior or 895 potentially malicious nodes can be reported to a management 896 system, or can be used as a trigger for taking corrective actions. 897 The information can be tracked by DetNet end systems and transit 898 nodes, and exported to a management system, for example using 899 NETCONF. 901 Related attacks 903 Performance analytics can be used to mitigate various attacks, 904 including the ones described in Section 3.2.1 (Delay Attack), 905 Section 3.2.3 (Resource Segmentation Attack), and Section 3.2.8 906 (Time Sync Attack). 908 For example, in the case of data plane delay attacks, one possible 909 mitigation is to timestamp the data at the source, and timestamp 910 it again at the destination, and if the resulting latency exceeds 911 the promised bound, discard that data and warn the operator (and/ 912 or enter a fail-safe mode). 914 5.7. Mitigation Summary 916 The following table maps the attacks of Section 3 to the impacts of 917 Section 4, and to the mitigations of the current section. Each row 918 specifies an attack, the impact of this attack if it is successfully 919 implemented, and possible mitigation methods. 921 +----------------------+---------------------+---------------------+ 922 | Attack | Impact | Mitigations | 923 +----------------------+---------------------+---------------------+ 924 |Delay Attack |-Non-deterministic |-Path redundancy | 925 | | delay |-Performance | 926 | |-Data disruption | analytics | 927 | |-Increased resource | | 928 | | consumption | | 929 +----------------------+---------------------+---------------------+ 930 |Reconnaissance |-Enabler for other |-Encryption | 931 | | attacks | | 932 +----------------------+---------------------+---------------------+ 933 |DetNet Flow Modificat-|-Increased resource |-Path redundancy | 934 |ion or Spoofing | consumption |-Integrity protection| 935 | |-Data disruption |-DetNet Node | 936 | | | authentication | 937 +----------------------+---------------------+---------------------+ 938 |Inter-Segment Attack |-Increased resource |-Path redundancy | 939 | | consumption |-Performance | 940 | |-Data disruption | analytics | 941 +----------------------+---------------------+---------------------+ 942 |Replication: Increased|-All impacts of other|-Integrity protection| 943 |attack surface | attacks |-DetNet Node | 944 | | | authentication | 945 +----------------------+---------------------+---------------------+ 946 |Replication-related |-Non-deterministic |-Integrity protection| 947 |Header Manipulation | delay |-DetNet Node | 948 | |-Data disruption | authentication | 949 +----------------------+---------------------+---------------------+ 950 |Path Manipulation |-Enabler for other |-Control message | 951 | | attacks | protection | 952 +----------------------+---------------------+---------------------+ 953 |Path Choice: Increased|-All impacts of other|-Control message | 954 |Attack Surface | attacks | protection | 955 +----------------------+---------------------+---------------------+ 956 |Control or Signaling |-Increased resource |-Control message | 957 |Packet Modification | consumption | protection | 958 | |-Non-deterministic | | 959 | | delay | | 960 | |-Data disruption | | 961 +----------------------+---------------------+---------------------+ 962 |Control or Signaling |-Increased resource |-Control message | 963 |Packet Injection | consumption | protection | 964 | |-Non-deterministic | | 965 | | delay | | 966 | |-Data disruption | | 967 +----------------------+---------------------+---------------------+ 968 |Attacks on Time Sync |-Non-deterministic |-Path redundancy | 969 |Mechanisms | delay |-Control message | 970 | |-Increased resource | protection | 971 | | consumption |-Performance | 972 | |-Data disruption | analytics | 973 +----------------------+---------------------+---------------------+ 975 Figure 3: Mapping Attacks to Impact and Mitigations 977 6. Association of Attacks to Use Cases 979 Different attacks can have different impact and/or mitigation 980 depending on the use case, so we would like to make this association 981 in our analysis. However since there is a potentially unbounded list 982 of use cases, we categorize the attacks with respect to the common 983 themes of the use cases as identified in the Use Case Common Themes 984 section of the DetNet Use Cases draft [RFC8578]. 986 See also Figure 2 for a mapping of the impact of attacks per use case 987 by industry. 989 6.1. Use Cases by Common Themes 991 In this section we review each theme and discuss the attacks that are 992 applicable to that theme, as well as anything specific about the 993 impact and mitigations for that attack with respect to that theme. 994 The table Figure 5 then provides a summary of the attacks that are 995 applicable to each theme. 997 6.1.1. Network Layer - AVB/TSN Ethernet 999 DetNet is expected to run over various transmission mediums, with 1000 Ethernet being explicitly supported. Attacks such as Delay or 1001 Reconnaissance might be implemented differently on a different 1002 transmission medium, however the impact on the DetNet as a whole 1003 would be essentially the same. We thus conclude that all attacks and 1004 impacts that would be applicable to DetNet over Ethernet (i.e. all 1005 those named in this draft) would also be applicable to DetNet over 1006 other transmission mediums. 1008 With respect to mitigations, some methods are specific to the 1009 Ethernet medium, for example time-aware scheduling using 802.1Qbv can 1010 protect against excessive use of bandwidth at the ingress - for other 1011 mediums, other mitigations would have to be implemented to provide 1012 analogous protection. 1014 6.1.2. Central Administration 1016 A DetNet network is expected to be controlled by a centralized 1017 network configuration and control system (CNC). Such a system may be 1018 in a single central location, or it may be distributed across 1019 multiple control entities that function together as a unified control 1020 system for the network. 1022 In this draft we distinguish between attacks on the DetNet Control 1023 plane vs. Data plane. But is an attack affecting control plane 1024 packets synonymous with an attack on the CNC itself? For purposes of 1025 this draft let us consider an attack on the CNC itself to be out of 1026 scope, and consider all attacks named in this draft which are 1027 relevant to control plane packets to be relevant to this theme, 1028 including Path Manipulation, Path Choice, Control Packet Modification 1029 or Injection, Reconaissance and Attacks on Time Sync Mechanisms. 1031 6.1.3. Hot Swap 1033 A DetNet network is not expected to be "plug and play" - it is 1034 expected that there is some centralized network configuration and 1035 control system. However, the ability to "hot swap" components (e.g. 1036 due to malfunction) is similar enough to "plug and play" that this 1037 kind of behavior may be expected in DetNet networks, depending on the 1038 implementation. 1040 An attack surface related to Hot Swap is that the DetNet network must 1041 at least consider input at runtime from devices that were not part of 1042 the initial configuration of the network. Even a "perfect" (or 1043 "hitless") replacement of a device at runtime would not necessarily 1044 be ideal, since presumably one would want to distinguish it from the 1045 original for OAM purposes (e.g. to report hot swap of a failed 1046 device). 1048 This implies that an attack such as Flow Modification, Spoofing or 1049 Inter-segment (which could introduce packets from a "new" device 1050 (i.e. one heretofore unknown on the network) could be used to exploit 1051 the need to consider such packets (as opposed to rejecting them out 1052 of hand as one would do if one did not have to consider introduction 1053 of a new device). 1055 Similarly if the network was designed to support runtime replacement 1056 of a clock device, then presence (or apparent presence) and thus 1057 consideration of packets from a new such device could affect the 1058 network, or the time sync of the network, for example by initiating a 1059 new Best Master Clock selection process. Thus attacks on time sync 1060 should be considered when designing hot swap type functionality. 1062 6.1.4. Data Flow Information Models 1064 Data Flow Information Models specific to DetNet networks are to be 1065 specified by DetNet. Thus they are "new" and thus potentially 1066 present a new attack surface. Does the threat take advantage of any 1067 aspect of our new Data Flow Info Models? 1069 This is TBD, thus there are no specific entries in our table, however 1070 that does not imply that there could be no relevant attacks. 1072 6.1.5. L2 and L3 Integration 1074 A DetNet network integrates Layer 2 (bridged) networks (e.g. AVB/TSN 1075 LAN) and Layer 3 (routed) networks via the use of well-known 1076 protocols such as IPv6, MPLS-PW, and Ethernet. Presumably security 1077 considerations applicable directly to those individual protocols is 1078 not specific to DetNet, and thus out of scope for this draft. 1079 However enabling DetNet to coordinate Layer 2 and Layer 3 behavior 1080 will require some additions to existing protocols (see draft-dt- 1081 detnet-dp-alt) and any such new work can introduce new attack 1082 surfaces. 1084 This is TBD, thus there are no specific entries in our table, however 1085 that does not imply that there could be no relevant attacks. 1087 6.1.6. End-to-End Delivery 1089 Packets sent over DetNet are guaranteed not to be dropped by the 1090 network due to congestion. (Packets may however be dropped for 1091 intended reasons, e.g. per security measures). 1093 A Data plane attack may force packets to be dropped, for example a 1094 "long" Delay or Replication/Elimination or Flow Modification attack. 1096 The same result might be obtained by a Control plane attack, e.g. 1097 Path Manipulation or Signaling Packet Modification. 1099 It may be that such attacks are limited to Internal MITM attackers, 1100 but other possibilities should be considered. 1102 An attack may also cause packets that should not be delivered to be 1103 delivered, such as by forcing packets from one (e.g. replicated) path 1104 to be preferred over another path when they should not be 1105 (Replication attack), or by Flow Modification, or by Path Choice or 1106 Packet Injection. A Time Sync attack could cause a system that was 1107 expecting certain packets at certain times to accept unintended 1108 packets based on compromised system time or time windowing in the 1109 scheduler. 1111 6.1.7. Proprietary Deterministic Ethernet Networks 1113 There are many proprietary non-interoperable deterministic Ethernet- 1114 based networks currently available; DetNet is intended to provide an 1115 open-standards-based alternative to such networks. In cases where a 1116 DetNet intersects with remnants of such networks or their protocols, 1117 such as by protocol emulation or access to such a network via a 1118 gateway, new attack surfaces can be opened. 1120 For example an Inter-Segment or Control plane attack such as Path 1121 Manipulation, Path Choice or Control Packet Modification/Injection 1122 could be used to exploit commands specific to such a protocol, or 1123 that are interpreted differently by the different protocols or 1124 gateway. 1126 6.1.8. Replacement for Proprietary Fieldbuses 1128 There are many proprietary "field buses" used in today's industrial 1129 and other industries; DetNet is intended to provide an open- 1130 standards-based alternative to such buses. In cases where a DetNet 1131 intersects with such fieldbuses or their protocols, such as by 1132 protocol emulation or access via a gateway, new attack surfaces can 1133 be opened. 1135 For example an Inter-Segment or Control plane attack such as Path 1136 Manipulation, Path Choice or Control Packet Modification/Injection 1137 could be used to exploit commands specific to such a protocol, or 1138 that are interpreted differently by the different protocols or 1139 gateway. 1141 6.1.9. Deterministic vs Best-Effort Traffic 1143 DetNet is intended to support coexistence of time-sensitive 1144 operational (OT, deterministic) traffic and information (IT, "best 1145 effort") traffic on the same ("unified") network. 1147 The presence of IT traffic on a network carrying OT traffic has long 1148 been considered insecure design [reference needed here]. With 1149 DetNet, this coexistance will become more common, and mitigations 1150 will need to be established. The fact that the IT traffic on a 1151 DetNet is limited to a corporate controlled network makes this a less 1152 difficult problem compared to being exposed to the open Internet, 1153 however this aspect of DetNet security should not be underestimated. 1155 Most of the themes described in this draft address OT (reserved) 1156 streams - this item is intended to address issues related to IT 1157 traffic on a DetNet. 1159 An Inter-segment attack can flood the network with IT-type traffic 1160 with the intent of disrupting handling of IT traffic, and/or the goal 1161 of interfering with OT traffic. Presumably if the stream reservation 1162 and isolation of the DetNet is well-designed (better-designed than 1163 the attack) then interference with OT traffic should not result from 1164 an attack that floods the network with IT traffic. 1166 However the DetNet's handling of IT traffic may not (by design) be as 1167 resilient to DOS attack, and thus designers must be otherwise 1168 prepared to mitigate DOS attacks on IT traffic in a DetNet. 1170 6.1.10. Deterministic Flows 1172 Reserved bandwidth data flows (deterministic flows) must provide the 1173 allocated bandwidth, and must be isolated from each other. 1175 A Spoofing or Inter-segment attack which adds packet traffic to a 1176 bandwidth-reserved stream could cause that stream to occupy more 1177 bandwidth than it is allocated, resulting in interference with other 1178 deterministic flows. 1180 A Flow Modification or Spoofing or Header Manipulation or Control 1181 Packet Modification attack could cause packets from one flow to be 1182 directed to another flow, thus breaching isolation between the flows. 1184 6.1.11. Unused Reserved Bandwidth 1186 If bandwidth reservations are made for a stream but the associated 1187 bandwidth is not used at any point in time, that bandwidth is made 1188 available on the network for best-effort traffic. If the owner of 1189 the reserved stream then starts transmitting again, the bandwidth is 1190 no longer available for best-effort traffic, on a moment-to-moment 1191 basis. (Such "temporarily available" bandwidth is not available for 1192 time-sensitive traffic, which must have its own reservation). 1194 An Inter-segment attack could flood the network with IT traffic, 1195 interfering with the intended IT traffic. 1197 A Flow Modification or Spoofing or Control Packet Modification or 1198 Injection attack could cause extra bandwidth to be reserved by a new 1199 or existing stream, thus making it unavailable for use by best-effort 1200 traffic. 1202 6.1.12. Interoperability 1204 The DetNet network specifications are intended to enable an ecosystem 1205 in which multiple vendors can create interoperable products, thus 1206 promoting device diversity and potentially higher numbers of each 1207 device manufactured. Does the threat take advantage of differences 1208 in implementation of "interoperable" products made by different 1209 vendors? 1211 This is TBD, thus there are no specific entries in our table, however 1212 that does not imply that there could be no relevant attacks. 1214 6.1.13. Cost Reductions 1216 The DetNet network specifications are intended to enable an ecosystem 1217 in which multiple vendors can create interoperable products, thus 1218 promoting higher numbers of each device manufactured, promoting cost 1219 reduction and cost competition among vendors. Does the threat take 1220 advantage of "low cost" HW or SW components or other "cost-related 1221 shortcuts" that might be present in devices? 1223 This is TBD, thus there are no specific entries in our table, however 1224 that does not imply that there could be no relevant attacks. 1226 6.1.14. Insufficiently Secure Devices 1228 The DetNet network specifications are intended to enable an ecosystem 1229 in which multiple vendors can create interoperable products, thus 1230 promoting device diversity and potentially higher numbers of each 1231 device manufactured. Does the threat attack "naivete" of SW, for 1232 example SW that was not designed to be sufficiently secure (or secure 1233 at all) but is deployed on a DetNet network that is intended to be 1234 highly secure? (For example IoT exploits like the Mirai video-camera 1235 botnet ([MIRAI]). 1237 This is TBD, thus there are no specific entries in our table, however 1238 that does not imply that there could be no relevant attacks. 1240 6.1.15. DetNet Network Size 1242 DetNet networks range in size from very small, e.g. inside a single 1243 industrial machine, to very large, for example a Utility Grid network 1244 spanning a whole country. 1246 The size of the network might be related to how the attack is 1247 introduced into the network, for example if the entire network is 1248 local, there is a threat that power can be cut to the entire network. 1249 If the network is large, perhaps only a part of the network is 1250 attacked. 1252 A Delay attack might be as relevant to a small network as to a large 1253 network, although the amount of delay might be different. 1255 Attacks sourced from IT traffic might be more likely in large 1256 networks, since more people might have access to the network. 1257 Similarly Path Manipulation, Path Choice and Time Sync attacks seem 1258 more likely relevant to large networks. 1260 6.1.16. Multiple Hops 1262 Large DetNet networks (e.g. a Utility Grid network) may involve many 1263 "hops" over various kinds of links for example radio repeaters, 1264 microwave links, fiber optic links, etc.. 1266 An attack that takes advantage of flaws (or even normal operation) in 1267 the device drivers for the various links (through internal knowledge 1268 of how the individual driver or firmware operates, perhaps like the 1269 Stuxnet attack) could take proportionately greater advantage of this 1270 topology. We don't currently have an attack like this defined; we 1271 have only "protocol" (time or packet) based attacks. Perhaps we need 1272 to define an attack like this? Or is that out of scope for DetNet? 1273 It is also possible that this DetNet topology will not be in as 1274 common use as other more homogeneous topologies so there may be more 1275 opportunity for attackers to exploit software and/or protocol flaws 1276 in the implementations which have not been wrung out by extensive 1277 use, particularly in the case of early adopters. 1279 Of the attacks we have defined, the ones identified above as relevant 1280 to "large" networks seem to be most relevant. 1282 6.1.17. Level of Service 1284 A DetNet is expected to provide means to configure the network that 1285 include querying network path latency, requesting bounded latency for 1286 a given stream, requesting worst case maximum and/or minimum latency 1287 for a given path or stream, and so on. It is an expected case that 1288 the network cannot provide a given requested service level. In such 1289 cases the network control system should reply that the requested 1290 service level is not available (as opposed to accepting the parameter 1291 but then not delivering the desired behavior). 1293 Control plane attacks such as Signaling Packet Modification and 1294 Injection could be used to modify or create control traffic that 1295 could interfere with the process of a user requesting a level of 1296 service and/or the network's reply. 1298 Reconnaissance could be used to characterize flows and perhaps target 1299 specific flows for attack via the Control plane as noted above. 1301 6.1.18. Bounded Latency 1303 DetNet provides the expectation of guaranteed bounded latency. 1305 Delay attacks can cause packets to miss their agreed-upon latency 1306 boundaries. 1308 Time Sync attacks can corrupt the system's time reference, resulting 1309 in missed latency deadlines (with respect to the "correct" time 1310 reference). 1312 6.1.19. Low Latency 1314 Applications may require "extremely low latency" however depending on 1315 the application these may mean very different latency values; for 1316 example "low latency" across a Utility grid network is on a different 1317 time scale than "low latency" in a motor control loop in a small 1318 machine. The intent is that the mechanisms for specifying desired 1319 latency include wide ranges, and that architecturally there is 1320 nothing to prevent arbitrarily low latencies from being implemented 1321 in a given network. 1323 Attacks on the Control plane (as described in the Level of Service 1324 theme) and Delay and Time attacks (as described in the Bounded 1325 Latency theme) both apply here. 1327 6.1.20. Bounded Jitter (Latency Variation) 1329 DetNet is expected to provide bounded jitter (packet to packet 1330 latency variation). 1332 Delay attacks can cause packets to vary in their arrival times, 1333 resulting in packet to packet latency variation, thereby violating 1334 the jitter specification. 1336 6.1.21. Symmetrical Path Delays 1338 Some applications would like to specify that the transit delay time 1339 values be equal for both the transmit and return paths. 1341 Delay attacks can cause path delays to differ. 1343 Time Sync attacks can corrupt the system's time reference, resulting 1344 in differing path delays (with respect to the "correct" time 1345 reference). 1347 6.1.22. Reliability and Availability 1349 DetNet based systems are expected to be implemented with essentially 1350 arbitrarily high availability (for example 99.9999% up time, or even 1351 12 nines). The intent is that the DetNet designs should not make any 1352 assumptions about the level of reliability and availability that may 1353 be required of a given system, and should define parameters for 1354 communicating these kinds of metrics within the network. 1356 Any attack on the system, of any type, can affect its overall 1357 reliability and availability, thus in our table we have marked every 1358 attack. Since every DetNet depends to a greater or lesser degree on 1359 reliability and availability, this essentially means that all 1360 networks have to mitigate all attacks, which to a greater or lesser 1361 degree defeats the purpose of associating attacks with use cases. It 1362 also underscores the difficulty of designing "extremely high 1363 reliability" networks. I hope that in future drafts we can say 1364 something more useful here. 1366 6.1.23. Redundant Paths 1368 DetNet based systems are expected to be implemented with essentially 1369 arbitrarily high reliability/availability. A strategy used by DetNet 1370 for providing such extraordinarily high levels of reliability is to 1371 provide redundant paths that can be seamlessly switched between, all 1372 the while maintaining the required performance of that system. 1374 Replication-related attacks are by definition applicable here. 1375 Control plane attacks can also interfere with the configuration of 1376 redundant paths. 1378 6.1.24. Security Measures 1380 A DetNet network must be made secure against devices failures, 1381 attackers, misbehaving devices, and so on. Does the threat affect 1382 such security measures themselves, e.g. by attacking SW designed to 1383 protect against device failure? 1385 This is TBD, thus there are no specific entries in our table, however 1386 that does not imply that there could be no relevant attacks. 1388 6.2. Attack Types by Use Case Common Theme 1390 The following table lists the attacks of Section 3, assigning a 1391 number to each type of attack. That number is then used as a short 1392 form identifier for the attack in Figure 5. 1394 +--+----------------------------------------+----------------------+ 1395 | | Attack | Section | 1396 +--+----------------------------------------+----------------------+ 1397 | 1|Delay Attack | Section 3.2.1 | 1398 +--+----------------------------------------+----------------------+ 1399 | 2|DetNet Flow Modification or Spoofing | Section 3.2.2 | 1400 +--+----------------------------------------+----------------------+ 1401 | 3|Inter-Segment Attack | Section 3.2.3 | 1402 +--+----------------------------------------+----------------------+ 1403 | 4|Replication: Increased attack surface | Section 3.2.4.1 | 1404 +--+----------------------------------------+----------------------+ 1405 | 5|Replication-related Header Manipulation | Section 3.2.4.2 | 1406 +--+----------------------------------------+----------------------+ 1407 | 6|Path Manipulation | Section 3.2.5.1 | 1408 +--+----------------------------------------+----------------------+ 1409 | 7|Path Choice: Increased Attack Surface | Section 3.2.5.2 | 1410 +--+----------------------------------------+----------------------+ 1411 | 8|Control or Signaling Packet Modification| Section 3.2.6.1 | 1412 +--+----------------------------------------+----------------------+ 1413 | 9|Control or Signaling Packet Injection | Section 3.2.6.2 | 1414 +--+----------------------------------------+----------------------+ 1415 |10|Reconnaissance | Section 3.2.7 | 1416 +--+----------------------------------------+----------------------+ 1417 |11|Attacks on Time Sync Mechanisms | Section 3.2.8 | 1418 +--+----------------------------------------+----------------------+ 1420 Figure 4: List of Attacks 1422 The following table maps the use case themes presented in this memo 1423 to the attacks of Figure 4. Each row specifies a theme, and the 1424 attacks relevant to this theme are marked with a '+'. 1426 +----------------------------+--------------------------------+ 1427 | Theme | Attack | 1428 | +--+--+--+--+--+--+--+--+--+--+--+ 1429 | | 1| 2| 3| 4| 5| 6| 7| 8| 9|10|11| 1430 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1431 |Network Layer - AVB/TSN Eth.| +| +| +| +| +| +| +| +| +| +| +| 1432 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1433 |Central Administration | | | | | | +| +| +| +| +| +| 1434 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1435 |Hot Swap | | +| +| | | | | | | | +| 1436 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1437 |Data Flow Information Models| | | | | | | | | | | | 1438 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1439 |L2 and L3 Integration | | | | | | | | | | | | 1440 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1441 |End-to-end Delivery | +| +| +| +| +| +| +| +| +| | +| 1442 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1443 |Proprietary Deterministic | | | +| | | +| +| +| +| | | 1444 |Ethernet Networks | | | | | | | | | | | | 1445 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1446 |Replacement for Proprietary | | | +| | | +| +| +| +| | | 1447 |Fieldbuses | | | | | | | | | | | | 1448 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1449 |Deterministic vs. Best- | | | +| | | | | | | | | 1450 |Effort Traffic | | | | | | | | | | | | 1451 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1452 |Deterministic Flows | | +| +| | +| +| | +| | | | 1453 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1454 |Unused Reserved Bandwidth | | +| +| | | | | +| +| | | 1455 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1456 |Interoperability | | | | | | | | | | | | 1457 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1458 |Cost Reductions | | | | | | | | | | | | 1459 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1460 |Insufficiently Secure | | | | | | | | | | | | 1461 |Devices | | | | | | | | | | | | 1462 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1463 |DetNet Network Size | +| | | | | +| +| | | | +| 1464 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1465 |Multiple Hops | +| +| | | | +| +| | | | +| 1466 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1467 |Level of Service | | | | | | | | +| +| +| | 1468 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1469 |Bounded Latency | +| | | | | | | | | | +| 1470 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1471 |Low Latency | +| | | | | | | +| +| +| +| 1472 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1473 |Bounded Jitter | +| | | | | | | | | | | 1474 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1475 |Symmetric Path Delays | +| | | | | | | | | | +| 1476 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1477 |Reliability and Availability| +| +| +| +| +| +| +| +| +| +| +| 1478 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1479 |Redundant Paths | | | | +| +| | | +| +| | | 1480 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1481 |Security Measures | | | | | | | | | | | | 1482 +----------------------------+--+--+--+--+--+--+--+--+--+--+--+ 1484 Figure 5: Mapping Between Themes and Attacks 1486 6.3. Security Considerations for OAM Traffic 1488 This section considers DetNet-specific security considerations for 1489 packet traffic that is generated and transmitted over a DetNet as 1490 part of OAM (Operations, Administration and Maintenance). For 1491 purposes of this discussion, OAM traffic falls into one of two basic 1492 types: 1494 o OAM traffic generated by the network itself. The additional 1495 bandwidth required for such packets is added by the network 1496 administration, presumably transparent to the customer. Security 1497 considerations for such traffic are not DetNet-specific (apart 1498 from such traffic being subject to the same DetNet-specific 1499 security considerations as any other DetNet data flow) and are 1500 thus not covered in this document. 1502 o OAM traffic generated by the customer. From a DetNet security 1503 point of view, DetNet security considerations for such traffic are 1504 exactly the same as for any other customer data flows. 1506 Thus OAM traffic presents no additional (i.e. OAM-specific) DetNet 1507 security considerations. 1509 7. DetNet Technology-Specific Threats 1511 Section 3 described threats which are independent of the DetNet 1512 implementation. This section considers threats related to the 1513 specific technologies referenced in 1514 [I-D.ietf-detnet-data-plane-framework] which have not already been 1515 enumerated in Section 3. 1517 As in this document in general, this section only enumerates security 1518 aspects which are unique to providing the specific quality of service 1519 aspects of DetNet, which are primarily to deliver data flows with 1520 extremely low packet loss rates and bounded end-to-end delivery 1521 latency. The primary considerations for the data plane specifically 1522 are to maintain integrity of data and delivery of the associated 1523 DetNet service traversing the DetNet network. 1525 As noted in [I-D.ietf-detnet-architecture], DetNet operates at the IP 1526 layer ([I-D.ietf-detnet-ip]) and delivers service over sub-layer 1527 technologies such as MPLS ([I-D.ietf-detnet-mpls]) and IEEE 802.1 1528 Time-Sensitive Networking (TSN) ([I-D.ietf-detnet-ip-over-tsn]). 1530 Application flows can be protected through whatever means is provided 1531 by the underlying technology. For example, technology-specific 1532 encryption may be used, such as that provided by IPSec [RFC4301] for 1533 IP flows and/or by an underlying sub-net using MACSec 1534 [IEEE802.1AE-2018] for IP over Ethernet (Layer-2) flows. 1536 Sections below discuss threats specific to IP, MPLS, and TSN in more 1537 detail. 1539 7.1. IP 1541 The IP protocol has a long history of security considerations and 1542 architectural protection mechanisms. From a data plane perspective 1543 DetNet does not add or modify any IP header information, and its use 1544 as a DetNet Data Plane does not introduce any new security issues 1545 that were not there before, apart from those already described in the 1546 data-plane-independent threats section Section 3. 1548 Thus the security considerations for a DetNet based on an IP data 1549 plane are purely inherited from the rich IP Security literature and 1550 code/application base, and the data-plane-independent section of this 1551 document. 1553 7.2. MPLS 1555 Editor's Note: To Be Written. 1557 7.3. TSN 1559 Editor's Note: To Be Written. 1561 8. Appendix A: DetNet Draft Security-Related Statements 1563 This section collects the various statements in the currently 1564 existing DetNet Working Group drafts. For each draft, the section 1565 name and number of the quoted section is shown. The text shown here 1566 is the work of the original draft authors, quoted verbatim from the 1567 drafts. The intention is to explicitly quote all relevant text, not 1568 to summarize it. 1570 8.1. Architecture (draft 8) 1572 8.1.1. Fault Mitigation (sec 4.5) 1574 One key to building robust real-time systems is to reduce the 1575 infinite variety of possible failures to a number that can be 1576 analyzed with reasonable confidence. DetNet aids in the process by 1577 providing filters and policers to detect DetNet packets received on 1578 the wrong interface, or at the wrong time, or in too great a volume, 1579 and to then take actions such as discarding the offending packet, 1580 shutting down the offending DetNet flow, or shutting down the 1581 offending interface. 1583 It is also essential that filters and service remarking be employed 1584 at the network edge to prevent non-DetNet packets from being mistaken 1585 for DetNet packets, and thus impinging on the resources allocated to 1586 DetNet packets. 1588 There exist techniques, at present and/or in various stages of 1589 standardization, that can perform these fault mitigation tasks that 1590 deliver a high probability that misbehaving systems will have zero 1591 impact on well-behaved DetNet flows, except of course, for the 1592 receiving interface(s) immediately downstream of the misbehaving 1593 device. Examples of such techniques include traffic policing 1594 functions (e.g. [RFC2475]) and separating flows into per-flow rate- 1595 limited queues. 1597 8.1.2. Security Considerations (sec 7) 1599 Security in the context of Deterministic Networking has an added 1600 dimension; the time of delivery of a packet can be just as important 1601 as the contents of the packet, itself. A man-in-the-middle attack, 1602 for example, can impose, and then systematically adjust, additional 1603 delays into a link, and thus disrupt or subvert a real-time 1604 application without having to crack any encryption methods employed. 1605 See [RFC7384] for an exploration of this issue in a related context. 1607 Furthermore, in a control system where millions of dollars of 1608 equipment, or even human lives, can be lost if the DetNet QoS is not 1609 delivered, one must consider not only simple equipment failures, 1610 where the box or wire instantly becomes perfectly silent, but bizarre 1611 errors such as can be caused by software failures. Because there is 1612 essential no limit to the kinds of failures that can occur, 1613 protecting against realistic equipment failures is indistinguishable, 1614 in most cases, from protecting against malicious behavior, whether 1615 accidental or intentional. 1617 Security must cover: 1619 o Protection of the signaling protocol 1621 o Authentication and authorization of the controlling nodes 1623 o Identification and shaping of the flows 1625 8.2. Data Plane Alternatives (draft 4) 1627 8.2.1. Security Considerations (sec 7) 1629 This document does not add any new security considerations beyond 1630 what the referenced technologies already have. 1632 8.3. Problem Statement (draft 5) 1634 8.3.1. Security Considerations (sec 5) 1636 Security in the context of Deterministic Networking has an added 1637 dimension; the time of delivery of a packet can be just as important 1638 as the contents of the packet, itself. A man-in-the-middle attack, 1639 for example, can impose, and then systematically adjust, additional 1640 delays into a link, and thus disrupt or subvert a real-time 1641 application without having to crack any encryption methods employed. 1642 See [RFC7384] for an exploration of this issue in a related context. 1644 Typical control networks today rely on complete physical isolation to 1645 prevent rogue access to network resources. DetNet enables the 1646 virtualization of those networks over a converged IT/OT 1647 infrastructure. Doing so, DetNet introduces an additional risk that 1648 flows interact and interfere with one another as they share physical 1649 resources such as Ethernet trunks and radio spectrum. The 1650 requirement is that there is no possible data leak from and into a 1651 deterministic flow, and in a more general fashion there is no 1652 possible influence whatsoever from the outside on a deterministic 1653 flow. The expectation is that physical resources are effectively 1654 associated with a given flow at a given point of time. In that 1655 model, Time Sharing of physical resources becomes transparent to the 1656 individual flows which have no clue whether the resources are used by 1657 other flows at other times. 1659 Security must cover: 1661 o Protection of the signaling protocol 1663 o Authentication and authorization of the controlling nodes 1665 o Identification and shaping of the flows 1667 o Isolation of flows from leakage and other influences from any 1668 activity sharing physical resources 1670 8.4. Use Cases (draft 11) 1672 8.4.1. (Utility Networks) Security Current Practices and Limitations 1673 (sec 3.2.1) 1675 Grid monitoring and control devices are already targets for cyber 1676 attacks, and legacy telecommunications protocols have many intrinsic 1677 network-related vulnerabilities. For example, DNP3, Modbus, 1678 PROFIBUS/PROFINET, and other protocols are designed around a common 1679 paradigm of request and respond. Each protocol is designed for a 1680 master device such as an HMI (Human Machine Interface) system to send 1681 commands to subordinate slave devices to retrieve data (reading 1682 inputs) or control (writing to outputs). Because many of these 1683 protocols lack authentication, encryption, or other basic security 1684 measures, they are prone to network-based attacks, allowing a 1685 malicious actor or attacker to utilize the request-and-respond system 1686 as a mechanism for command-and-control like functionality. Specific 1687 security concerns common to most industrial control, including 1688 utility telecommunication protocols include the following: 1690 o Network or transport errors (e.g. malformed packets or excessive 1691 latency) can cause protocol failure. 1693 o Protocol commands may be available that are capable of forcing 1694 slave devices into inoperable states, including powering-off 1695 devices, forcing them into a listen-only state, disabling 1696 alarming. 1698 o Protocol commands may be available that are capable of restarting 1699 communications and otherwise interrupting processes. 1701 o Protocol commands may be available that are capable of clearing, 1702 erasing, or resetting diagnostic information such as counters and 1703 diagnostic registers. 1705 o Protocol commands may be available that are capable of requesting 1706 sensitive information about the controllers, their configurations, 1707 or other need-to-know information. 1709 o Most protocols are application layer protocols transported over 1710 TCP; therefore it is easy to transport commands over non-standard 1711 ports or inject commands into authorized traffic flows. 1713 o Protocol commands may be available that are capable of 1714 broadcasting messages to many devices at once (i.e. a potential 1715 DoS). 1717 o Protocol commands may be available to query the device network to 1718 obtain defined points and their values (i.e. a configuration 1719 scan). 1721 o Protocol commands may be available that will list all available 1722 function codes (i.e. a function scan). 1724 o These inherent vulnerabilities, along with increasing connectivity 1725 between IT an OT networks, make network-based attacks very 1726 feasible. 1728 o Simple injection of malicious protocol commands provides control 1729 over the target process. Altering legitimate protocol traffic can 1730 also alter information about a process and disrupt the legitimate 1731 controls that are in place over that process. A man-in-the-middle 1732 attack could provide both control over a process and 1733 misrepresentation of data back to operator consoles. 1735 8.4.2. (Utility Networks) Security Trends in Utility Networks (sec 1736 3.3.3) 1738 Although advanced telecommunications networks can assist in 1739 transforming the energy industry by playing a critical role in 1740 maintaining high levels of reliability, performance, and 1741 manageability, they also introduce the need for an integrated 1742 security infrastructure. Many of the technologies being deployed to 1743 support smart grid projects such as smart meters and sensors can 1744 increase the vulnerability of the grid to attack. Top security 1745 concerns for utilities migrating to an intelligent smart grid 1746 telecommunications platform center on the following trends: 1748 o Integration of distributed energy resources 1750 o Proliferation of digital devices to enable management, automation, 1751 protection, and control 1753 o Regulatory mandates to comply with standards for critical 1754 infrastructure protection 1756 o Migration to new systems for outage management, distribution 1757 automation, condition-based maintenance, load forecasting, and 1758 smart metering 1760 o Demand for new levels of customer service and energy management 1762 This development of a diverse set of networks to support the 1763 integration of microgrids, open-access energy competition, and the 1764 use of network-controlled devices is driving the need for a converged 1765 security infrastructure for all participants in the smart grid, 1766 including utilities, energy service providers, large commercial and 1767 industrial, as well as residential customers. Securing the assets of 1768 electric power delivery systems (from the control center to the 1769 substation, to the feeders and down to customer meters) requires an 1770 end-to-end security infrastructure that protects the myriad of 1771 telecommunications assets used to operate, monitor, and control power 1772 flow and measurement. 1774 "Cyber security" refers to all the security issues in automation and 1775 telecommunications that affect any functions related to the operation 1776 of the electric power systems. Specifically, it involves the 1777 concepts of: 1779 o Integrity : data cannot be altered undetectably 1781 o Authenticity : the telecommunications parties involved must be 1782 validated as genuine 1784 o Authorization : only requests and commands from the authorized 1785 users can be accepted by the system 1787 o Confidentiality : data must not be accessible to any 1788 unauthenticated users 1790 When designing and deploying new smart grid devices and 1791 telecommunications systems, it is imperative to understand the 1792 various impacts of these new components under a variety of attack 1793 situations on the power grid. Consequences of a cyber attack on the 1794 grid telecommunications network can be catastrophic. This is why 1795 security for smart grid is not just an ad hoc feature or product, 1796 it's a complete framework integrating both physical and Cyber 1797 security requirements and covering the entire smart grid networks 1798 from generation to distribution. Security has therefore become one 1799 of the main foundations of the utility telecom network architecture 1800 and must be considered at every layer with a defense-in-depth 1801 approach. Migrating to IP based protocols is key to address these 1802 challenges for two reasons: 1804 o IP enables a rich set of features and capabilities to enhance the 1805 security posture 1807 o IP is based on open standards, which allows interoperability 1808 between different vendors and products, driving down the costs 1809 associated with implementing security solutions in OT networks. 1811 Securing OT (Operation technology) telecommunications over packet- 1812 switched IP networks follow the same principles that are foundational 1813 for securing the IT infrastructure, i.e., consideration must be given 1814 to enforcing electronic access control for both person-to-machine and 1815 machine-to-machine communications, and providing the appropriate 1816 levels of data privacy, device and platform integrity, and threat 1817 detection and mitigation. 1819 Existing power automation security standards can inform network 1820 security. For example the NERC CIP (North American Electric 1821 Reliability Corporation Critical Infrastructure Protection) plan is a 1822 set of requirements designed to secure the assets required for 1823 operating North America's bulk electric system. Another standardized 1824 security control technique is Segmentation (zones and conduits 1825 including access control). 1827 The requirements in Industrial Automation and Control Systems (IACS) 1828 are quite similar, especially in new scenarios such as Industry 4.0/ 1829 Digital Factory where workflows and protocols cross zones, segments, 1830 and entities. IEC 62443 (ISA99) defines security for IACS, typically 1831 for installations in other critical infrastructure such as oil and 1832 gas. 1834 Availability and integrity are the most important security objectives 1835 for the lower layers of such networks; confidentiality and privacy 1836 are relevant if customer or market data is involved, typically 1837 handled by higher layers. 1839 8.4.3. (BAS) Security Considerations (sec 4.2.4) 1841 When BAS field networks were developed it was assumed that the field 1842 networks would always be physically isolated from external networks 1843 and therefore security was not a concern. In today's world many BASs 1844 are managed remotely and are thus connected to shared IP networks and 1845 so security is definitely a concern, yet security features are not 1846 available in the majority of BAS field network deployments . 1848 The management network, being an IP-based network, has the protocols 1849 available to enable network security, but in practice many BAS 1850 systems do not implement even the available security features such as 1851 device authentication or encryption for data in transit. 1853 8.4.4. (6TiSCH) Security Considerations (sec 5.3.3) 1855 On top of the classical requirements for protection of control 1856 signaling, it must be noted that 6TiSCH networks operate on limited 1857 resources that can be depleted rapidly in a DoS attack on the system, 1858 for instance by placing a rogue device in the network, or by 1859 obtaining management control and setting up unexpected additional 1860 paths. 1862 8.4.5. (Cellular radio) Security Considerations (sec 6.1.5) 1864 Establishing time-sensitive streams in the network entails reserving 1865 networking resources for long periods of time. It is important that 1866 these reservation requests be authenticated to prevent malicious 1867 reservation attempts from hostile nodes (or accidental 1868 misconfiguration). This is particularly important in the case where 1869 the reservation requests span administrative domains. Furthermore, 1870 the reservation information itself should be digitally signed to 1871 reduce the risk of a legitimate node pushing a stale or hostile 1872 configuration into another networking node. 1874 Note: This is considered important for the security policy of the 1875 network, but does not affect the core DetNet architecture and design. 1877 8.4.6. (Industrial M2M) Communication Today (sec 7.2) 1879 Industrial network scenarios require advanced security solutions. 1880 Many of the current industrial production networks are physically 1881 separated. Preventing critical flows from be leaked outside a domain 1882 is handled today by filtering policies that are typically enforced in 1883 firewalls. 1885 9. IANA Considerations 1887 This memo includes no requests from IANA. 1889 10. Security Considerations 1891 The security considerations of DetNet networks are presented 1892 throughout this document. 1894 11. Informative References 1896 [ARINC664P7] 1897 ARINC, "ARINC 664 Aircraft Data Network, Part 7, Avionics 1898 Full-Duplex Switched Ethernet Network", 2009. 1900 [I-D.ietf-detnet-architecture] 1901 Finn, N., Thubert, P., Varga, B., and J. Farkas, 1902 "Deterministic Networking Architecture", draft-ietf- 1903 detnet-architecture-13 (work in progress), May 2019. 1905 [I-D.ietf-detnet-data-plane-framework] 1906 Varga, B., Farkas, J., Berger, L., Fedyk, D., Malis, A., 1907 Bryant, S., and J. Korhonen, "DetNet Data Plane 1908 Framework", draft-ietf-detnet-data-plane-framework-01 1909 (work in progress), July 2019. 1911 [I-D.ietf-detnet-ip] 1912 Varga, B., Farkas, J., Berger, L., Fedyk, D., Malis, A., 1913 Bryant, S., and J. Korhonen, "DetNet Data Plane: IP", 1914 draft-ietf-detnet-ip-01 (work in progress), July 2019. 1916 [I-D.ietf-detnet-ip-over-tsn] 1917 Varga, B., Farkas, J., Malis, A., Bryant, S., and J. 1918 Korhonen, "DetNet Data Plane: IP over IEEE 802.1 Time 1919 Sensitive Networking (TSN)", draft-ietf-detnet-ip-over- 1920 tsn-00 (work in progress), May 2019. 1922 [I-D.ietf-detnet-mpls] 1923 Varga, B., Farkas, J., Berger, L., Fedyk, D., Malis, A., 1924 Bryant, S., and J. Korhonen, "DetNet Data Plane: MPLS", 1925 draft-ietf-detnet-mpls-01 (work in progress), July 2019. 1927 [I-D.varga-detnet-service-model] 1928 Varga, B. and J. Farkas, "DetNet Service Model", draft- 1929 varga-detnet-service-model-02 (work in progress), May 1930 2017. 1932 [IEEE1588] 1933 IEEE, "IEEE 1588 Standard for a Precision Clock 1934 Synchronization Protocol for Networked Measurement and 1935 Control Systems Version 2", 2008. 1937 [IEEE802.1AE-2018] 1938 IEEE Standards Association, "IEEE Std 802.1AE-2018 MAC 1939 Security (MACsec)", 2018, 1940 . 1942 [MIRAI] krebsonsecurity.com, "https://krebsonsecurity.com/2016/10/ 1943 hacked-cameras-dvrs-powered-todays-massive-internet- 1944 outage/", 2016. 1946 [RFC3552] Rescorla, E. and B. Korver, "Guidelines for Writing RFC 1947 Text on Security Considerations", BCP 72, RFC 3552, 1948 DOI 10.17487/RFC3552, July 2003, 1949 . 1951 [RFC4301] Kent, S. and K. Seo, "Security Architecture for the 1952 Internet Protocol", RFC 4301, DOI 10.17487/RFC4301, 1953 December 2005, . 1955 [RFC7384] Mizrahi, T., "Security Requirements of Time Protocols in 1956 Packet Switched Networks", RFC 7384, DOI 10.17487/RFC7384, 1957 October 2014, . 1959 [RFC8578] Grossman, E., Ed., "Deterministic Networking Use Cases", 1960 RFC 8578, DOI 10.17487/RFC8578, May 2019, 1961 . 1963 Authors' Addresses 1965 Tal Mizrahi 1966 Huawei Network.IO Innovation Lab 1968 Email: tal.mizrahi.phd@gmail.com 1970 Ethan Grossman (editor) 1971 Dolby Laboratories, Inc. 1972 1275 Market Street 1973 San Francisco, CA 94103 1974 USA 1976 Phone: +1 415 645 4726 1977 Email: ethan.grossman@dolby.com 1978 URI: http://www.dolby.com 1980 Andrew J. Hacker 1981 MistIQ Technologies, Inc 1982 Harrisburg, PA 1983 USA 1985 Email: ajhacker@mistiqtech.com 1986 URI: http://www.mistiqtech.com 1988 Subir Das 1989 Applied Communication Sciences 1990 150 Mount Airy Road, Basking Ridge 1991 New Jersey, 07920 1992 USA 1994 Email: sdas@appcomsci.com 1996 John Dowdell 1997 Airbus Defence and Space 1998 Celtic Springs 1999 Newport NP10 8FZ 2000 United Kingdom 2002 Email: john.dowdell.ietf@gmail.com 2003 Henrik Austad 2004 Cisco Systems 2005 Philip Pedersens vei 1 2006 Lysaker 1366 2007 Norway 2009 Email: henrik@austad.us 2011 Kevin Stanton 2012 Intel 2014 Email: kevin.b.stanton@intel.com 2016 Norman Finn 2017 Huawei 2019 Email: norman.finn@mail01.huawei.com