idnits 2.17.1 draft-bryant-arch-fwd-layer-uc-03.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 : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (January 24, 2022) is 823 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Outdated reference: A later version (-05) exists of draft-bryant-arch-fwd-layer-ps-03 == Outdated reference: A later version (-17) exists of draft-ietf-teas-enhanced-vpn-09 Summary: 0 errors (**), 0 flaws (~~), 3 warnings (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 INTAREA Working Group S. Bryant 3 Internet-Draft University of Surrey 5/6GIC 4 Intended status: Informational U. Chunduri 5 Expires: July 28, 2022 Intel 6 T. Eckert 7 A. Clemm 8 Futurewei Technologies Inc. 9 January 24, 2022 11 Forwarding Layer Use Cases 12 draft-bryant-arch-fwd-layer-uc-03 14 Abstract 16 This document considers the new and emerging use cases for IP. These 17 use cases are difficult to address with IP in its current format and 18 demonstrate the need to evolve the protocol. 20 Status of This Memo 22 This Internet-Draft is submitted in full conformance with the 23 provisions of BCP 78 and BCP 79. 25 Internet-Drafts are working documents of the Internet Engineering 26 Task Force (IETF). Note that other groups may also distribute 27 working documents as Internet-Drafts. The list of current Internet- 28 Drafts is at https://datatracker.ietf.org/drafts/current/. 30 Internet-Drafts are draft documents valid for a maximum of six months 31 and may be updated, replaced, or obsoleted by other documents at any 32 time. It is inappropriate to use Internet-Drafts as reference 33 material or to cite them other than as "work in progress." 35 This Internet-Draft will expire on July 28, 2022. 37 Copyright Notice 39 Copyright (c) 2022 IETF Trust and the persons identified as the 40 document authors. All rights reserved. 42 This document is subject to BCP 78 and the IETF Trust's Legal 43 Provisions Relating to IETF Documents 44 (https://trustee.ietf.org/license-info) in effect on the date of 45 publication of this document. Please review these documents 46 carefully, as they describe your rights and restrictions with respect 47 to this document. Code Components extracted from this document must 48 include Simplified BSD License text as described in Section 4.e of 49 the Trust Legal Provisions and are provided without warranty as 50 described in the Simplified BSD License. 52 Table of Contents 54 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 55 1.1. Forwarding Layer . . . . . . . . . . . . . . . . . . . . 3 56 2. New Use Cases for packet networks . . . . . . . . . . . . . . 4 57 2.1. Role of Fixed Networks in 5G and Beyond 5G . . . . . . . 4 58 2.2. Convergence of Industrial Control Networks . . . . . . . 5 59 2.3. Cloud Based Industrial Automation . . . . . . . . . . . . 5 60 2.4. Volumetric Data Transmission . . . . . . . . . . . . . . 6 61 2.5. ITU-T Focus Group Network-2030 . . . . . . . . . . . . . 6 62 2.6. Emerging and New Media Applications . . . . . . . . . . . 7 63 3. Deployment Models . . . . . . . . . . . . . . . . . . . . . . 8 64 3.1. Traditional Deployment Models . . . . . . . . . . . . . . 9 65 3.1.1. Best-effort Internet . . . . . . . . . . . . . . . . 9 66 3.1.2. Enhanced Service . . . . . . . . . . . . . . . . . . 10 67 3.1.3. Over-the-top (OTT) Providers . . . . . . . . . . . . 10 68 3.1.4. Cooperating Providers . . . . . . . . . . . . . . . . 11 69 3.2. Emerging Deployment Models . . . . . . . . . . . . . . . 11 70 3.2.1. Embedded Service . . . . . . . . . . . . . . . . . . 11 71 3.2.2. Embedded Global Service . . . . . . . . . . . . . . . 12 72 3.2.3. Changing Fixed Access Models (1 or 2 Providers) . . . 13 73 3.2.4. Single "Underlay" provider E2E for 5G/B5G network 74 (Cellular/Access Networks) . . . . . . . . . . . . . 13 75 3.3. Envisioned New Deployment Models . . . . . . . . . . . . 14 76 3.3.1. Network Slicing . . . . . . . . . . . . . . . . . . . 14 77 3.3.2. Private 5G Networks . . . . . . . . . . . . . . . . . 15 78 3.4. Limited Domains . . . . . . . . . . . . . . . . . . . . . 15 79 4. New Network Services and Capabilities . . . . . . . . . . . 16 80 4.1. New Services . . . . . . . . . . . . . . . . . . . . . . 16 81 4.2. New Capabilities . . . . . . . . . . . . . . . . . . . . 17 82 5. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 18 83 6. Security Considerations . . . . . . . . . . . . . . . . . . . 18 84 7. Appendix 1: Expanded Summary of Sub-G1 Use Cases . . . . . . 19 85 7.1. Holographic-type communications . . . . . . . . . . . . . 19 86 7.2. Tactile Internet for Remote Operations . . . . . . . . . 19 87 7.3. Space-Terrestrial Integrated Networks . . . . . . . . . . 20 88 7.4. ManyNets . . . . . . . . . . . . . . . . . . . . . . . . 20 89 8. Appendix 2: Expanded Summary of Sub-G2 New Network 90 Capabilities and Services . . . . . . . . . . . . . . . . . . 21 91 8.1. New Services . . . . . . . . . . . . . . . . . . . . . . 21 92 8.1.1. High-Precision Communications Services . . . . . . . 22 93 8.1.2. In-time Services . . . . . . . . . . . . . . . . . . 22 94 8.1.3. On-time Services . . . . . . . . . . . . . . . . . . 23 95 8.1.4. Coordinated Services . . . . . . . . . . . . . . . . 23 96 8.1.5. Qualitative Communication Services . . . . . . . . . 23 98 8.2. New Capabilities . . . . . . . . . . . . . . . . . . . . 23 99 8.2.1. Manage ability . . . . . . . . . . . . . . . . . . . 24 100 8.2.2. High Programmability and Agile Life-cycle . . . . . . 25 101 8.2.3. Security . . . . . . . . . . . . . . . . . . . . . . 26 102 8.2.4. Trustworthiness . . . . . . . . . . . . . . . . . . . 27 103 8.2.5. Resilience . . . . . . . . . . . . . . . . . . . . . 27 104 8.2.6. Privacy-Sensitive . . . . . . . . . . . . . . . . . . 27 105 8.2.7. Accountability and Verifiability . . . . . . . . . . 28 106 9. Informative References . . . . . . . . . . . . . . . . . . . 29 107 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 31 109 1. Introduction 111 There is an emerging set of new requirements that exceed the network 112 and transport services of the current Internet, which currently only 113 delivers "best effort" service. While many controlled or private 114 networks include further services, such as other DiffServ QoS in 115 addition to best effort and traffic engineering with bandwidth 116 guarantees, the solutions used today only support walled gardens and 117 are thus they are not available to application service providers and 118 consumers across the Internet. 120 The purpose of this document is to look at current, evolving and 121 future use cases that need to addressed by the Internet forwarding 122 layer. In parallel with this use case study, a study of the gaps 123 between the capability of the existing IP forwarding layer and the 124 requirements described in this use case study is provided in 125 [I-D.bryant-arch-fwd-layer-ps]. It is thus the purpose of this text 126 to provide the wider context for the forwarding layer problem 127 statement. 129 The purpose of this text is thus to stimulate discussion on the 130 emerging contexts in which the forwarding layer will need to operate 131 in the future. 133 1.1. Forwarding Layer 135 The term "forwarding layer" is used in this document to indicate that 136 that development work will likely need to reach down to layer 2.5 in 137 order to ensure that packets are handled correctly down to the 138 physical layer, and that it is equally it is possible that 139 development work will need to reach into the transport layer. This 140 is described in more detail in [I-D.bryant-arch-fwd-layer-ps]. 142 2. New Use Cases for packet networks 144 This section summarizes the use case areas that have been observed by 145 the authors, and are considered relevant to any analysis of the gaps 146 in forwarding layer capabilities. 148 This section is structured into sub-sections discussing either group 149 of use cases directly or the work of specific groups that are 150 identifying use cases and that may also work on identifying issues 151 and or proposed architectures or solutions for them. 153 Subsections are ordered from what might be considered to be the most 154 near-term use cases to the potentially most far reaching ones. 156 2.1. Role of Fixed Networks in 5G and Beyond 5G 158 The 5G and beyond 5G (B5G) services are not meant to be limited to 159 the 5G-NR (new-radio). In fact for those services relating to uRLLC, 160 and mMTC packet networks have evolve along with the radio 161 technologies. While 5G-NR protocol stack has evolved to provide per- 162 frame reliability and latency guarantees, the IP/MPLS transport 163 network by-and-large remains best-effort. It is no longer possible 164 to solve network problems simply by increasing the capacity 165 [SysArch5G]. The expectations 5G devices have of 5G networks, can 166 not be met without improving IP/MPLS based back-haul networks. For 167 example, the 5G based systems involve machine to machine 168 communications, generally using command-based smaller payloads. In 169 this case the overheads of packet headers and overlays become 170 apparent when computing latency budget of such packets. 172 The IETF has produced a large body of work on the deterministic needs 173 of network applications [RFC8578]. These range from refinements and 174 expansions of above summarized Audio/Video and AR/VR use cases over 175 gaming into many more "industrial" use cases. Industrial use cases 176 generally involve industrial controllers for high-precision machinery 177 and equipment, such as robotic arms, centrifuges, or manufacturing 178 equipment for the assembly of electronic components. 179 These use cases have in common that they require delivery of packets 180 with very precise and "deterministic" performance characteristics, as 181 the controlled equipment and the control loops involved have very 182 exact timing requirements and are not tolerant of any latency 183 variations, as otherwise control loop issues and other undesired 184 effects may occur. 185 Specifically, the use cases involve curtailing maximum latency that 186 could be incurred. However, deterministic networking, by itself, 187 does not appear to be sufficient to meet all of the emerging needs. 189 2.2. Convergence of Industrial Control Networks 191 Industrial control networks exist to serve specialist applications 192 and are deployed in well controlled networks subject to tight timing 193 and reliability constrains and tight security constraints. They 194 mostly use bespoke, application specific proprietary protocols. 195 There is a desire to achieve economy of scale by using a single 196 protocol, and to integrate the production network with the back- 197 office network. The obvious protocol to use would be IP, but to be 198 deployed in this mixed application environment IP needs to satisfy 199 the non-negotiable needs of the industrial control network such as 200 timing, reliability and security. 202 2.3. Cloud Based Industrial Automation 204 Future industrial networks are significantly different from best 205 effort networks in terms of performance and reliability requirements. 206 This is discussed in [NET2030SubG1]. These networks need more than 207 basic connectivity between the back office and the factory floors, 208 instead they require integration from devices all the way through to 209 the business systems. This permits many new types of UI and full 210 automatic operation and control of industrial processes without 211 significant human intervention. These networks need to deliver 212 better than best effort performance, and require real-time, secure, 213 and reliable factory-wide connectivity, as well as inter-factory 214 connectivity at large scale. 216 Such systems typically require low end-to-end latency to meet closed 217 loop control requirements. Such system also need low jitter 218 connectivity. IIoT systems, as an example contain many control sub- 219 systems that run at cycle times ranging from sub-ms to 10 ms. In 220 such systems, end-to-end control requires in-time signaling delay at 221 the same cycle time level, without malfunctions. These low latency 222 requirements of IIoT applications are increasingly not only relevant 223 to internal system communications, but also becoming essential for 224 the interconnection of remote systems. 226 As another example, it is a fundamental requirement for multiple-axis 227 applications to have time synchronization in order to permit 228 cooperation between various devices, sometimes remotely. In order to 229 recover the clock signal and reach precise time synchronization, the 230 machine control, especially the motion control sub-system, requires 231 very small jitter at sub- microsecond level, and such small jitter is 232 expected to have bounded limits under some critical situations. 234 In some IIOT systems a service availability of 99.999999% is needed, 235 as any break in communications may be reflected as a million-dollar 236 loss. At the same time, as part of the Industry 4.0 evolution, 237 operational technologies (OT) and information technology (IT) are 238 converging. In this model control functions traditionally carried 239 out by customized hardware platforms, such as Programmable Logic 240 Controllers (PLC), have been slowly virtualized and moved onto the 241 edge or into the cloud in order to reduce the CAPEX and OPEX, and to 242 provide increased system flexibility and capability and to allow 'big 243 data' approaches. This move of industrial system to the cloud places 244 higher requirements on the underlying networks, as the latency, 245 jitter, security and reliability requirements previously needed 246 locally have to be implemented at larger scales. 248 2.4. Volumetric Data Transmission 250 Volumetric Data refers to cases where very large data sets need to be 251 transferred continuously in real time. One example is Immersive AR/ 252 VR media transmission Section 2.6. Another example is V2X with many 253 sensors continuously generating data which needs to made available 254 for, amongst other reasons, technical analysis by the manufacturer as 255 part of product development, and insurance purposes. 257 2.5. ITU-T Focus Group Network-2030 259 The ITU-T has been running a Focus Group (FG) Network-2030 260 [FGNETWORK2030] to analyze the needs of networks in the period post 261 2030. This work started in July 2018 and submitted it report to 262 ITU-T Study Group 13 in June 2020. It has been an open process with 263 contribution by a cross-section of the networking industry. Because 264 this is non-IETF work, this section summarizes the currently 265 finalized key findings of the ITU-T Focus Group Network-2030 to make 266 it easier for the reader to better understand the work. Note that 267 this work is still ongoing and additional findings may be published. 269 The Focus Group Network 2030 considered a number of use cases that it 270 was postulated would need to be addressed in the 2030 time-frame and 271 the technology gaps that need to be bridged in order to address these 272 needs. It then considered a number of new network services that 273 would be needed to support these services. 275 An ongoing piece of work on the architecture of the network post 2030 276 has not yet been completed at the time of writing and is only 277 partially discussed in this document. 279 The reader is referred to [WP], [NET2030SubG2], [UC] for information 280 beyond that provided in this summary. 282 ITU-T FG NET2030 Sub-group Sub-G1 (Sub-G1) considered a number of use 283 cases that it considered to be representative of the network needs 284 post 2030. These needs are legitimate needs in their own right, but 285 as is always the case act as poster-children for new applications 286 that will inevitable conceived in the light of the new network 287 capabilities that we postulate to be necessary. 289 o Holographic-type communications (HCT) 291 o Tactile Internet for Remote Operations (TIRO) 293 o Network and Computing Convergence (NCC) 295 o Digital Twin (DT) 297 o Space-Terrestrial Integrated Networks (STIN) 299 o ManyNets 301 o Industrial IoT (IIoT) with cloudification. 303 Further information on these use cases is provided in Section 7, and 304 in the ITU documents [UC] and [WP]. 306 Note to the reader: Unlike ITU-T Study Groups which are restricted to 307 members, ITU-T Focus Groups are open to anyone without payment. At 308 the time of writing, ITU-T Focus Group Network-2030 material that is 309 not available for anonymous download, is accessible for free by 310 joining the Study Group. 312 2.6. Emerging and New Media Applications 314 Audio/Video streaming for production, entertainment, remote 315 observation, and interactive audio/video are the most ubiquitous 316 applications on the Internet and private IP networks after web- 317 services. They have grown primarily through an evolution of the 318 applications to work with the constraints of todays Internet and 319 adopting pre-existing infrastructure such as content caches: best- 320 effort streaming with adaptive video, no service guarantees for most 321 services, and co-location of caches with large user communities. In 322 environments where more than best-effort services for these 323 applications are required and deployment of current technologies to 324 support them is feasible, it is done. Examples include DiffServ or 325 even on or off-path bandwidth reservations in controlled networks. 327 Networked AR/VR is a very near term set of use cases, where solution 328 models are very much attempting to use and expand existing solution 329 approaches for video network streaming but where the limits of above 330 current best practices are also amplified by the larger bandwidth 331 requirements and stricter latency and jitter requirements of AR/VR. 333 To ensure a good user experience, for live Virtual Reality (VR), a 334 much higher resolution than 8K video is required. In addition to the 335 high bandwidth requirements of VR, there needs to be a supporting 336 transmission network to provide a communications path with bounded 337 low latency as well. This stringent VR latency requirement is a 338 challenge to existing networks. 340 In cellular networks, even though the the air interface link latency 341 needed is significantly reduced e.g. with New Radio (5GNR), the end- 342 to-end (E2E) requirements for live VR is harder to meet. This is 343 because of the fixed L2/IP/MPLS networks in front/mid/backhaul 344 components, and because of the best effort nature of the packet 345 delivery systems in these networks. 347 3. Deployment Models 349 In this section we look at a number of network deployment models. We 350 group these deployment models into three types: 352 o The traditional deployment models 354 o Emerging deployment model models 356 o Envisioned new deployment models 358 The service requirements demanded from the networks and security 359 implications vastly differ in these different deployment models. 361 A few general observations are useful in providing context to this 362 section: 364 o End to End traffic over the Internet backbone is becoming minority 365 traffic. 367 o Commercial deployments do not operate the way they used to when 368 many of the original Internet protocols and invariants were 369 established. 371 o The application trajectory is for the applications to be hosted on 372 (protected) servers a few hops from the user. 374 o Applications are becoming self-contained and use their own stack 375 which is tunneled over UDP/IP to the server. 377 3.1. Traditional Deployment Models 379 In this section we look at the traditional deployment models that 380 have been in existence for many year and formed the foundation of 381 Internet. 383 3.1.1. Best-effort Internet 385 In this model connectivity is edge-to-edge, and in the general case 386 the edge connectivity is provided by a service provider who peers 387 with a transit provider that provides connectivity to other service 388 providers possibly via other transit providers. This is shown in 389 Figure 1. 391 +---+ +---+ 392 | H | |Svr| 393 +-+-+ +-+-+ 394 | SP1 Internet SP2 | 395 | .......... ..................... ......... | 396 +-+--+ .+----+ . .+---+ +---+ +---+' . +----+. +-+-+ 397 | CE +--+ PE +------+AS1+--+AS2+--+AS3+------+ PE +---+ CE| 398 +----+ .+----+ . .+---+ +---+ +---+. . +----+. +---+ 399 .......... ..................... ......... 401 Figure 1: An Edge-2-Edge Classical Internet 403 This service is generally known as "best-effort" in that each element 404 of the service path undertakes to do no more that try its best to 405 provide equitable service to all traffic. These are traditional E2E 406 deployments where communication endpoints of the data traffic on 407 different provider networks with regional, transit network providers 408 through Internet Exchange Providers (IXPs) providing the global inter 409 connection. The term lower-common-denominator might be a better term 410 in that the service quality is the service of the worst element of 411 the path on a packet by packet basis. 413 This model is in the process of being replaced by a model in which 414 the most popular and important service are provided at the edge with 415 Internet transit traffic being used where there is no alternative. 417 In this case the provider controls only the path to the CE and can 418 certify the correct operation of the service according to contract 419 from that point but the user is responsible for providing the 420 required service characteristics into their own network. 422 In this network environment it is difficult to support any form of 423 enhanced service since it is unlikely that the whole path is know to 424 support extended capabilities in the forwarding plane. It is not 425 infeasible, and it would be possible to set up such paths in 426 principle given suitable enhancements to the routing system. However 427 such a scenario must be considered infeasible for the foreseeable 428 future. 430 3.1.2. Enhanced Service 432 This is the traditional service provider deployment where various 433 network services (VPN, security, Bandwidth..) are offered to the 434 endpoints of the communication and other providers. Such 435 capabilities are purchased through contract with the service provider 436 and are typically expensive. 438 These networks predominantly use MPLS technology though native IP 439 (IPv4/IPv6) with GRE and IPv6 with routing extension headers with 440 SRv6 are being deployed recently. 442 .................................. 443 +---+ . +---+ Single +---+ . +---+ 444 |CE1|---|PE1|---.. Provider ..---|PE2|---|CE2| 445 +---+ . +---+ Network +---+ . +---+ 446 .................................. 448 Figure 2: An Edge-2-Edge Network 450 In this case there is a single provider network in which E2E 451 offerings and host session are initiated and terminated with in the 452 single provider network. 454 3.1.3. Over-the-top (OTT) Providers 456 In this model the endpoints of the communication (virtual or physical 457 hosts) consuming services through with in the OTT provider network 458 servers (Cloud and Data Center (DC) networks); where the other 459 endpoint can be in the same server form or on the DC Gateway or on 460 the other end of the DC Server Farm connected through Data Center 461 Interconnect (DCI). 463 The local provider is thus just a connectivity provider to opaque 464 traffic with no ability to enhance the service. However the 465 corollary to this is that whilst the the OTT provider has full 466 control of what happens whilst the user data is within their network 467 they have no control over how the user traffic transits to them 468 across the "public" network. 470 3.1.4. Cooperating Providers 472 Where two providers interconnect with no Internet Transit Network: 473 Another variant of the E2E connectivity can be seen as evolving 474 comprises only endpoints provider (access) network and receiver 475 access provider network with global transit provided by one ISP. 476 This case is more tractable provided there is co-operation between 477 the providers. 479 3.2. Emerging Deployment Models 481 The emerging model is to provide the service close to the user by 482 embedding that service with the service provider network. This has 483 three advantages, firstly that the service latency is lower, secondly 484 that that there is less transit traffic that the network provider 485 needs to manage or pay for, and thirdly that the service availability 486 and reliability is in the hands of the network provider that the 487 customer is contracted to. 489 3.2.1. Embedded Service 491 The industry move is towards content and application service 492 providers embedding themselves within the edge network. This is 493 currently done to save bandwidth and improve response time. As the 494 need for high precision low latency networking develops the need for 495 edge computing rises since the closer the client and the server the 496 less the scope for network induced performance degradation. 498 +---+ 499 | H | 500 +-+-+ 501 | 502 | ..................................... 503 +-+--+ .+----+ +---+ . 504 | CE +--+ PE |--------+Svr| . 505 +----+ .+----+ +---+ Provider 1 . 506 ..................................... 508 Figure 3: An Edge-2-Provider 510 In this network the server S (owned by the content and applications 511 provider) has a contractual relationship with provider 1 and is thus 512 able to negotiate the network characteristics needed to meet its 513 service requirement. This model in which the server brokers the user 514 to network interface (UNI) requirements removes many of the 515 objections to the classical UNI model in which the client requests 516 the service requirements. In this model the host authenticates 517 itself with the server, having formed a previous business 518 relationship (for example by purchasing a holographic conferencing 519 service). The server has a relationship with Provider1, and thus is 520 a trusted party able to request that the service be set up between 521 itself and and its client, paying as necessary. As this is a 522 requested paid service traversing a limited distance over a defined 523 network, a bespoke packet protocol can, if necessary, be used with in 524 a contained and constrained way. 526 How the server communicates with any other part of the application 527 domain is out of scope for this document and possibly out of scope 528 for Provider 1. 530 This takes us to consider the embedded global service described in 531 {#EGS}. 533 3.2.2. Embedded Global Service 535 +---+ 536 | H1| 537 +-+-+ 538 | 539 | ...................................... 540 +-+--+ . +----+ +---+ . 541 | CE +---+ PE |--------+ S1| . 542 +----+ . +----+ +-+-+ Provider 1 . 543 ..................|................... 544 | 545 |Private Peering 546 | 547 ..................|................... 548 +----+ . +----+ +-+-+ . 549 | CE +---+ PE |--------+ S2| . 550 +----+ . +----+ +---+ Provider 2 . 551 | ...................................... 552 | 553 +-+-+ 554 | H | 555 +-+-+ 557 Figure 4: Edge-2-Edge via Provider 559 In this network model, the server S1 (owned by the content and 560 applications provider) has a contractual relationship with provider 1 561 and is thus able to negotiate the network characteristics needed to 562 meet its service requirement. It is servicing the needs of host H1. 564 Similarly that same provider has a contractual relationship with 565 provider 2 where it is servicing the needs of host H2. 567 By a method outside the scope of this document and outside the scope 568 of the global Internet the contents and applications provider has a 569 private path between S1 and S2. 571 This scenario shown in Figure 4 is important because it removes the 572 overwhelming issues associated with providing enhanced service across 573 the global Internet. Furthermore it describes a model where there is 574 commercial incentive, at scale, for the edge providers (Provider 1 575 and 2 above) to invest in providing and enhanced access service. 577 3.2.3. Changing Fixed Access Models (1 or 2 Providers) 579 The preceding sections are the basis for a change in the network 580 fixed access model. 582 The access network either connects to a data center gateway or one is 583 embedded in the access network. This gateway either passes the 584 traffic to a locally connected data center that provides the required 585 service or passes it over a private global data center interconnect 586 to a partner data center for service provision. Such a connection 587 provides service model in which the required service level cane be 588 more readily addressed. 590 H H 591 | | 592 | | 593 Access NW 594 \ 595 \ 596 DC-GW==Private Global==DC-GW 597 // DCI \\ 598 DC Fabric DC Fabric 599 | | | | | | | | | | 600 S S S S S S S S S S 602 Figure 5: Changing Fixed Access Model 604 3.2.4. Single "Underlay" provider E2E for 5G/B5G network (Cellular/ 605 Access Networks) 607 The preceding sections are the basis for the emerging change in the 608 structure of the 5B and Beyond 5G (B5G) network design. 610 Endpoints (UE's) connecting to the provider wireless or wired 611 networks, where service is terminated inside the provider network end 612 points. Based on the service offerings connection termination can 613 happen close to the Radio/access nodes with multi-access edge 614 computing (MEC) clouds or in the provider core network (core-cloud) 615 before going to the Internet eventually. Example of these 616 deployments include BNG, 4G and 5G wireless access/RAN/backhaul 617 networks. 619 Thus in Figure 6 user equipment connects to the customer site 620 provider edge via the radio network. This in turn is connected to 621 the aggregation PE which in turn determines if the traffic should be 622 routed to a local data center for processing, or passed to a core 623 data center. At the core DC the traffic may be processed locally, 624 passed out to the Internet, passed to a peer DC via a private DCI, or 625 processed locally with the help of resources access via that external 626 interconnects. 628 User Equipment(UE) 629 Phone/eMBB 630 / Compute Compute 631 \ Vehicle Storage Storage 632 / / | | / Internet 633 \ \ Drone/UAV | | / 634 / / / DC Fabric DC Fabric{ 635 \ \ \ IIOT | | \ 636 / / / / | | \ Private Global 637 \ \ \ \ | | DCI 638 Radio --------CS PE----Aggr PE-----Core PE 640 Figure 6: 4G and 5G underlay provider network 642 3.3. Envisioned New Deployment Models 644 The emerging network deployment models are a potential vector for 645 fundamental change in the way the network operates. 647 3.3.1. Network Slicing 649 Network slicing is a method of creating a private subset of a public 650 network. Unlike VPNs it is not a simple over the top approach, 651 instead it is more integrated with the base network in terms of the 652 way the base network provides services and allocates resources. A 653 network slice provides significant isolation between one slice and 654 another and between the slice and best effort users of the network. 655 In an ideal slice, the users of one slice have no way of knowing 656 anything about the traffic in any other slice. Such a service could 657 be offered through statistical multiplexing techniques with real 658 bandwidth permanently allocated to each slice, but this would not 659 easily offer the statistical multiplexing that make packet networking 660 so economic and so flexible. In particular it would not be easy to 661 transparently "borrow" unused committed bandwidth in a way that was 662 undetectable. It seems likely that to create a high fidelity slice 663 will require new properties in the packet layer, either through 664 extension of the existing packet protocols, or through the 665 introduction of an alternative design. A useful discussion of 666 network slicing relevant to this context can be found in 667 [I-D.ietf-teas-enhanced-vpn]. 669 Largely popularized as part of 5G the concept of network slicing has 670 wider applicability. 672 3.3.2. Private 5G Networks 674 A use case is emerging for 5G technology in private networks. The 675 interest is in the protection and security that comes with the use of 676 licensed spectrum. Unlicensed spectrum offers no protection against 677 other users of that spectrum and thus another aspect of best effort 678 comes into play, not only is the network best effort with respect to 679 traffic within the network (an addressable problem) but the radio is 680 best effort with respect to radio traffic from adjacent networks. 681 Without extensive radio shielding of the facility a user cannot know 682 if the spectrum is available for their use at any time, and they have 683 to suffer interference from adjacent users, who may be benignly using 684 the spectrum for legitimate purposes, as is their equal right, or may 685 be using it to cause service disruption to a commercial enterprise. 687 5G runs on licensed and hence protected spectrum. In return for the 688 paying the license fee the spectrum owner has a statutory protection 689 against interference. 691 Thus it is interesting to note that a major UK car plant just 692 announced the use of 5G to provide connectivity for equipment at 693 their manufacturing facility. 695 Such applications of 5G are not as architecturally constrained as 696 public 5G deployments and thus have the ability to make different 697 fundamental choices regarding their packet protocols. 699 3.4. Limited Domains 701 [RFC8799] provides a useful insight into the emergence of limited 702 domains in which fewer (or different) constraints on protocol design 703 and operation apply. Limited domains offer an opportunity to deploy 704 specialist forwarding layer protocols, designed to meet specific 705 objectives, which are not readily addressed by general purpose 706 protocols such as IPv4|6 without the need to worry about inter- 707 working and inter-operation across the big I Internet. 709 Such domains can be considered sandboxes in which new proposals can 710 be deployed without the wider concerns of full-scale Internet 711 deployment. 713 4. New Network Services and Capabilities 715 In order to support the use cases presented in Section 2, a number of 716 new network services will be needed. Likewise, a number of 717 additional more general network capabilities will becoming 718 increasingly important. Neither services nor capabilities are 719 sufficiently supported to the degree that will be required by 720 Internet technology in use today. 722 This section describes these services and capabilities at a high 723 level. It builds on a corresponding analysis that was conducted at 724 ITU-T FG-NET2030; readers are referred Section 8 for further detail 725 and, of course, to output produced by that group [NET2030SubG2] for a 726 more complete explanation of their considerations. 728 4.1. New Services 730 [NET2030SubG2] identifies a number of network services that will be 731 needed to support many of the new use cases. These network services 732 are divided into two categories: 734 o Foundational Services (FS) require which dedicated support on some 735 or all network system nodes which are delivering the service 736 between two or more application system nodes. 738 o Compound Services (CS) are composed of one or more foundational 739 services, and are used to make network services easier to consume 740 by certain applications or categories of use cases. An example of 741 a CS would be a Tactile Internet Service which consisted of 742 tactile control channel and a haptic feedback channel. 744 The following are a set of Foundational Services : 746 o High-Precision Communications Services: services with precisely 747 defined service level objectives related to end-to-end latency. 748 Three high-precision communications services that have so far been 749 proposed: 751 * In-time Services: services that require end-to-end latency 752 within a quantifiable limit. This service is similar to the 753 service provided by DetNet [RFC8655] but with more demanding 754 applications which need to be satisfied over IP. 756 * On-time Services: services require end-to-end-latency to be of 757 an exact duration. 759 * Coordinated Services: Coordinated services require multiple 760 interdependent flows to be delivered with the same end-to-end 761 latency, regardless of any (potential additional) service level 762 objective. 764 o Qualitative Communication Services: services that are able to 765 suppress retransmission of less relevant portions of the payload 766 in order to meet requirements on latency by applications that are 767 tolerant to this. 769 These are described in more detain in Section 8.1. 771 4.2. New Capabilities 773 [NET2030SubG2] identifies also a number of network capabilities that 774 will become increasingly important going forward, in addition to the 775 support for any particular services. 776 A number of those need to be taken into consideration from the very 777 beginning when thinking about how future data-planes need to evolve. 778 These capabilities are described in more detail in Section 8.2. 780 o Manageability: Many of the services that need to be supported in 781 the future will require advances in measurements and telemetry 782 will be required in order to monitor and validate that promised 783 service levels are indeed being delivered. These will requires 784 advanced instrumentation that is ideally built. 786 o High Programmability and Agile Life-cycle: Methods to provide 787 operators need to be able to rapidly nd easily introduce new 788 network services and adapt to new contexts and application needs. 790 o Security and Trustworthiness: New mechanisms are needed to 791 authorize packets to enter the network from a host or from another 792 network, and for them to then receive the required premium service 793 that can operate. This must operate without impacting the latency 794 and MTU requirements. This security mechanism has to protect both 795 the network, the user data and the user privacy, but still expose 796 sufficient information to the network that the correct premium 797 service can be delivered. 799 o Resilience: Ultra-low-latency requirements and the huge increase 800 of bandwidth demands of new services such as holographic type 801 communication services make retransmission as a mechanism to 802 recover data that was lost in transit increasingly less feasible. 804 Therefore, network resilience and avoidance of loss becomes more 805 importance that it is for best effort networks. 807 o Privacy-Sensitive: There is a growing awareness of the lack of 808 privacy in the Internet and its implications. 809 New network services have to be sensitive to and comply with 810 heightened user privacy expectations. 811 At the same time, the need for privacy needs to be balanced with 812 legitimate needs of network providers to operate and maintain 813 their networks, which requires some visibility into what is 814 happening on the network and how it is being used. There are a 815 variety of privacy-related requirements that ensue, such as: 817 * Anonymization 819 * Opaque User data 821 * Secured Storage 823 * Flow anonymization 825 o Accountability and Verifiability: Provision of the methods to 826 account for an verify delivery of premium services. 828 5. IANA Considerations 830 This document does not request any allocations from IANA. 832 6. Security Considerations 834 Security is likely to be more significant with the applications being 835 considered in this work. With interest in tightly controlled access 836 and latency, and contractual terms of business it is going to be 837 necessary to have provable right of access to network resources. 838 However heavyweight security is a contra-requirement to the light- 839 weight process needed for power efficiency, fast forwarding and low 840 latency. Addressing this will require new insights into network 841 security. 843 Further information on the issue of providing security in latency 844 sensitive environments can be found in [RFC9055] which are a sub-set 845 of the considerations applicable to the new use cases considered in 846 this text. 848 7. Appendix 1: Expanded Summary of Sub-G1 Use Cases 850 7.1. Holographic-type communications 852 This work projects that we will move towards a holographic society 853 where users remotely interact with the physical world over the 854 network. In industry the digital twin model will enable the control 855 of real objects through digital replicas. Tele-presence will move to 856 a new level with multi-site collaborations becoming much closer to 857 physical meetings that can take place without the time and 858 environmental cost of physical travel. 3D medical scans will become 859 full 3D views rather than the body/ organ slices that too many of us 860 are regrettably familiar with. It is easy to imagine that this 861 technology will take message delivery to a completely new level. 863 Analysis of these concepts results in the conclusion that the 864 following key network requirements are necessary: 866 o Ultra-high bandwidth (BPS class) 868 o Ultra-low latency (sub-ms) 870 o Multi-stream synchronization 872 o Enhanced network security 874 o Enhanced network reliability 876 o Edge computation 878 7.2. Tactile Internet for Remote Operations 880 Two cases were proposed as examples of this class of application. 881 The first is remote industrial management which involves the real- 882 time monitoring and control of industrial infrastructure operations. 883 The second involves remote robotic surgery. Remote robotic surgery 884 within an operating suite complex is a standard practice today, 885 however there are cases where it would be desirable to extend the 886 range of this facility. 888 Analysis of these concepts results in the conclusion that the 889 following key network requirements are necessary: 891 o Ultra-high bandwidth (Tbps class) 893 o Ultra-low latency (sub-ms) 895 o Sensory input synchronization 896 o Enhanced network security 898 o Enhanced network reliability 900 o Differentiated prioritization levels 902 7.3. Space-Terrestrial Integrated Networks 904 The game-changer in the area of space-terrestrial networking is the 905 active deployment of huge clusters of cheap Low Earth Orbit (LEO) 906 satellite constellations. These LEOs have a number of properties 907 that make them attractive, but arguably the most important is that 908 they combine global coverage with low latency. Studies [Handley] 909 show that for distances over 3000Km latency via a LEO cluster is 910 lower than the latency of terrestrial networks. The up-link to a LEO 911 cluster has to constantly change the point of attachment to the 912 cluster as the satellites that form the cluster rapidly move across 913 the sky relative to both the ground and relative to the satellites in 914 other orbits. In this scenario a number of access and connection 915 models need to be considered. 917 Analysis of these concepts results in the conclusion that the 918 following key network requirements are necessary: 920 o A suitable addressing and routing mechanism to deal with a network 921 that is constantly in flux. 923 o Sufficient bandwidth capacity on the satellite side to support the 924 new application needs 926 o A suitable satellite admission system 928 o Edge computation and storage 930 7.4. ManyNets 932 There is evidence that there is a change in direction from the 933 Internet as a single hetrogenious network back to a true Internet, 934 that is an interconnection of a number of networks each optimized for 935 its local use but capable of inter-working. 937 For example, satellite and the terrestrial networks adopt different 938 protocol architecture, which causes the difficulty to internetwork 939 between them, yet the common goal is to provide access to the 940 Internet. Secondly, there will be a massive number of IoT-type 941 devices connecting to the networks but the current interconnection 942 schemes are too complex for these services. There are further trends 943 in 5G/B5G back-haul infrastructure, requiring diverse set of resource 944 guarantees in networks to support different industry verticals. The 945 collection of such special purpose networks, existing together and 946 requiring interconnection among themselves are called ManyNets. 948 Much closer the traditional Internet model is the move to edge 949 computing services in which the client traffic is terminated at a 950 compute node very close to access edge. [DOT] Any resultant 951 application traffic is a private matter between the application on 952 the edge server and the servers it communicates with in the 953 fulfillment of those needs. Furthermore the application on the 954 client may be using a tunnel to the edge compute server. In such a 955 network the protocol used inside the tunnel and the protocol used 956 between the servers executing the service is a private matter. 958 The ManyNets concept aims to support flexible methods to support the 959 communication among such heterogeneous devices and their networks. 961 8. Appendix 2: Expanded Summary of Sub-G2 New Network Capabilities and 962 Services 964 This appendix expands on the ITU-T Sub-G2 new network capabilities 965 and services introduced in Section 4 It builds upon the analysis that 966 was conducted at ITU-T FG-NET2030; readers are also referred to 967 output produced by that group [NET2030SubG2] for more detail. 969 8.1. New Services 971 [NET2030SubG2] identifies a number of network services that will be 972 needed to support many of the new use cases. These network services 973 are divided into two categories: 975 o Foundational Services (FS) require which dedicated support on some 976 or all network system nodes which are delivering the service 977 between two or more application system nodes. FS cannot be 978 decomposed into other services. For example, IP packet routing 979 and forwarding are is a (pre-existing) foundational network 980 services. 982 o Compound Services (CS) are composed of one or more foundational 983 services. CS are "convenience services" that make network 984 services easier to consume by certain applications or categories 985 of use cases, but do not by themselves introduce new network 986 services or requirements into network system nodes. One example 987 would be a Tactile Internet Service which consists of two 988 communications channels, one for tactile control and the other for 989 haptic feedback. 991 The following sections focus on Foundational Services only, as these 992 are the ones that provide the basic building blocks with which the 993 needs of all other services can be addressed, and which are the ones 994 that potentially introduce new foundational requirements on network 995 system nodes. 997 8.1.1. High-Precision Communications Services 999 High-Precision Communications Services refers to services that have 1000 precisely defined service level objectives related to end-to-end 1001 latency, in many cases coupled with stringent requirements regarding 1002 to packet loss and to bandwidth needs. These requirements are in 1003 stark contrast to the best effort nature with related to existing 1004 network services. 1005 Of course, existing services often go to great lengths in order to 1006 optimize service levels and minimize latency, and QoS techniques aim 1007 to mitigate adverse effects of e.g. congestion by applying various 1008 forms of prioritization and admission control. However, 1009 fundamentally all of these techniques still constitute patches that, 1010 while alleviating the symptoms of the underlying best-effort nature, 1011 do not address the underlying cause and fall short of providing 1012 service level guarantees that will not be just of a statistical 1013 nature but that will be met by design. 1015 The high-precision communications services that have been identified 1016 are described in the following three sub-sections. 1018 8.1.2. In-time Services 1020 In-time services require end-to-end latency within a quantifiable 1021 limit. They specific a service level objective that is not to be 1022 exceeded, such as a maximum acceptable latency (putting a hard 1023 boundary on the worst case). In-time services are required by 1024 applications and use cases that have clear bounds on acceptable 1025 latency, beyond which the Quality of Experience would deteriorate 1026 rapidly, rendering the application unusable. An example concerns use 1027 cases that involve providing tactile feedback to users. Creating an 1028 illusion of touch requires a control loop with a hard-bounded round- 1029 trip time that is determined by human / biological factors, beyond 1030 which the sense of touch is lost and with it the ability to e.g. 1031 operate a piece of machinery from remote. Because many such use 1032 cases are mission-critical (such as tele-driving or remote surgery), 1033 in addition any loss or need for retransmission is unacceptable. 1035 This service is similar to the service provided by DetNet [RFC8655] 1036 but with more demanding applications which need to be satisfied over 1037 IP. 1039 8.1.3. On-time Services 1041 On-time services require end-to-end-latency to be of an exact 1042 duration, with the possibility of a small quantifiable variance as 1043 can be specified either by an acceptable window around the targeted 1044 latency or by a lower bound in addition to an upper bound. Examples 1045 of use cases include applications that require synchronization 1046 between multiple flows that have the same in-time latency target, or 1047 applications requiring fairness between multiple participants 1048 regardless of path lengths, such as gaming or market exchanges when 1049 required by regulatory authorities. The concept of a lowest 1050 acceptable latency imposes new requirements on networks to 1051 potentially slow down packets by buffering or other means, which 1052 introduces challenges due to high data rates and the cost e.g. of 1053 associated memory. 1055 8.1.4. Coordinated Services 1057 Coordinated services require multiple interdependent flows to be 1058 delivered with the same end-to-end latency, regardless of any 1059 (potential additional) service level objective. Use cases and 1060 applications include applications that require synchronization 1061 between multiple flows, such as use cases involving data streams from 1062 multiple cameras and telemetry sources. In the special case where an 1063 on-time service is required, no additional service is needed (as 1064 synchronization occurs by virtue of the fact that each flow adheres 1065 to the same SLO), but coordination may also be required in cases 1066 where no specific end-to-end latency is required, as long as all 1067 flows are serviced with service levels that are identical. 1069 8.1.5. Qualitative Communication Services 1071 Qualitative communication services (QCS) are able to suppress 1072 retransmission of portions of the payload that are deemed less 1073 relevant when necessary in order to meet requirements on latency by 1074 applications that are tolerant of certain quality degradation. They 1075 may involve the application of network coding schemes. 1077 QCS is a new service type that is needed to support AR/VR, 1078 holographic-type communications Industrial Internet and services such 1079 as autonomous driving. This needs the support of a new network 1080 capability that is as yet to be developed. 1082 8.2. New Capabilities 1084 [NET2030SubG2] identifies also a number of network capabilities that 1085 will become increasingly important going forward, in addition to the 1086 support for any particular services. These were introduced in 1087 Section 4.2. A number of these capabilities need to be taken into 1088 consideration from the very beginning when thinking about how future 1089 data-planes need to evolve. 1090 While many of those capabilities are well known, the past has shown 1091 that retrofitting data-planes with such capabilities after the fact 1092 and in a way that is adequate to the problem at hand is very hard. 1094 8.2.1. Manage ability 1096 Many of the services that need to be supported in the future have in 1097 common that they place very high demands on latency and precision 1098 that need to be supported at very high scales, coupled with 1099 expectations of zero packet loss and much higher availability than 1100 today. 1102 In order to assure in-time and on-time services with high levels of 1103 accuracy, advances in measurements and telemetry will be required in 1104 order to monitor and validate that promised service levels are indeed 1105 being delivered. This requires advanced instrumentation that is 1106 ideally built-in all the way to the protocol level. 1108 For example, the ability to identify and automatically eliminate 1109 potential sources of service-level degradations and fluctuations will 1110 become of increasing importance. This requires the ability to 1111 generate corresponding telemetry data and the ability to observe the 1112 performance of packets as they traverse the network. Some of the 1113 challenges that need to be addressed include the very high volume of 1114 data that gets generated and needs to be assessed, and the effects of 1115 the collection itself on performance. In general, greater emphasis 1116 will need to be placed on the ability to monitor, observe, and 1117 validate packet performance and behavior than is the case today. For 1118 seamless support, these capabilities will be inherently integrated 1119 with the forwarding function itself, for example delivered together 1120 with the packets. Today's solution approach, IOAM, is a promising 1121 technology currently that points in the right direction, and that 1122 also highlights some of the challenges - from MTU considerations due 1123 to extending packet sizes to the ability to customize and obtain the 1124 "right" data. It will therefore be not sufficient by itself. Data 1125 to be generated from the network will need to be "smarter", i.e. more 1126 insightful and action-able. This will require additional abilities 1127 to process data "on-device". In additional, the need for new 1128 management functions may arise, such as functions that allow to 1129 validate adherence with agreed-upon service levels for a flow as a 1130 whole, and to prevent data or privacy leakage as well as provide 1131 evidence for the possibility or absence of such leakage. 1133 8.2.2. High Programmability and Agile Life-cycle 1135 Operators need to be able to rapidly introduce new network services 1136 and adapt to new contexts and application needs. This will require 1137 advances in network programmability. Today's model of vendor-defined 1138 (supporting service features via new firmware or hardware-based 1139 networking features) or operator-defined (supporting service features 1140 via programmable software-defined networking (SDN) controllers, 1141 virtualized network functions (VNF) and Network Function 1142 Virtualization (NFV), and service function chaining (SFC) will no 1143 longer be sufficient. 1145 Software Defined Networking and Network Function Virtualization (NFV) 1146 have opened up the possibility to accelerate development life-cycles 1147 and enable network providers to develop new networking features on 1148 their own if needed. Segment Routing is being evolved for that 1149 purpose as well. Furthermore, network slicing promises more agility 1150 in the introduction of new network services. However, the complexity 1151 of the associated controller software results in its own challenges 1152 with software development cycles that, while more agile than life- 1153 cycles before, are still prohibitive and that can only be undertaken 1154 by network providers, not by their customers. Rapid customization of 1155 networking services for specific needs or adaptation to unique 1156 deployments are out of reach for network provider customers. What is 1157 lacking is the ability for applications to rapidly introduce and 1158 customize novel behavior at the network flow level, without need to 1159 introduce application-level over-the-top (OTT) overlays. Such a 1160 capability would be analogous to server-less computing that is 1161 revolutionizing cloud services today. In addition, it should be 1162 noted that softwarized networks are built on relatively stable (and 1163 slowly evolving) underlying physical commodity hardware network 1164 infrastructure. This is insufficient to deliver on new high- 1165 precision network services, which require hardware advances at many 1166 levels to provide programmable flow and QoS behavior at line rate, 1167 affecting everything from queuing and scheduling to packet processing 1168 pipelines. 1170 The evolution of forwarding planes should allow development life- 1171 cycles that are much more agile than today and move from "Dev Ops" to 1172 "Flow Ops" (i.e. dynamic programmability of networks at the flow 1173 level). 1174 This requires support of novel network and data-plane programming 1175 models which can possibly be delivered and effected via the 1176 forwarding plane itself. 1178 8.2.3. Security 1180 The possibility of security threats increases with complexity of 1181 networks, the potential ramifications of attacks are growing more 1182 serious with increasing mission-criticality of networking services 1183 and applications. 1184 The forwarding plane plays a large role in the ability to thwart 1185 attacks. 1186 For example, the fact that source addresses are not authenticated in 1187 existing IP is at the root of a wide range security problems from 1188 phishing and fraudulent impersonation designed to compromise and 1189 steal user assets to amplification attacks designed to bring down 1190 services. 1191 Going forward, it is absolutely critical, then, to minimize the 1192 attack surface of the forwarding plane as it evolves. 1194 A key security aspects needed from the network point of view includes 1195 to verify if the packet is authorized to enter into the network and 1196 if it is sufficiently integrity protected. However, when packets are 1197 emitted from the host for these new communication services, the 1198 network portion of the packet (e.g., an extension header or an 1199 overlay header) should not be encrypted because network nodes may 1200 need to interpret the header and provide the desired service. 1201 Lack of encryption and integrity validation, of course, would at the 1202 same time increase the threat surface and open up the possibility for 1203 attacks. 1204 Mechanisms for authorization and integrity protection must be 1205 developed to meet the line rate performance as services delivered can 1206 be time sensitive. At the same time, the size of packets should not 1207 be significantly increased to avoid negative impact on utilization 1208 and overhead tax. 1209 This limits the options for additional security collateral that can 1210 be included with packets. 1212 Homomorphic forms of encryption may need to be devised in which 1213 network operations can be performed in privacy-preserving manner on 1214 encrypted packet headers and tunneled packets without exposing any of 1215 their contents. 1217 Another dimension to security arises when the end to end service that 1218 needs to be delivered crosses the administrative boundary of the 1219 originating host. For those cases, additional mechanisms need to be 1220 specified to sufficiently ensure the privacy and confidentiality of 1221 the network layer information. While there are lot of avenues to 1222 tackle these issues and some aspects are being looked into by various 1223 Standards Development Organizations, e.g. IRTF PANRG on Path-Aware 1224 Networking, comprehensive solutions are yet to be worked out. 1226 Any mechanisms specified for authorization, integrity protection, and 1227 network header confidentiality should be orthogonal to the transport 1228 layer and above transport layer security mechanisms set in place by 1229 the end host/user. Regardless of whether or not the latest security 1230 advances in transport and layers above (e.g. TLS1.3, QUIC or HTTPSx) 1231 are applied on the payload, network nodes should not have to act on 1232 this information to deliver new services to avoid layer violations. 1234 8.2.4. Trustworthiness 1236 As future network services are deployed, deployment scenarios will 1237 include cases in which packets need to traverse trust boundaries 1238 which are under different administrative domains. As the forwarding 1239 plane evolves, it should do so in such a way that trustworthiness of 1240 packets is maintained - i.e. integrity of data is protected, 1241 tampering with packet meta-data (such as source authentication or 1242 service level telemetry) would be evident, and privacy of users is 1243 guarded. 1245 8.2.5. Resilience 1247 Ultra-low-latency requirements and the huge increase of bandwidth 1248 demands of new services such as holographic type communication 1249 services make retransmission as a mechanism to recover data that was 1250 lost in transit increasingly less feasible. Therefore, network 1251 resilience and avoidance of loss becomes of paramount importance. 1253 There are many methods for providing network resilience. This 1254 includes providing redundancy and diversity of both physical (e.g. 1255 ports, routers, line cards) and logical (e.g. shapers, policers, 1256 classifiers) entities. It also includes the use of protocols that 1257 provide quick re-convergence and maintain high availability of 1258 existing connections after a failure event occurs in the network. 1259 Other techniques include packet replication or network coding and 1260 error correction techniques to overcome packet loss. 1261 As the forwarding plane evolves, mechanisms to provide network 1262 resilience should be inherently supported. 1264 8.2.6. Privacy-Sensitive 1266 Today, there is a growing awareness of the lack of privacy in the 1267 Internet and its implications. 1268 New network services have to be sensitive to and comply with 1269 heightened user privacy expectations. 1270 At the same time, the need for privacy needs to be balanced with 1271 legitimate needs of network providers to operate and maintain their 1272 networks, which requires some visibility into what is happening on 1273 the network and how it is being used. 1275 Likewise, mechanisms to provide privacy must be provided in such a 1276 way to not compromise security, such as allowing anonymous attackers 1277 to prey on other users. 1279 An evolved forwarding plane must provide mechanisms that ensure users 1280 privacy by design and prevent illegitimate exposing of personally- 1281 identifiable information (PII), while preventing abuse of those 1282 mechanisms by attack exploits and while affording network providers 1283 with legitimate visibility into use of their network and services. 1285 There are a variety of privacy-related requirements that ensue, such 1286 as: 1288 o Anonymization: To prevent tracking by eavesdropper by packet 1289 capture, visible information in packets such as source and 1290 destination addresses should be difficult (ideally: impossible) to 1291 directly correlate to PII. 1293 o Opaque User data: Networks must not rely on the user data to 1294 provide or improve the service. However, network providers may 1295 use specific service-visible data in packets. 1297 o Secured Storage: Some services may require the network to slow 1298 down the delivery of the packets, implying the possibility that 1299 packets are temporarily buffered on the router. The storage of 1300 those packets must be secured and prevented from extraction for 1301 deep inspection or analysis. 1303 o Flow anonymization: Flows of information should be randomized in a 1304 dynamic manner so that it is difficult through traffic analysis to 1305 deduce patterns and identify the type of traffic. 1307 Potential mechanisms to consider include (but are not limited to) 1308 avoiding the need for long-lived addresses (to prevent trackablity) 1309 and the use of homomorphic encryption for packet headers and tunneled 1310 packets (in addition to traditional payload encryption) that allow to 1311 perform network operations in privacy-preserving manner without 1312 exposing meta-data carried in headers. 1314 8.2.7. Accountability and Verifiability 1316 Many new services demand guarantees instead of being accepting of 1317 "best effort". 1318 As a result, today's "best effort" accounting may no longer be 1319 sufficient. 1321 Today's accounting technology largely relies on interface statistics 1322 and flow records. 1324 Those statistics and records may not be entirely accurate. 1325 For example, in many cases their generation involves sampling and is 1326 thus subject to sampling inaccuracies. 1327 In addition, this data largely accounts for volume but not so much 1328 for actual service levels (e.g. latencies, let alone coordination 1329 across flows) that are delivered. 1330 Service level measurements can be used to complement other statistics 1331 but come with significant overhead and also have various limitations, 1332 from sampling to the consumption of network and edge node processing 1333 bandwidth. 1334 Techniques that rely on passive measurements are infeasible in many 1335 network deployments and hampered by encryption as well as issues 1336 relating to privacy. 1338 Guarantees demand their price. This makes it increasingly important 1339 both for providers and users of services to be able to validate that 1340 promised service levels were delivered on. 1341 For example, proof of service delivery (including proof of service 1342 level delivery) may need to be provided to account and charge for 1343 network services. 1344 This will require advances in accounting technology that should be 1345 considered as forwarding technology evolves, possibly providing 1346 accounting as a function that is intrinsically coupled with 1347 forwarding itself. 1349 9. Informative References 1351 [DOT] Huston, G., "The Death of Transit and Beyond", n.d., 1352 . 1355 [FGNETWORK2030] 1356 "Focus Group on Technologies for Network 2030", n.d., 1357 . 1360 [Handley] Handley, M., "Delay is Not an Option: Low Latency Routing 1361 in Space", n.d., 1362 . 1364 [I-D.bryant-arch-fwd-layer-ps] 1365 Bryant, S., Chunduri, U., Eckert, T., and A. Clemm, 1366 "Forwarding Layer Problem Statement", draft-bryant-arch- 1367 fwd-layer-ps-03 (work in progress), June 2021. 1369 [I-D.ietf-teas-enhanced-vpn] 1370 Dong, J., Bryant, S., Li, Z., Miyasaka, T., and Y. Lee, "A 1371 Framework for Enhanced Virtual Private Network (VPN+) 1372 Services", draft-ietf-teas-enhanced-vpn-09 (work in 1373 progress), October 2021. 1375 [NET2030SubG1] 1376 ITU-T FGNet2030, "FG NET-2030 Sub-G1 Representative use 1377 cases and key network requirements for Network 2030", 1378 January 2021, 1379 . 1381 [NET2030SubG2] 1382 ITU-T FGNET2030, "New Services and Capabilities for 1383 Network 2030: Description, Technical Gap and Performance 1384 Target Analysis", October 2019, . 1388 [RFC8578] Grossman, E., Ed., "Deterministic Networking Use Cases", 1389 RFC 8578, DOI 10.17487/RFC8578, May 2019, 1390 . 1392 [RFC8655] Finn, N., Thubert, P., Varga, B., and J. Farkas, 1393 "Deterministic Networking Architecture", RFC 8655, 1394 DOI 10.17487/RFC8655, October 2019, 1395 . 1397 [RFC8799] Carpenter, B. and B. Liu, "Limited Domains and Internet 1398 Protocols", RFC 8799, DOI 10.17487/RFC8799, July 2020, 1399 . 1401 [RFC9055] Grossman, E., Ed., Mizrahi, T., and A. Hacker, 1402 "Deterministic Networking (DetNet) Security 1403 Considerations", RFC 9055, DOI 10.17487/RFC9055, June 1404 2021, . 1406 [SysArch5G] 1407 "System architecture for the 5G System (5GS)", n.d., 1408 . 1411 [UC] ITU-T FGNET2030, "Use Cases and Requirements for Network 1412 2030 Summary report "Representative use cases and key 1413 network requirements for Network 2030"", January 2020, 1414 . 1417 [WP] "Network 2030 - A Blueprint of Technology, Applications, 1418 and Market Drivers towards the Year 2030 and Beyond, a 1419 White Paper on Network 2030, ITU-T", May 2019, 1420 . 1423 Authors' Addresses 1425 Stewart Bryant 1426 University of Surrey 5/6GIC 1428 Email: sb@stewartbryant.com 1430 Uma Chunduri 1431 Intel 1433 Email: umac.ietf@gmail.com 1435 Toerless Eckert 1436 Futurewei Technologies Inc. 1438 Email: tte@cs.fau.de 1440 Alexander Clemm 1441 Futurewei Technologies Inc. 1443 Email: ludwig@clemm.org