idnits 2.17.1 draft-ietf-manet-smf-12.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 seems to contain a disclaimer for pre-RFC5378 work, and may have content which was first submitted before 10 November 2008. The disclaimer is necessary when there are original authors that you have been unable to contact, or if some do not wish to grant the BCP78 rights to the IETF Trust. If you are able to get all authors (current and original) to grant those rights, you can and should remove the disclaimer; otherwise, the disclaimer is needed and you can ignore this comment. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (July 11, 2011) is 4673 days in the past. Is this intentional? Checking references for intended status: Experimental ---------------------------------------------------------------------------- ** Obsolete normative reference: RFC 2460 (Obsoleted by RFC 8200) -- Obsolete informational reference (is this intentional?): RFC 4601 (Obsoleted by RFC 7761) Summary: 1 error (**), 0 flaws (~~), 1 warning (==), 3 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group J. Macker, editor 3 Internet-Draft NRL 4 Intended status: Experimental July 11, 2011 5 Expires: January 12, 2012 7 Simplified Multicast Forwarding 8 draft-ietf-manet-smf-12 10 Abstract 12 This document describes a Simplified Multicast Forwarding (SMF) 13 mechanism that provides basic IP multicast forwarding suitable for 14 limited wireless mesh and mobile ad hoc network (MANET) use. It is 15 mainly applicable in situations where efficient flooding represents 16 an acceptable engineering design trade-off. It defines techniques 17 for multicast duplicate packet detection (DPD), to be applied in the 18 forwarding process, for both IPv4 and IPv6 protocol use. This 19 document also specifies optional mechanisms for applying reduced 20 relay sets to achieve more efficient multicast data distribution 21 within a mesh topology, as compared to classic flooding. 22 Interactions with other protocols, such as use of information 23 provided by concurrently running unicast routing protocols, or 24 interaction with other multicast protocols, as well as multiple 25 deployment approaches are described. Distributed algorithms for 26 selecting reduced relay sets and related discussion are provided in 27 the appendices. Basic issues relating to the operation of multicast 28 MANET border routers are discussed, but ongoing work remains in this 29 area, and is beyond the scope of this document. 31 Status of this Memo 33 This Internet-Draft is submitted in full conformance with the 34 provisions of BCP 78 and BCP 79. 36 Internet-Drafts are working documents of the Internet Engineering 37 Task Force (IETF). Note that other groups may also distribute 38 working documents as Internet-Drafts. The list of current Internet- 39 Drafts is at http://datatracker.ietf.org/drafts/current/. 41 Internet-Drafts are draft documents valid for a maximum of six months 42 and may be updated, replaced, or obsoleted by other documents at any 43 time. It is inappropriate to use Internet-Drafts as reference 44 material or to cite them other than as "work in progress." 46 This Internet-Draft will expire on January 12, 2012. 48 Copyright Notice 49 Copyright (c) 2011 IETF Trust and the persons identified as the 50 document authors. All rights reserved. 52 This document is subject to BCP 78 and the IETF Trust's Legal 53 Provisions Relating to IETF Documents 54 (http://trustee.ietf.org/license-info) in effect on the date of 55 publication of this document. Please review these documents 56 carefully, as they describe your rights and restrictions with respect 57 to this document. Code Components extracted from this document must 58 include Simplified BSD License text as described in Section 4.e of 59 the Trust Legal Provisions and are provided without warranty as 60 described in the Simplified BSD License. 62 This document may contain material from IETF Documents or IETF 63 Contributions published or made publicly available before November 64 10, 2008. The person(s) controlling the copyright in some of this 65 material may not have granted the IETF Trust the right to allow 66 modifications of such material outside the IETF Standards Process. 67 Without obtaining an adequate license from the person(s) controlling 68 the copyright in such materials, this document may not be modified 69 outside the IETF Standards Process, and derivative works of it may 70 not be created outside the IETF Standards Process, except to format 71 it for publication as an RFC or to translate it into languages other 72 than English. 74 Table of Contents 76 1. Introduction and Scope . . . . . . . . . . . . . . . . . . . . 5 77 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 5 78 3. Applicability Statement . . . . . . . . . . . . . . . . . . . 6 79 4. Overview and Functioning . . . . . . . . . . . . . . . . . . . 7 80 5. SMF Packet Processing and Forwarding . . . . . . . . . . . . . 9 81 6. SMF Duplicate Packet Detection . . . . . . . . . . . . . . . . 11 82 6.1. IPv6 Duplicate Packet Detection . . . . . . . . . . . . . 12 83 6.1.1. IPv6 SMF_DPD Header Option . . . . . . . . . . . . . . 12 84 6.1.2. IPv6 Identification-based DPD . . . . . . . . . . . . 15 85 6.1.3. IPv6 Hash-based DPD . . . . . . . . . . . . . . . . . 17 86 6.2. IPv4 Duplicate Packet Detection . . . . . . . . . . . . . 18 87 6.2.1. IPv4 Identification-based DPD . . . . . . . . . . . . 18 88 6.2.2. IPv4 Hash-based DPD . . . . . . . . . . . . . . . . . 20 89 7. Relay Set Selection . . . . . . . . . . . . . . . . . . . . . 20 90 7.1. Non-Reduced Relay Set Forwarding . . . . . . . . . . . . . 21 91 7.2. Reduced Relay Set Forwarding . . . . . . . . . . . . . . . 21 92 8. SMF Neighborhood Discovery Requirements . . . . . . . . . . . 23 93 8.1. SMF Relay Algorithm TLV Types . . . . . . . . . . . . . . 24 94 8.1.1. SMF Message TLV Type . . . . . . . . . . . . . . . . . 24 95 8.1.2. SMF Address Block TLV Type . . . . . . . . . . . . . . 26 96 9. SMF Border Gateway Considerations . . . . . . . . . . . . . . 26 97 9.1. Forwarded Multicast Groups . . . . . . . . . . . . . . . . 27 98 9.2. Multicast Group Scoping . . . . . . . . . . . . . . . . . 28 99 9.3. Interface with Exterior Multicast Routing Protocols . . . 28 100 9.4. Multiple Border Routers . . . . . . . . . . . . . . . . . 29 101 10. Security Considerations . . . . . . . . . . . . . . . . . . . 30 102 11. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 32 103 11.1. IPv6 SMF_DPD Header Extension Option Type . . . . . . . . 32 104 11.2. TaggerId Types (TidTy) . . . . . . . . . . . . . . . . . . 33 105 11.3. Well-known Multicast Address . . . . . . . . . . . . . . . 34 106 11.4. SMF Type-Length-Values . . . . . . . . . . . . . . . . . . 34 107 11.4.1. Expert Review for created Type Extension Registries . 34 108 11.4.2. SMF Message TLV Type (SMF_TYPE) . . . . . . . . . . . 34 109 11.4.3. SMF Address Block TLV Type (SMF_NBR_TYPE) . . . . . . 35 110 11.4.4. SMF Relay Algorithm ID Registry . . . . . . . . . . . 35 111 12. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 36 112 13. References . . . . . . . . . . . . . . . . . . . . . . . . . . 36 113 13.1. Normative References . . . . . . . . . . . . . . . . . . . 36 114 13.2. Informative References . . . . . . . . . . . . . . . . . . 37 115 Appendix A. Essential Connecting Dominating Set (E-CDS) 116 Algorithm . . . . . . . . . . . . . . . . . . . . . . 39 117 A.1. E-CDS Relay Set Selection Overview . . . . . . . . . . . . 39 118 A.2. E-CDS Forwarding Rules . . . . . . . . . . . . . . . . . . 40 119 A.3. E-CDS Neighborhood Discovery Requirements . . . . . . . . 40 120 A.4. E-CDS Selection Algorithm . . . . . . . . . . . . . . . . 43 121 Appendix B. Source-based Multipoint Relay (S-MPR) . . . . . . . . 45 122 B.1. S-MPR Relay Set Selection Overview . . . . . . . . . . . . 45 123 B.2. S-MPR Forwarding Rules . . . . . . . . . . . . . . . . . . 46 124 B.3. S-MPR Neighborhood Discovery Requirements . . . . . . . . 47 125 B.4. S-MPR Selection Algorithm . . . . . . . . . . . . . . . . 49 126 Appendix C. Multipoint Relay Connected Dominating Set 127 (MPR-CDS) Algorithm . . . . . . . . . . . . . . . . . 50 128 C.1. MPR-CDS Relay Set Selection Overview . . . . . . . . . . . 50 129 C.2. MPR-CDS Forwarding Rules . . . . . . . . . . . . . . . . . 51 130 C.3. MPR-CDS Neighborhood Discovery Requirements . . . . . . . 52 131 C.4. MPR-CDS Selection Algorithm . . . . . . . . . . . . . . . 52 132 Author's Address . . . . . . . . . . . . . . . . . . . . . . . . . 53 134 1. Introduction and Scope 136 Unicast routing protocols, designed for MANET and wireless mesh use, 137 often apply distributed algorithms to flood routing control plane 138 messages within a MANET routing domain. For example, algorithms 139 specified within [RFC3626] and [RFC3684] provide distributed methods 140 of dynamically electing reduced relay sets that attempt to 141 efficiently flood routing control messages while maintaining a 142 connected set under dynamic topological conditions. 144 Simplified Multicast Forwarding (SMF) extends the efficient flooding 145 concept to the data forwarding plane, providing an appropriate 146 multicast forwarding capability for use cases where localized, 147 efficient flooding is considered an effective design approach. The 148 baseline design is intended to provide a basic, best effort multicast 149 forwarding capability that is constrained to operate within a single 150 MANET routing domain. 152 An SMF routing domain is an instance of a SMF routing protocol with 153 common policies, under a single network administration authority. 154 The main design goals of this document are to: 156 o adapt efficient relay sets in MANET environments [RFC2501]; 157 o define the needed IPv4 and IPv6 multicast duplicate packet 158 detection (DPD) mechanisms to support multi-hop, packet 159 forwarding. 161 2. Terminology 163 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 164 "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and 165 "OPTIONAL" in this document are to be interpreted as described in 166 [RFC2119]. 168 The terms introduced in [RFC5444], including "packet", "message", 169 "TLV Block", "TLV", and "address" are to be interpreted as described 170 therein. 172 The following abbreviations are used throughout this document: 174 +--------------+----------------------------------------------------+ 175 | Abbreviation | Definition | 176 +--------------+----------------------------------------------------+ 177 | MANET | Mobile Ad hoc Network | 178 | SMF | Simplified Multicast Forwarding | 179 | CF | Classic Flooding | 180 | CDS | Connected Dominating Set | 181 | MPR | Multi-Point Relay | 182 | S-MPR | Source-based MPR | 183 | MPR-CDS | MPR-based CDS | 184 | E-CDS | Essential CDS | 185 | NHDP | Neighborhood Discovery Protocol | 186 | SMF_DPD | SMF-Duplicate Packet Detection | 187 | I-DPD | Identification-based DPD | 188 | H-DPD | Hash-based DPD | 189 | HAV | Hash-assist Value | 190 | FIB | Forwarding Information Base | 191 | TLV | type-length-value encoding | 192 | DoS | Denial of Service | 193 | SMF Router | A MANET Router, implementing the protocol | 194 | | specified in this document | 195 | SMF Routing | A MANET Routing Domain wherein the protocol, | 196 | Domain | specified in this document, is operating | 197 +--------------+----------------------------------------------------+ 199 3. Applicability Statement 201 Within dynamic wireless routing topologies, maintaining traditional 202 forwarding trees to support a multicast routing protocol is often not 203 as effective as in wired networks due to the reduced reliability and 204 increased dynamics of mesh topologies [MGL04] [GM99]. A basic packet 205 forwarding service reaching all connected routers running the SMF 206 protocol within a MANET routing domain may provide a useful group 207 communication paradigm for various classes of applications, for 208 example multimedia streaming, interactive group-based messaging and 209 applications, peer-to-peer middleware multicasting, and multi-hop 210 mobile discovery or registration services. SMF is likely only 211 appropriate for deployment in limited dynamic MANET routing domains 212 so that the flooding process can be contained, further defined as 213 administratively scoped multicast forwarding domains in Section 9.2. 215 A design goal is, that hosts may also participate in multicast 216 traffic transmission and reception with standard IP network layer 217 semantics (e.g., special or unnecessary encapsulation of IP packets 218 should be avoided in this case). SMF deployments are able to connect 219 and interoperate with existing standard multicast protocols operating 220 within more conventional Internet infrastructures. To this end, a 221 multicast border router or proxy mechanism MUST be used when deployed 222 alongside more fixed-infrastructure IP multicast routing such 223 Protocol Independent Multicast (PIM) variants [RFC3973] and 224 [RFC4601]. Experimental SMF implementations and deployments have 225 demonstrated gateway functionality at MANET border routers operating 226 with existing external IP multicast routing protocols [CDHM07], 227 [DHS08],and [DHG09]. SMF may be extended or combined with other 228 mechanisms to provide increased reliability and group specific 229 filtering; the details for this are out of scope for this document. 231 This document does not presently support forwarding of packets with 232 directed broadcast addresses as a destination [RFC2644]. 234 4. Overview and Functioning 236 Figure 1 provides an overview of the logical SMF router architecture, 237 consisting of "Neighborhood Discovery", "Relay Set Selection" and 238 "Forwarding Process" components. Typically, relay set selection (or 239 self-election) occurs based on dynamic input from a neighborhood 240 discovery process. SMF supports the case where neighborhood 241 discovery and/or relay set selection information is obtained from a 242 coexistent process (e.g., a lower layer mechanism or a unicast 243 routing protocol using relay sets). In some algorithm designs, the 244 forwarding decision for a packet can also depend on previous hop or 245 incoming interface information. The asterisks (*) in Figure 1 mark 246 the primitives and relationships, needed by relay set algorithms 247 requiring previous-hop packet forwarding knowledge. 248 ______________ _____________ 249 | | | | 250 | Neighborhood | | Relay Set | 251 | Discovery |------------->| Selection | 252 | Protocol | neighbor | Algorithm | 253 |______________| info |_____________| 254 \ / 255 \ / 256 neighbor\ /forwarding 257 info* \ ____________ / status 258 \ | | / 259 `-->| Forwarding |<--' 260 | Process | 261 ~~~~~~~~~~~~~~~~~>|____________|~~~~~~~~~~~~~~~~~> 262 incoming packet, forwarded packets 263 interface id*, and 264 previous hop* 266 Figure 1: SMF router Architecture 268 Certain IP multicast packets, defined in Section 9.2 and Section 5, 269 are "non-forwardable". These multicast packets MUST be ignored by 270 the SMF forwarding engine. The SMF forwarding engine MAY also work 271 with policies and management interfaces to allow additional filtering 272 control over which multicast packets are considered for potential SMF 273 forwarding. This interface would allow more refined dynamic 274 forwarding control once such techniques are matured for MANET 275 operation. At present, further discussion of dynamic control is left 276 to future work. 278 Interoperable SMF implementations MUST use a common DPD approach and 279 be able to process the header options defined in this document for 280 IPv6 operation. 282 Classic Flooding (CF) is defined as the simplest case of SMF 283 multicast forwarding. With CF, each SMF router forwards each 284 received multicast packet exactly once. In this case, the need for 285 any relay set selection or neighborhood topology information is 286 eliminated, at the expense of additional network overhead incurring 287 from unnecessary packet retransmissions. With CF, the SMF_DPD 288 functionality is still required. While SMF supports CF as a mode of 289 operation, the use of more efficient relay set modes is RECOMMENDED 290 in order to reduce contention and congestion caused by unnecessary 291 packet retransmissions [NTSC99]. 293 An efficient, reduced relay set is constructed by selecting and 294 maintaining a subset of all possible routers in a MANET routing 295 domain. Known distributed relay set selection algorithms have 296 demonstrated the ability to provide and maintain a dynamic connected 297 set for forwarding multicast IP packets [MDC04]. A few such relay 298 set selection algorithms are described in the Appendices of this 299 document and the basic designs borrow directly from previously 300 documented IETF work. SMF relay set configuration is extensible and 301 additional relay set algorithms beyond those specified here can be 302 accommodated in future work. 304 Determining and maintaining an optimized set of relays generally 305 requires dynamic neighborhood topology information. Neighborhood 306 topology discovery functions MAY be provided by a MANET unicast 307 routing protocol or by using the MANET NeighborHood Discovery 308 Protocol (NHDP) [RFC6130], operating concurrently with SMF. This 309 specification allows alternative lower layer interfaces (e.g., radio 310 router interface) to provide the necessary neighborhood information 311 to aid in supporting more effective relay set election. An SMF 312 implementation SHOULD provide the ability for multicast forwarding 313 state to be dynamically managed per operating network interface. The 314 relay state maintenance options and interactions are outlined in 315 Section 7. This document states specific requirements for 316 neighborhood discovery with respect to the forwarding process and the 317 relay set selection algorithms described herein. For determining 318 dynamic relay sets in the absence of other control interfaces, SMF 319 relies on NHDP to assist in IP layer 2-hop neighborhood discovery and 320 maintenance for relay set election. "SMF_TYPE" and "SMF_NBR_TYPE" 321 Message and Address Block TLV structures (per [RFC5444]) are defined 322 by this document for use with NHDP for carrying SMF specific 323 information. It is RECOMMENDED that all routers performing SMF 324 operation in conjunction with NHDP, include these TLV types in any 325 NHDP HELLO messages generated. This capability allows for routers 326 participating in SMF to be explicitly identified along with their 327 respective dynamic relay set algorithm. 329 5. SMF Packet Processing and Forwarding 331 The SMF Packet Processing and Forwarding actions are conducted with 332 the following packet handling activities: 334 1. Processing of outbound, locally-generated multicast packets. 335 2. Reception and processing of inbound packets on specific network 336 interfaces. 338 The purpose of intercepting outbound, locally-generated multicast 339 packets is to apply any added packet marking needed to satisfy the 340 DPD requirements so that proper forwarding may be conducted. Note 341 that for some system configurations the interception of outbound 342 packets for this purpose is not necessary. 344 Inbound multicast packets are received by the SMF implementation and 345 processed for possible forwarding. SMF implementations MUST be 346 capable of forwarding IP multicast packets with destination addresses 347 that are not router-local and link-local for IPv6, as defined in 348 [RFC4291], and that are not within the local network control block as 349 defined by [RFC5771]. 351 This will support generic multi-hop multicast application needs or 352 distribute designated multicast traffic ingressing the SMF routing 353 domain via border routers. The multicast addresses to be forwarded 354 should be maintained by an a priori list or a dynamic forwarding 355 information base (FIB) that MAY interact with future MANET dynamic 356 group membership extensions or management functions. 358 The SL-MANET-ROUTERS multicast group is defined to contain all 359 routers within an SMF routing domain, so that packets transmitted to 360 the multicast address associated with the group will be attempted 361 delivered to all connected routers running SMF. Due the mobile 362 nature of a MANET, routers running SMF may not be topologically 363 connected at particular times. For IPv6, SL-MANET-ROUTERS is 364 specified to be "site-local". Minimally SMF MUST forward, as 365 instructed by the relay set selection algorithm, unique (non- 366 duplicate) packets received for SL-MANET-ROUTERS when the TTL/ 367 hop-limit or hop limit value in the IP header is greater than 1. SMF 368 MUST forward all additional global scope multicast addresses 369 specified within the dynamic FIB or configured list as well. In all 370 cases, the following rules MUST be observed for SMF multicast 371 forwarding: 373 1. Any IP packets not addressed to an IP multicast address MUST NOT 374 be forwarded by the SMF forwarding engine 375 2. IP multicast packets with TTL/hop-limit <= 1 MUST NOT be 376 forwarded. 377 3. Link local IP multicast packets MUST NOT be forwarded. 378 4. Incoming IP multicast packets with an IP source address matching 379 one of those of the local SMF router interface(s) MUST NOT be 380 forwarded. 381 5. Received frames with the MAC source address matching any MAC 382 address of the router's interfaces MUST NOT be forwarded. 383 6. Received packets for which SMF cannot reasonably ensure temporal 384 DPD uniqueness MUST NOT be forwarded. 385 7. Prior to being forwarded, the TTL/hop-limit of the forwarded 386 packet MUST be decremented by one. 388 Note that rule #3 is important because over some types of wireless 389 interfaces, the originating SMF router may receive re-transmissions 390 of its own packets when they are forwarded by adjacent routers. This 391 rule avoids unnecessary retransmission of locally-generated packets 392 even when other forwarding decision rules would apply. 394 An additional processing rule also needs to be considered based upon 395 a potential security threat. As discussed in Section 10, there may 396 be a concern in some SMF deployments that malicious routers may 397 conduct a denial-of-service attack by remotely "previewing" (e.g., 398 via a directional receive antenna) packets that an SMF router would 399 be forwarding and conduct a "pre-play" attack by transmitting the 400 packet before the SMF router would otherwise receive it but with a 401 reduced TTL/hop-limit field value. This form of attack can cause an 402 SMF router to create a DPD entry that would block the proper 403 forwarding of the valid packet (with correct TTL/hop-limit) through 404 the SMF routing domain. A RECOMMENDED approach to prevent this 405 attack, when it is a concern, would be to cache temporal packet TTL/ 406 hop-limit values along with the per-packet DPD state (hash value(s) 407 and/or identifier as described in Section 6). Then, if a subsequent 408 matching (with respect to DPD) packet arrives with a larger TTL/ 409 hop-limit value than the packet that was previously forwarded, SMF 410 should forward the new packet and update the TTL/hop-limit value 411 cached with corresponding DPD state to the new, larger TTL/hop-limit 412 value. There may be temporal cases where SMF would unnecessarily 413 forward some duplicate packets using this approach, but those cases 414 are expected to be minimal and acceptable when compared with the 415 potential threat of denied service. 417 Once the SMF multicast forwarding rules have been applied, an SMF 418 implementation MUST make a forwarding decision dependent upon the 419 relay set selection algorithm in use. If the SMF implementation is 420 using Classic Flooding (CF), the forwarding decision is implicit once 421 DPD uniqueness is determined. Otherwise, a forwarding decision 422 depends upon the current interface-specific relay set state. The 423 descriptions of the relay set selection algorithms in the Appendices 424 to this document specify the respective heuristics for multicast 425 packet forwarding and specific DPD or other processing required to 426 achieve correct SMF behavior in each case. For example, one class of 427 forwarding is based upon relay set election status and the packet's 428 previous hop, while other classes designate the local SMF router as a 429 forwarder for all neighboring routers. 431 6. SMF Duplicate Packet Detection 433 Duplicate packet detection (DPD) is often a requirement in MANET or 434 wireless mesh packet forwarding mechanisms because packets may be 435 transmitted out via the same physical interface as the one over which 436 they were received. Routers may also receive multiple copies of the 437 same packets from different neighbors. SMF operation requires DPD 438 and implementations MUST provide mechanisms to detect and reduce the 439 likelihood of forwarding duplicate multicast packets using temporal 440 packet identification. It is RECOMMENDED this be implemented by 441 keeping a history of recently-processed multicast packets for 442 comparison with incoming packets. A DPD packet cache history SHOULD 443 be kept long enough so as to span the maximum network traversal 444 lifetime, MAX_PACKET_LIFETIME, of multicast packets being forwarded 445 within an SMF routing domain. The DPD mechanism SHOULD avoid keeping 446 unnecessary state for packet flows such as those that are locally- 447 generated or link-local destinations that would not be considered for 448 forwarding, as presented in Section 5. 450 For IPv4 and IPv6, both, this document describes two basic multicast 451 duplicate packet detection mechanisms: header content identification- 452 based (I-DPD) and hash-based (H-DPD) duplicate detection. I-DPD is a 453 mechanism using specific packet headers, and option headers in the 454 case of IPv6, in combination with flow state to estimate the temporal 455 uniqueness of a packet. H-DPD uses hashing over header fields and 456 payload of a multicast packet, in order to provide an estimation of 457 temporal uniqueness. 459 Trade-offs of the two approaches to DPD merit different 460 considerations dependent upon the specific SMF deployment scenario. 462 Because of the potential addition of a hop-by-hop option header with 463 IPv6, all SMF routers in the same SMF deployments MUST be configured 464 so as to use a common mechanism and DPD algorithm. The main 465 difference between IPv4 and IPv6 SMF_DPD specification is the 466 avoidance of any additional header options for IPv4. 468 For each network interface, SMF implementations MUST maintain DPD 469 packet state as needed to support the forwarding heuristics of the 470 relay set algorithm used. In general, this involves keeping track of 471 previously forwarded packets so that duplicates are not forwarded, 472 but some relay techniques have additional considerations, such as 473 discussed in Appendix B.2. 475 Additional details of I-DPD and H-DPD processing and maintenance for 476 different classes of packets are described in the following. 478 6.1. IPv6 Duplicate Packet Detection 480 This section describes the mechanisms and options for SMF IPv6 DPD. 481 The base IPv6 packet header does not provide any explicit packet 482 identification header field that can be exploited for I-DPD. The 483 following options are therefore described, to support IPv6 DPD: 484 1. a hop-by-hop SMF_DPD option header, defined in this document 485 (Section 6.1.1), 486 2. the use of IPv6 fragment header fields for I-DPD, if one is 487 present (Section 6.1.2), 488 3. the use of IPsec sequencing for I-DPD when a non-fragmented, 489 IPsec header is detected (Section 6.1.2), and 490 4. an H-DPD approach assisted, as needed, by the SMF_DPD option 491 header (Section 6.1.3). 493 SMF MUST provide a DPD marking module that can insert the hop-by-hop 494 IPv6 header option, defined Section 6.1.1. This module MUST be 495 invoked after any source-based fragmentation that may occur with 496 IPv6, so as to ensure that all fragments are suitably marked. SMF 497 IPv6 DPD is presently specified to allow either a packet hash or 498 header identification method for DPD. An SMF implementation MUST be 499 configured to operate either in H-DPD or I-DPD mode, and perform the 500 corresponding tasks, outlined in Section 6.1.1 and Section 6.1.3. 502 6.1.1. IPv6 SMF_DPD Header Option 504 This section defines an IPv6 Hop-by-Hop Option [RFC2460], SMF_DPD, to 505 serve the purpose of unique packet identification for IPv6 I-DPD. 506 Additionally, the SMF_DPD header option provides a mechanism to 507 guarantee non-collision of hash values for different packets when 508 H-DPD is used. 510 If this is the only hop-by-hop option present, the optional 511 "TaggerId" field (see below) is not included, and the size of the DPD 512 packet identifier (sequence number) or hash token is 24 bits or less, 513 this will result in the addition of 8 bytes to the IPv6 packet header 514 including the "Next Header", "Header Extension Length", SMF_DPD 515 option fields, and padding. 517 0 1 2 3 518 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 519 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 520 ... |0|0|0| SMF_DPD | Opt. Data Len | 521 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 522 |H| DPD Identifier Option Fields or Hash Assist Value ... 523 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 525 IPv6 SMF_DPD Hop-by-Hop Header Option 527 "Option Type" = (Lower 5 bits pending IANA assignment, highest order 528 MUST be 000). By having these three bits be zero, this specification 529 requires that routers not recognizing this option type should skip 530 over this option and continue processing the header and that the 531 option must not change en route [RFC2460]. 533 "Opt. Data Len" = Length of option content (I.e., 1 + ( ? 534 ( + 1): 0) + Length(DPD ID)). 536 "H-bit" = a hash indicator bit value identifying DPD marking type. 0 537 == sequence-based approach with optional TaggerId and a tuple-based 538 sequence number. 1 == indicates a hash assist value (HAV) field 539 follows to aid in avoiding hash-based DPD collisions. 541 When the "H-bit" is cleared (zero value), the SMF_DPD format to 542 support I-DPD operation is specified as shown in Figure 2 and defines 543 the extension header in accordance with [RFC2460]. 545 0 1 2 3 546 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 547 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 548 ... |0|0|0| OptType | Opt. Data Len | 549 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 550 |0|TidTy| TidLen| TaggerId (optional) ... | 551 +-+-+-+-+-+-+-+-+ +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 552 | | Identifier ... 553 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 555 Figure 2: IPv6 SMF_DPD Header Option in I-DPD mode 557 "TidTy" = a 3-bit field indicating the presence and type of the 558 optional "TaggerId" field. "TidLen" = a 4-bit field indicating the 559 length (in octets) of the following TaggerId field. "TaggerId" = a 560 field, is used to differentiate multiple ingressing border gateways 561 that may commonly apply the SMF_DPD option header to packets from a 562 particular source. Table 1 lists the TaggerId types, used in this 563 document: 565 +---------+---------------------------------------------------------+ 566 | Name | Purpose | 567 +---------+---------------------------------------------------------+ 568 | NULL | Indicates no "TaggerId" field is present. "TidLen" | 569 | | MUST also be set to ZERO. | 570 | DEFAULT | A "TaggerId" of non-specific context is present. | 571 | | "TidLen + 1" defines the length of the TaggerId field | 572 | | in bytes. | 573 | IPv4 | A "TaggerId" representing an IPv4 address is present. | 574 | | The "TidLen" MUST be set to 3. | 575 | IPv6 | A "TaggerId" representing an IPv6 address is present. | 576 | | The "TidLen" MUST be set to 15. | 577 +---------+---------------------------------------------------------+ 579 Table 1: TaggerId Types 581 This format allows a quick check of the "TidTy" field to determine if 582 a "TaggerId" field is present. If "TidTy" is NULL, then the length 583 of the DPD packet field corresponds to ( 584 - 1). If the is non-NULL, then the length of the "TaggerId" 585 field is equal to ( - 1) and the remainder of the option data 586 comprises the DPD packet field. When the "TaggerId" 587 field is present, the field can be considered a unique 588 packet identifier in the context of the 589 tuple. When the "TaggerId" field is not present, then it is assumed 590 the source applied the SMF_DPD option and the can be 591 considered unique in the context of the IPv6 packet header tuple. IPv6 I-DPD operation details are described in 593 Section 6.1.2. 595 When the "H-bit" in the SMF_DPD option data is set, the data content 596 value is interpreted as a Hash-Assist Value (HAV) used to facilitate 597 H-DPD operation. In this case, the source or ingressing gateways 598 apply the SMF_DPD with an HAV only when required to differentiate the 599 hash value of a new packet with respect to hash values in the DPD 600 cache. This situation can be detected locally on the router by 601 running the hash algorithm and checking the DPD cache, prior to 602 ingressing a previously unmarked packet or a locally sourced packet. 603 This helps to guarantee the uniqueness of generated hash values when 604 H-DPD is used. Additionally, this also avoids the added overhead of 605 applying the SMF_DPD option header to every packet. For many hash 606 algorithms, it is expected that only sparse use of the SMF_DPD option 607 may be required. The format of the SMF_DPD header option for H-DPD 608 operation is given in Figure 3. 610 0 1 2 3 611 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 612 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 613 ... |0|0|0| OptType | Opt. Data Len | 614 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 615 |1| Hash Assist Value (HAV) ... 616 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 618 Figure 3: IPv6 SMF_DPD Header Option in H-DPD Mode 620 The SMF_DPD option should be applied with an HAV to produce a unique 621 hash digest for packets within the context of the IPv6 packet header 622 . The size of the HAV field is implied by the "Opt. Data 623 Len". The appropriate size of the field depends upon the collision 624 properties of the specific hash algorithm used. More details on IPv6 625 H-DPD operation are provided in Section 6.1.3. 627 6.1.2. IPv6 Identification-based DPD 629 Table 2 summarizes the IPv6 I-DPD processing and forwarding decision 630 approach. Within the table '*' indicates an ignore field condition. 632 +-------------+-----------+-----------+-----------------------------+ 633 | IPv6 | IPv6 | IPv6 | SMF IPv6 I-DPD Mode Action | 634 | Fragment | IPsec | I-DPD | | 635 | Header | Header | Header | | 636 +-------------+-----------+-----------+-----------------------------+ 637 | Present | * | Not | Use Fragment Header I-DPD | 638 | | | Present | Check and Process for | 639 | | | | Forwarding | 640 | Not Present | Present | Not | Use IPsec Header I-DPD | 641 | | | Present | Check and Process for | 642 | | | | Forwarding | 643 | Present | * | Present | Invalid, do not Forward | 644 | Not Present | Present | Present | Invalid, do not Forward | 645 | Not Present | Not | Not | Add I-DPD Header,and | 646 | | Present | Present | Process for Forwarding | 647 | Not Present | Not | Present | Use I-DPD Header Check and | 648 | | Present | | Process for Forwarding | 649 +-------------+-----------+-----------+-----------------------------+ 651 Table 2: IPv6 I-DPD Processing Rules 653 1. If a received IPv6 multicast packet is an IPv6 fragment, SMF MUST 654 use the fragment extension header fields for packet 655 identification. This identifier can be considered unique in the 656 context of the of the IP packet. 657 2. If the packet is an unfragmented IPv6 IPsec packet, SMF MUST use 658 IPsec fields for packet identification. The IPsec header 659 field can be considered a unique identifier in the 660 context of the where the 661 "IPsecType" is either AH or ESP [RFC4302]. 662 3. For unfragmented, non-IPsec, IPv6 packets, the use of the SMF_DPD 663 header option is necessary to support I-DPD operation. The 664 SMF_DPD header option is applied in the context of the 665 of the IP packet. hosts or ingressing SMF gateways are 666 responsible for applying this option to support DPD. Table 3 667 summarizes these packet identification types: 669 +-----------+---------------------------------+---------------------+ 670 | IPv6 | Packet DPD ID Context | Packet DPD ID | 671 | Packet | | | 672 | Type | | | 673 +-----------+---------------------------------+---------------------+ 674 | Fragment | | | 675 | IPsec | | | 676 | Packet | | | 677 | Regular | <[TaggerId:]srcAddr:dstAddr> | | 679 +-----------+---------------------------------+---------------------+ 680 Table 3: IPv6 I-DPD Packet Identification Types 682 "IPsecType" is either Authentication Header (AH) or Encapsulating 683 Security Payload (ESP). 685 The "TaggerId" is an optional field of the IPv6 SMF_DPD header 686 option. 688 6.1.3. IPv6 Hash-based DPD 690 A default hash-based DPD approach (H-DPD) for use by SMF is specified 691 as follows. An SHA-1 [RFC3174] hash of the non-mutable header 692 fields, options fields, and data content of the IPv6 multicast packet 693 is used to produce a 128-bit digest. The approach for calculating 694 this hash value SHOULD follow the same guidelines described for 695 calculating the Integrity Check Value (ICV) described in [RFC4302] 696 with respect to non-mutable fields. This approach should have a 697 reasonably low probability of digest collision when packet headers 698 and content are varying. SHA-1 is being applied in SMF only to 699 provide a low probability of collision and is not being used for 700 cryptographic or authentication purposes. A history of the packet 701 hash values SHOULD be maintained within the context of the IPv6 702 packet header . SMF ingress points (i.e., source hosts or 703 gateways) use this history to confirm that new packets are unique 704 with respect to their hash value. The Hash-assist Value (HAV) field 705 described in Section 6.1.1 is provided as a differentiating field 706 when a digest collision would otherwise occur. Note that the HAV is 707 an immutable option field and SMF MUST process any included HAV 708 values (see Section 6.1.1) in its hash calculation. 710 If a packet results in a digest collision (i.e., by checking the 711 H-DPD digest history) within the DPD cache kept by SMF forwarders, 712 the packet SHOULD be silently dropped. If a digest collision is 713 detected at an SMF ingress point the H-DPD option header is 714 constructed with a randomly generated HAV. A HAV is recalculated as 715 needed to produce a non-colliding hash value prior to forwarding. 716 The multicast packet is then forwarded with the added IPv6 SMF_DPD 717 header option. 719 The SHA-1 indexing and IPv6 HAV approaches are specified at present 720 for consistency and robustness to suit experimental uses. Future 721 approaches and experimentation may discover design tradeoffs in hash 722 robustness and efficiency worth considering. Enhancements MAY 723 include reducing the maximum payload length that is processed, 724 determining shorter indexes, or applying more efficient hashing 725 algorithms. Use of the HAV functionality may allow for application 726 of "lighter-weight" hashing techniques that might not have been 727 initially considered due to poor collision properties otherwise. 729 Such techniques could reduce packet processing overhead and memory 730 requirements. 732 6.2. IPv4 Duplicate Packet Detection 734 This section describes the mechanisms and options for IPv4 DPD. The 735 IPv4 packet header [RFC0791] 16-bit "Identification" field MAY be 736 used for DPD assistance, but practical limitations may require 737 alternative approaches in some situations. The following areas are 738 described to support IPv4 DPD: 740 1. the use of IPv4 fragment header fields for I-DPD when they exist 741 (Section 6.2.1), 742 2. the use of IPsec sequencing for I-DPD when a non-fragmented IPv4 743 IPsec packet is detected (Section 6.2.1), and 744 3. an H-DPD approach(Section 6.2.2). 746 A specific SMF_DPD marking option is not specified for IPv4 since 747 header options are not as tractable for hosts as they are for IPv6. 748 IPv4 packets from a particular source are assumed to be marked with a 749 temporally unique value in the "Identification" field of the packet 750 header that can serve for SMF_DPD purposes. However, in present 751 operating system networking kernels, the IPv4 header "Identification" 752 value is not always generated properly, especially when the "don't 753 fragment" (DF) bit is set. The IPv4 I-DPD mode of this specification 754 requires that IPv4 "Identification" fields are managed reasonably by 755 hosts, and that temporally unique values are set within the context 756 of the packet header tuple. If this is 757 not expected during an SMF deployment, then it is RECOMMENDED that 758 the H-DPD method be used as a more reliable approach. 760 Since IPv4 SMF does not specify an option header, the 761 interoperability constraints are looser than the IPv6 version and 762 forwarders may be operate with mixed H-DPD and I-DPD modes as long as 763 they consistently perform the appropriate DPD routines outlined in 764 the following sections. However, it is RECOMMENDED that a deployment 765 be configured with a common mode for operational consistency. 767 6.2.1. IPv4 Identification-based DPD 769 Table 4 summarizes the IPv4 I-DPD processing approach once a packet 770 has passed the basic forwardable criteria described in Section 5. 771 Within the table '*' indicates an ignore field condition. DF, MF, 772 Fragment offset correspond to related fields and flags defined in 773 [RFC0791]. 775 +------+------+----------+---------+--------------------------------+ 776 | DF | MF | Fragment | IPsec | IPv4 I-DPD Action | 777 | flag | flag | offset | | | 778 +------+------+----------+---------+--------------------------------+ 779 | 1 | 1 | * | * | Invalid, Do Not Forward | 780 | 1 | 0 | nonzero | * | Invalid, Do Not Forward | 781 | * | 0 | zero | not | Tuple I-DPD Check and Process | 782 | | | | Present | for Forwarding | 783 | * | 0 | zero | Present | IPsec enhanced Tuple I-DPD | 784 | | | | | Check and Process for | 785 | | | | | Forwarding | 786 | 0 | 0 | nonzero | * | Extended Fragment Offset Tuple | 787 | | | | | I-DPD Check and Process for | 788 | | | | | Forwarding | 789 | 0 | 1 | zero or | * | Extended Fragment Offset Tuple | 790 | | | nonzero | | I-DPD Check and Process for | 791 | | | | | Forwarding | 792 +------+------+----------+---------+--------------------------------+ 794 Table 4: IPv4 I-DPD Processing Rules 796 For performance reasons, IPv4 network fragmentation and reassembly of 797 multicast packets within wireless MANET networks should be minimized, 798 yet SMF provides the forwarding of fragments when they occur. If the 799 IPv4 multicast packet is a fragment, SMF MUST use the fragmentation 800 header fields for packet identification. This identification can be 801 considered temporally unique in the context of the of the IPv4 packet. If the packet is an unfragmented IPv4 803 IPsec packet, SMF MUST use IPsec fields for packet identification. 804 The IPsec header field can be considered a unique 805 identifier in the context of the 806 where the "IPsecType" is either AH or ESP [RFC4302]. Finally, for 807 unfragmented, non-IPsec, IPv4 packets, the "Identification" field can 808 be used for I-DPD purposes. The "Identification" field can be 809 considered unique in the context of the IPv4 tuple. Table 5 summarizes these packet identification 811 types: 813 +-----------+---------------------------------+---------------------+ 814 | IPv4 | Packet Identification Context | Packet Identifier | 815 | Packet | | | 816 | Type | | | 817 +-----------+---------------------------------+---------------------+ 818 | Fragment | | | 819 | IPsec | | | 820 | Packet | | | 821 | Regular | | | 823 +-----------+---------------------------------+---------------------+ 825 Table 5: IPv4 I-DPD Packet Identification Types 827 "IPsecType" is either Authentication Header (AH) or Encapsulating 828 Security Payload (ESP). 830 The limited size (16 bits) of the IPv4 header "Identification" field 831 [RFC0791] may result in more frequent value field wrapping, 832 particularly if a common sequence space is used by a source for 833 multiple destinations. If I-DPD operation is required, the use of 834 the "internal hashing" technique described in Section 10 may mitigate 835 this limitation of the IPv4 "Identification" field for SMF_DPD. In 836 this case the "internal hash" value would be concatenated with the 837 "Identification" value for I-DPD operation. 839 6.2.2. IPv4 Hash-based DPD 841 To ensure consistent IPv4 H-DPD operation among SMF routers, a 842 default hashing approach is specified. This is similar to that 843 specified for IPv6 in Section 6.1.3, but the H-DPD header option with 844 HAV is not considered. SMF MUST perform an SHA-1 [RFC3174] hash of 845 the immutable header fields, option fields and data content of the 846 IPv4 multicast packet resulting in a 128-bit digest. The approach 847 for calculating the hash value SHOULD follow the same guidelines 848 described for calculating the Integrity Check Value (ICV) described 849 in [RFC4302] with respect to non-mutable fields. A history of the 850 packet hash values SHOULD be maintained in the context of . The context for IPv4 is more specific than that of 852 IPv6 since the SMF_DPD HAV cannot be employed to mitigate hash 853 collisions. 855 The SHA-1 hash is specified at present for consistency and 856 robustness. Future approaches and experimentation may discover 857 design tradeoffs in hash robustness and efficiency worth considering 858 for future revisions of SMF. This MAY include reducing the packet 859 payload length that is processed, determining shorter indexes, or 860 applying a more efficient hashing algorithm. 862 7. Relay Set Selection 864 SMF is flexible in its support of different reduced relay set 865 mechanism for efficient flooding, the constraints imposed hereon 866 being detailed in this section. 868 7.1. Non-Reduced Relay Set Forwarding 870 SMF implementations MUST support CF as a basic forwarding mechanism 871 when reduced relay set information is not available or not selected 872 for operation. In CF mode, each router transmits a packet once that 873 has passed the SMF forwarding rules. The DPD techniques described in 874 Section 6 are critical to proper operation and prevent duplicate 875 packet retransmissions by the same relays. 877 7.2. Reduced Relay Set Forwarding 879 MANET reduced relay sets are often achieved by distributed algorithms 880 that can dynamically calculate a topological connected dominating set 881 (CDS). 883 A goal of SMF is to apply reduced relay sets for more efficient 884 multicast dissemination within dynamic topologies. To accomplish 885 this an SMF implementation MUST support the ability to modify its 886 multicast packet forwarding rules based upon relay set state received 887 dynamically during operation. In this way, SMF operates effectively 888 as neighbor adjacencies or multicast forwarding policies within the 889 topology change. 891 In early SMF experimental prototyping, the relay set information has 892 been derived from coexistent unicast routing control plane traffic 893 flooding processes [MDC04]. From this experience, extra pruning 894 considerations were sometimes required when utilizing a relay set 895 from a separate routing protocol process. As an example, relay sets 896 formed for the unicast control plane flooding MAY include additional 897 redundancy that may not be desired for multicast forwarding use 898 (e.g., biconnected relay set). 900 Here is a recommended criteria list for SMF relay set selection 901 algorithm candidates: 903 1. Robustness to topological dynamics and mobility 904 2. Localized election or coordination of any relay sets 905 3. Reasonable minimization of CDS relay set size given above 906 constraints 907 4. Heuristic support for preference or election metrics 909 Some relay set algorithms meeting these criteria are described in the 910 Appendices of this document. Additional relay set selection 911 algorithms may be specified in separate specifications in the future. 912 Each Appendix subsection in this document can serve as a template for 913 specifying additional relay algorithms. 915 Figure 4 depicts an information flow diagram of possible relay set 916 control options. The SMF Relay Set State represents the information 917 base that is used by SMF in the forwarding decision process. The 918 relay set control option diagram demonstrates that the SMF relay set 919 state may be determined by fundamentally three different methods: 921 o Independent operation with NHDP [RFC6130] input providing dynamic 922 network neighborhood adjacency information, used by a particular 923 relay set selection algorithm. 924 o Slave operation with an existing unicast MANET routing protocol, 925 capable of providing CDS election information for use by SMF. 926 o Cross layer operation that may involve L2 triggers / Information 927 describing neighbors or links. 929 Other heuristics to influence and control election can come from 930 network management or other interfaces as shown on the right of 931 Figure 4. CF mode simplifies the control and does not require other 932 input but relies solely on DPD. 934 Possible L2 Trigger/Information 935 | 936 | 937 ______________ ______v_____ __________________ 938 | MANET | | | | | 939 | Neighborhood | | Relay Set | | Other Heuristics | 940 | Discovery |----------->| Selection |<------| (Preference,etc) | 941 | Protocol | neighbor | Algorithm | | Net Management | 942 |______________| info |____________| |__________________| 943 \ / 944 \ / 945 neighbor\ / Dynamic Relay 946 info* \ ____________ / Set Status 947 \ | SMF | / (State, {neighbor info}) 948 `-->| Relay Set |<--' 949 | State | 950 -->|____________| 951 / 952 / 953 ______________ 954 | Coexistent | 955 | MANET | 956 | Unicast | 957 | Process | 958 |______________| 960 Figure 4: SMF Reduced Relay Set Information Flow 962 More discussion is provided on the three styles of SMF operation with 963 reduced relay sets as illustrated in Figure 4: 965 1. Independent operation: In this case, SMF operates independently 966 from any unicast routing protocols. To support reduced relay 967 sets SMF MUST perform its own relay set selection using 968 information gathered from signaling. It is RECOMMENDED that an 969 associated NHDP process be used for this signaling. NHDP 970 messaging SHOULD be appended with additional [RFC5444] type- 971 length-value (TLV) content to support SMF-specific requirements 972 as discussed in [RFC6130] and for the applicable relay set 973 algorithm described in the Appendices of this document or future 974 specifications. Unicast routing protocols may co-exist, even 975 using the same NHDP process, but signaling that supports reduced 976 relay set selection for SMF is independent of these protocols. 977 2. Operation with CDS-aware unicast routing protocol: In this case, 978 a coexistent unicast routing protocol provides dynamic relay set 979 state based upon its own control plane CDS or neighborhood 980 discovery information. 981 3. Cross-layer Operation: In this case, SMF operates using 982 neighborhood status and triggers from a cross-layer information 983 base for dynamic relay set selection and maintenance (e.g., lower 984 link layer). 986 8. SMF Neighborhood Discovery Requirements 988 This section defines the requirements for use of the MANET 989 Neighborhood Discovery Protocol (NHDP) [RFC6130] to support SMF 990 operation. Note that basic CF forwarding requires no neighborhood 991 topology knowledge since in this configured mode every SMF router 992 relays all traffic. Supporting more reduced SMF relay set operation 993 requires the discovery and maintenance of dynamic neighborhood 994 topology information. NHDP can be used to provide this necessary 995 information, however there are SMF-specific requirements for NHDP 996 use. This is the case for both "independent" SMF operation where 997 NHDP is being used specifically to support SMF or when one NHDP 998 instance is used, both, for SMF and a coexistent MANET unicast 999 routing protocol. 1001 NHDP HELLO messages and the resultant neighborhood information base 1002 are described separately within the NHDP specification. To 1003 summarize, NHDP provides the following basic functions: 1005 1. 1-hop neighbor link sensing and bidirectionality checks of 1006 neighbor links, 1007 2. 2-hop neighborhood discovery including collection of 2-hop 1008 neighbors and connectivity information, 1009 3. Collection and maintenance of the above information across 1010 multiple interfaces, and 1012 4. A method for signaling SMF information throughout the 2-hop 1013 neighborhood through the use of TLV extensions. 1015 Appendices (A-C) of this document describe CDS-based relay set 1016 selection algorithms that can achieve efficient SMF operation, even 1017 in dynamic, mobile networks and each of the algorithms has been 1018 initially experimented within a working SMF prototype [MDDA07]. When 1019 using these algorithms in conjunction with NHDP, a method verifying 1020 neighbor SMF operation is required in order to insure correct relay 1021 set selection. NHDP along with SMF operation verification provides 1022 the necessary information required by these algorithms to conduct 1023 relay set selection. Verification of SMF operation may be done 1024 administratively or through the use of the SMF relay algorithms TLVs 1025 defined in the following subsections. Use of the SMF relay algorithm 1026 TLVs is RECOMMENDED when using NHDP for SMF neighborhood discovery. 1028 Section 8.1 specifies SMF-specific TLV types, supporting general SMF 1029 operation or supporting the algorithms described in the Appendices. 1030 The Appendices describing several relay set algorithms also specify 1031 any additional requirements for use with NHDP and reference the 1032 applicable TLV types as needed. 1034 8.1. SMF Relay Algorithm TLV Types 1036 This section specifies TLV types to be used within NHDP messages to 1037 identify the CDS relay set selection algorithm(s) in use. Two TLV 1038 types are defined, one Message TLV type and one Address Block TLV 1039 type. 1041 8.1.1. SMF Message TLV Type 1043 The Message TLV type denoted SMF_TYPE is used to identify the 1044 existence of an SMF instance operating in conjunction with NHDP. 1045 This Message TLV type makes use of the extended type field as defined 1046 by [RFC5444] to convey the CDS relay set selection algorithm 1047 currently in use by the SMF message originator. When NHDP is used to 1048 support SMF operation, the SMF_TYPE TLV, containing the extended type 1049 field with the appropriate value, SHOULD be included in NHDP_HELLO 1050 messages (HELLO messages as defined in [RFC6130]). This allows SMF 1051 routers to learn when neighbors are configured to use NHDP for 1052 information exchange including algorithm type and related algorithm 1053 information. This information can be used to take action, such as 1054 ignoring neighbor information using incompatible algorithms. It is 1055 possible that SMF neighbors MAY be configured differently and still 1056 operate cooperatively, but these cases will vary dependent upon the 1057 algorithm types designated. 1059 This document defines a Message TLV type as specified in Table 6 1060 conforming to [RFC5444]. The TLV extended type field is used to 1061 contain the sender's "Relay Algorithm Type". The interpretation of 1062 the "value" content of these TLVs is defined per "Relay Algorithm 1063 Type" and may contain algorithm specific information. 1065 +---------------+----------------+--------------------+ 1066 | | TLV Syntax | Field Values | 1067 +---------------+----------------+--------------------+ 1068 | type | | SMF_TYPE | 1069 | extended type | | | 1070 | length | | variable | 1071 | value | | variable | 1072 +---------------+----------------+--------------------+ 1074 Table 6: SMF Type Message TLV 1076 In Table 6 is an 8-bit field containing a number 1077 0-255 representing the "Relay Algorithm Type" of the originator 1078 address of the corresponding NHDP message. 1080 Values for the are defined in Table 7. The table 1081 provides value assignments, future IANA assignment spaces, and an 1082 experimental space. The experimental space use MUST NOT assume 1083 uniqueness and thus SHOULD NOT be used for general interoperable 1084 deployment prior to official IANA assignment. 1086 +-------------+--------------------+--------------------------------+ 1087 | Type Value | Extended Type | Algorithm | 1088 | | Value | | 1089 +-------------+--------------------+--------------------------------+ 1090 | SMF_TYPE | 0 | CF | 1091 | SMF_TYPE | 1 | S-MPR | 1092 | SMF_TYPE | 2 | E-CDS | 1093 | SMF_TYPE | 3 | MPR-CDS | 1094 | SMF_TYPE | 4-127 | Future Assignment STD action | 1095 | SMF_TYPE | 128-239 | No STD action required | 1096 | SMF_TYPE | 240-255 | Experimental Space | 1097 +-------------+--------------------+--------------------------------+ 1099 Table 7: SMF Relay Algorithm Type Values 1101 Acceptable and fields of an SMF_TYPE TLV are 1102 dependent on the extended type value (i.e. relay algorithm type). 1103 The appropriate algorithm type, as conveyed in the 1104 field, defines the meaning and format of its TLV field. For 1105 the algorithms defined by this document, see the appropriate appendix 1106 for the field format. 1108 8.1.2. SMF Address Block TLV Type 1110 An address block TLV type, denoted SMF_NBR_TYPE (i.e., SMF neighbor 1111 relay algorithm) is specified in Table 8. This TLV enables CDS relay 1112 algorithm operation and configuration to be shared among 2-hop 1113 neighborhoods. Some relay algorithms require two hop neighbor 1114 configuration in order to correctly select relay sets. It is also 1115 useful when mixed relay algorithm operation is possible, some 1116 examples of mixed use are outlined in the Appendices. 1118 The message SMF_TYPE TLV and address block SMF_NBR_TYPE TLV types 1119 share a common format. 1121 +---------------+----------------+--------------------+ 1122 | | TLV syntax | Field Values | 1123 +---------------+----------------+--------------------+ 1124 | type | | SMF_NBR_TYPE | 1125 | extended type | | | 1126 | length | | variable | 1127 | value | | variable | 1128 +---------------+----------------+--------------------+ 1130 Table 8: SMF Type Address Block TLV 1132 in Table 8 is an 8-bit unsigned integer field 1133 containing a number 0-255 representing the "Relay Algorithm Type" 1134 value that corresponds to any associated address in the address 1135 block. Note that "Relay Algorithm Type" values for 2-hop neighbors 1136 can be conveyed in a single TLV or multiple value TLVs as described 1137 in [RFC5444]. It is expected that SMF routers using NHDP construct 1138 address blocks with SMF_NBR_TYPE TLVs to advertise "Relay Algorithm 1139 Type" and to advertise neighbor algorithm values received in SMF_TYPE 1140 TLVs from those neighbors. 1142 Again values for the are defined in Table 7. 1144 The interpretation of the "value" field of SMF_NBR_TYPE TLVs is 1145 defined per "Relay Algorithm Type" and may contain algorithm specific 1146 information. See the appropriate appendix for definitions of value 1147 fields for the algorithms defined by this document. 1149 9. SMF Border Gateway Considerations 1151 It is expected that SMF will be used to provide simple forwarding of 1152 multicast traffic within a MANET or mesh routing topology. A border 1153 router gateway approach should be used to allow interconnection of 1154 SMF routing domains with networks using other multicast routing 1155 protocols, such as PIM. It is important to note that there are many 1156 scenario-specific issues that should be addressed when discussing 1157 border multicast routers. At the present time, experimental 1158 deployments of SMF and PIM border router approaches have been 1159 demonstrated [DHS08]. Some of the functionality border routers may 1160 need to address includes the following: 1162 1. Determining which multicast group traffic transits the border 1163 router whether entering or exiting the attached SMF routing 1164 domain. 1165 2. Enforcement of TTL/hop-limit threshold or other scoping policies. 1166 3. Any marking or labeling to enable DPD on ingressing packets. 1167 4. Interface with exterior multicast routing protocols. 1168 5. Possible operation with multiple border routers (presently beyond 1169 scope of this document). 1170 6. Provisions for participating non-SMF devices (routers or hosts). 1172 Each of these areas is discussed in more detail in the following 1173 subsections. Note the behavior of SMF border routers is the same as 1174 that of non-border SMF routers when forwarding packets on interfaces 1175 within the SMF routing domain. Packets that are passed outbound to 1176 interfaces operating fixed-infrastructure multicast routing protocols 1177 SHOULD be evaluated for duplicate packet status since present 1178 standard multicast forwarding mechanisms do not usually perform this 1179 function. 1181 9.1. Forwarded Multicast Groups 1183 Mechanisms for dynamically determining groups for forwarding into a 1184 MANET SMF routing domain is an evolving technology area. Ideally, 1185 only traffic for which there is active group membership should be 1186 injected into the SMF domain. This can be accomplished by providing 1187 an IPv4 Internet Group Membership Protocol (IGMP) or IPv6 Multicast 1188 Listener Discovery (MLD) proxy protocol so that MANET SMF routers can 1189 inform attached border routers (and hence multicast networks) of 1190 their current group membership status. For specific systems and 1191 services it may be possible to statically configure group membership 1192 joins in border routers, but it is RECOMMENDED that some form of 1193 IGMP/MLD proxy or other explicit, dynamic control of membership be 1194 provided. Specification of such an IGMP/MLD proxy protocol is beyond 1195 the scope of this document. 1197 For outbound traffic, SMF border routers perform duplicate packet 1198 detection and forward non-duplicate traffic that meets TTL/hop limit 1199 and scoping criteria to interfaces external to the SMF routing 1200 domain. Appropriate IP multicast routing (e.g., PIM-based solutions) 1201 on those interfaces can make further forwarding decisions with 1202 respect to the multicast packet. Note that the presence of multiple 1203 border routers associated with a MANET routing domain raises 1204 additional issues. This is further discussed in Section 9.4 but 1205 further work is expected to be needed here. 1207 9.2. Multicast Group Scoping 1209 Multicast scoping is used by network administrators to control the 1210 network routing domains reachable by multicast packets. This is 1211 usually done by configuring external interfaces of border routers in 1212 the border of a routing domain to not forward multicast packets which 1213 must be kept within the SMF routing domain. This is commonly done 1214 based on TTL/hop-limit of messages or the basis of group addresses. 1215 These schemes are known respectively as: 1217 1. TTL scoping. 1218 2. Administrative scoping. 1220 For IPv4, network administrators can configure border routers with 1221 the appropriate TTL/hop-limit thresholds or administratively scoped 1222 multicast groups for the router interfaces as with any traditional 1223 multicast router. However, for the case of TTL/hop-limit scoping it 1224 SHOULD be taken into account that the packet could traverse multiple 1225 hops within the MANET SMF routing domain before reaching the border 1226 router. Thus, TTL thresholds SHOULD be selected carefully. 1228 For IPv6, multicast address spaces include information about the 1229 scope of the group. Thus, border routers of an SMF routing domain 1230 know if they must forward a packet based on the IPv6 multicast group 1231 address. For the case of IPv6, it is RECOMMENDED that a MANET SMF 1232 routing domain be designated a site-scoped multicast domain. Thus, 1233 all IPv6 site-scoped multicast packets in the range FF05::/16 SHOULD 1234 be kept within the MANET SMF routing domain by border routers. IPv6 1235 packets in any other wider range scopes (i.e. FF08::/16, FF0B::/16 1236 and FF0E::16) MAY traverse border routers unless other restrictions 1237 different from the scope applies. 1239 Given that scoping of multicast packets is performed at the border 1240 routers, and given that existing scoping mechanisms are not designed 1241 to work with mobile routers, it is assumed that non-border routers 1242 running SMF will not stop forwarding multicast data packets of an 1243 appropriate site scoping. That is, it is assumed that an SMF routing 1244 domain is a site-scoped multicast area. 1246 9.3. Interface with Exterior Multicast Routing Protocols 1248 The traditional operation of multicast routing protocols is tightly 1249 integrated with the group membership function. Leaf routers are 1250 configured to periodically gather group membership information, while 1251 intermediate routers conspire to create multicast trees connecting 1252 routers with directly-connected multicast sources and routers with 1253 active multicast receivers. In the concrete case of SMF, border 1254 routers can be considered leaf routers. Mechanisms for multicast 1255 sources and receivers to interoperate with border routers over the 1256 multihop MANET SMF routing domain as if they were directly connected 1257 to the router need to be defined. The following issues need to be 1258 addressed: 1260 1. A mechanism by which border routers gather membership information 1261 2. A mechanism by which multicast sources are known by the border 1262 router 1263 3. A mechanism for exchange of exterior routing protocol messages 1264 across the SMF routing domain if the SMF routing domain is to 1265 provide transit connectivity for multicast traffic. 1267 It is beyond the scope of this document to address implementation 1268 solutions to these issues. As described in Section 9.1, IGMP/MLD 1269 proxy mechanisms can address some of these issues. Similarly, 1270 exterior routing protocol messages could be tunneled or conveyed 1271 across an SMF routing domain but doing this robustly in a distributed 1272 wireless environment likely requires additional considerations 1273 outside the scope of this document. 1275 The need for the border router to receive traffic from recognized 1276 multicast sources within the SMF routing domain is important to 1277 potentially achieve interoperability with existing routing protocols. 1278 For instance, PIM-S requires routers with locally attached multicast 1279 sources to register them to the Rendezvous Point (RP) so that routers 1280 can join the multicast tree. In addition, if those sources are not 1281 advertised to other autonomous systems (AS) using Multicast Source 1282 Discovery Protocol (MSDP), receivers in those external networks are 1283 not able to join the multicast tree for that source. 1285 9.4. Multiple Border Routers 1287 An SMF routing domain might be deployed with multiple participating 1288 routers having connectivity to external, fixed-infrastructure 1289 networks. Allowing multiple routers to forward multicast traffic to/ 1290 from the SMF routing domain can be beneficial since it can increase 1291 reliability, and provide better service. For example, if the SMF 1292 routing domain were to fragment with different SMF routers 1293 maintaining connectivity to different border routers, multicast 1294 service could still continue successfully. But, the case of multiple 1295 border routers connecting a SMF routing domain to external networks 1296 presents several challenges for SMF: 1298 1. Handling duplicate unmarked IPv4 or IPv6 (without IPsec 1299 encapsulation or DPD option) packets possibly injected by 1300 multiple border routers. 1301 2. Source-based relay algorithms handling of duplicate traffic 1302 injected by multiple border routers. 1303 3. Determination of which border router(s) will forward outbound 1304 multicast traffic. 1305 4. Additional challenges with interfaces to exterior multicast 1306 routing protocols. 1308 When multiple border routers are present they may be alternatively 1309 (due to route changes) or simultaneously injecting common traffic 1310 into the SMF routing domain that has not been previously marked for 1311 SMF_DPD. Different border routers would not be able to implicitly 1312 synchronize sequencing of injected traffic since they may not receive 1313 exactly the same messages due to packet losses. For IPv6 I-DPD 1314 operation, the optional "TaggerId" field described for the SMF_DPD 1315 header option can be used to mitigate this issue. When multiple 1316 border routers are injecting a flow into a SMF routing domain, there 1317 are two forwarding policies that SMF routers running I-DPD may 1318 implement: 1320 1. Redundantly forward the multicast flows (identified by ) from each border router, performing DPD processing on a 1322 or basis, or 1323 2. Use some basis to select the flow of one tagger (border router) 1324 over the others and forward packets for applicable flows 1325 (identified by ) only for the selected 1326 "Tagger ID" until timeout or some other criteria to favor another 1327 tagger occurs. 1329 It is RECOMMENDED that the first approach be used in the case of 1330 I-DPD operation. Additional specification may be required to 1331 describe an interoperable forwarding policy based on this second 1332 option. Note that the implementation of the second option requires 1333 that per-flow (i.e., ) state be maintained for the 1334 selected "Tagger ID". 1336 The deployment of H-DPD operation may alleviate DPD resolution when 1337 ingressing traffic comes from multiple border routers. Non-colliding 1338 hash indexes (those not requiring the H-DPD options header in IPv6) 1339 should be resolved effectively. 1341 10. Security Considerations 1343 Gratuitous use of option headers can cause problems in routers. 1344 Other IP routers external to an SMF routing domain that might receive 1345 forwarded multicast SHOULD ignore SMF-specific IPv6 header options 1346 when encountered. The header options types are encoded appropriately 1347 to allow for this behavior. 1349 This section briefly discusses several SMF denial-of-service (DoS) 1350 attack scenarios and we provide some initial recommended mitigation 1351 strategies. 1353 A potential denial-of-service attack against SMF forwarding is 1354 possible when a malicious router has a form of wormhole access to 1355 non-adjacent parts of a network topology. In the wireless ad hoc 1356 case, a directional antenna is one way to provide such a wormhole 1357 physically. If such a router can preview forwarded packets in a non- 1358 adjacent part of the network and forward modified versions to another 1359 part of the network it can perform the following attack. The 1360 malicious router could reduce the TTL/hop-limit or Hop Limit of the 1361 packet and transmit it to the SMF router causing it to forward the 1362 packet with a limited TTL/hop-limit (or even drop it) and make a DPD 1363 entry that could block or limit the subsequent forwarding of later- 1364 arriving valid packets with correct TTL/hop-limit values. This would 1365 be a relatively low-cost, high-payoff attack that would be hard to 1366 detect and thus attractive to potential attackers. An approach of 1367 caching TTL/hop-limit information with DPD state and taking 1368 appropriate forwarding actions is identified in Section 5 to mitigate 1369 this form of attack. 1371 Sequence-based packet identifiers are predictable and thus provide an 1372 opportunity for a DoS attack against forwarding. Forwarding 1373 protocols that use DPD techniques, such as SMF, may be vulnerable to 1374 DoS attacks based on spoofing packets with apparently valid packet 1375 identifier fields. In wireless environments, where SMF will most 1376 likely be used, the opportunity for such attacks may be more 1377 prevalent than in wired networks. In the case of IPv4 packets, 1378 fragmented IP packets or packets with IPsec headers applied, the DPD 1379 "identifier portions" of potential future packets that might be 1380 forwarded is highly predictable and easily subject to DoS attacks 1381 against forwarding. A RECOMMENDED technique to counter this concern 1382 is for SMF implementations to generate an "internal" hash value that 1383 is concatenated with the explicit I-DPD packet identifier to form a 1384 unique identifier that is a function of the packet content as well as 1385 the visible identifier. SMF implementations could seed their hash 1386 generation with a random value to make it unlikely that an external 1387 observer could guess how to spoof packets used in a denial-of-service 1388 attack against forwarding. Since the hash computation and state is 1389 kept completely internal to SMF routers, the cryptographic properties 1390 of this hashing would not need to be extensive and thus possibly of 1391 low complexity. Experimental implementations may determine that a 1392 lightweight hash of even only portions of packets may suffice to 1393 serve this purpose. 1395 While H-DPD is not as readily susceptible to this form of DoS attack, 1396 it is possible that a sophisticated adversary could use side 1397 information to construct spoofing packets to mislead forwarders using 1398 a well-known hash algorithm. Thus, similarly, a separate "internal" 1399 hash value could be concatenated with the well-known hash value to 1400 alleviate this security concern. 1402 The support of forwarding IPsec packets without further modification 1403 for both IPv4 and IPv6 is supported by this specification. 1405 Authentication mechanisms to identify the source of IPv6 option 1406 headers should be considered to reduce vulnerability to a variety of 1407 attacks. 1409 Furthermore, when the MANET Neighborhood Discovery Protocol [RFC6130] 1410 is used, the security considerations described there also applies. 1412 11. IANA Considerations 1414 This document defines one IPv6 Hop-by-Hop Option, a type for which 1415 which must be allocated from the IPv6 "Destination Options and Hop- 1416 by-Hop Options" registry of [RFC2780]. 1418 This document creates one registry for recording TaggerId types, 1419 (TidTy). 1421 This document requests registration of one well-known multicast 1422 address from each of the IPv4 and IPv6 multicast address spaces. 1424 This document defines one Message TLV, a type for which must be 1425 allocated from the "Message TLV Types" registry of [RFC5444]. 1427 Finally, this document defines one Address Block TLV, a type for 1428 which must be allocated from the "Address Block TLV Types" registry 1429 of [RFC5444]. 1431 11.1. IPv6 SMF_DPD Header Extension Option Type 1433 IANA is requested to allocate an IPv6 Option Type from the IPv6 1434 "Destination Options and Hop-by-Hop Options" registry of [RFC2780], 1435 as specified in Table 9. 1437 +----------+-----+-----+------+-------------------------+-----------+ 1438 | Mnemonic | act | chg | rest | Description | Reference | 1439 +----------+-----+-----+------+-------------------------+-----------+ 1440 | SMF_DPD | 00 | 0 | TBD | Multicast Duplicate | This | 1441 | | | | | Packet Detection (DPD) | Document | 1442 +----------+-----+-----+------+-------------------------+-----------+ 1444 Table 9: IPv6 Option Type Allocation 1446 11.2. TaggerId Types (TidTy) 1448 A portion of the option data content in the SMF_DPD is the Taggger 1449 Identifier Type (TidTy), that provides a context for the optionally 1450 included "TaggerId". 1452 IANA is requested to create a registry for recording TaggerId Types 1453 (TidTy), with initial assignments and allocation policies, as 1454 specified in Table 10. 1456 +----------+------+-------------------------------+-----------------+ 1457 | Mnemonic | Type | Description | Reference | 1458 +----------+------+-------------------------------+-----------------+ 1459 | NULL | 0 | No "TaggerId" field is | This document | 1460 | | | present | | 1461 | DEFAULT | 1 | A "TaggerId" of non-specific | This document | 1462 | | | context is present | | 1463 | IPv4 | 2 | A "TaggerId" representing an | This document | 1464 | | | IPv4 address is present | | 1465 | IPv6 | 3 | A "TaggerId" representing an | This document | 1466 | | | IPv6 address is present | | 1467 | | 4-6 | | Unassigned | 1468 | | | | (IETF Review) | 1469 | ExtId | 7 | | Unassigned | 1470 | | | | (Expert Review) | 1471 +----------+------+-------------------------------+-----------------+ 1473 Table 10: TaggerId Types 1475 For allocation of unassigned values 4-6, IETF Review is required. 1477 For allocation of unassigned value 7, Expert Review is required, with 1478 the following specific guidelines to the Expert: this value is 1479 intended for use in case a future development of this specification 1480 requires extending the type space (e.g. by way of providing a pointer 1481 to a subsequent field). 1483 11.3. Well-known Multicast Address 1485 IANA is requested to allocate an IPv4 multicast address "SL-MANET- 1486 ROUTERS" from the "Internetwork Control Block (224.0.1/24)" sub- 1487 registry of the "IPv4 Multicast Address" registry. 1489 IANA is requested to allocate an IPv6 multicast address "SL-MANET- 1490 ROUTERS" from the "Site-Local Scope Multicast Addresses" sub-sub- 1491 registry of the "Fixed Scope Multicast Addresses" sub-registry of the 1492 "INTERNET PROTOCOL VERSION 6 MULTICAST ADDRESSES" registry. 1494 11.4. SMF Type-Length-Values 1496 11.4.1. Expert Review for created Type Extension Registries 1498 Creation of Address Block TLV Types and Message TLV Types in 1499 registries of [RFC5444], and hence in the HELLO message specific 1500 registries of [RFC6130], entails creation of corresponding Type 1501 Extension registries for each such type. For such Type Extension 1502 registries, where an Expert Review is required, the designated expert 1503 SHOULD take the same general recommendations into consideration as 1504 are specified by [RFC5444]. 1506 11.4.2. SMF Message TLV Type (SMF_TYPE) 1508 This document defines one Message TLV Type, "SMF_TYPE", which must be 1509 allocated from the "HELLO Message-Type-specific Message TLV Types" 1510 registry, defined in [RFC6130]. 1512 This will create a new Type Extension registry, with initial 1513 assignments as specified in Table 11. 1515 +----------+------+-----------+--------------------+----------------+ 1516 | Name | Type | Type | Description | Allocation | 1517 | | | Extension | | Policy | 1518 +----------+------+-----------+--------------------+----------------+ 1519 | SMF_TYPE | TBD2 | 0-255 | Specifies relay | Section 11.4.4 | 1520 | | | | algorithm | | 1521 | | | | supported by the | | 1522 | | | | SMF router, | | 1523 | | | | originating the | | 1524 | | | | HELLO message, | | 1525 | | | | according to | | 1526 | | | | Section 11.4.4. | | 1527 +----------+------+-----------+--------------------+----------------+ 1529 Table 11: SMF_TYPE Message TLV Type Extension Registry 1531 11.4.3. SMF Address Block TLV Type (SMF_NBR_TYPE) 1533 This document defines one Address Block TLV Type, "SMF_NBR_TYPE", 1534 which must be allocated from the "HELLO Message-Type-specific Address 1535 Block TLV Types" registry, defined in [RFC6130]. 1537 This will create a new Type Extension registry, with initial 1538 assignments as specified in Table 12. 1540 +----------+------+-----------+--------------------+----------------+ 1541 | Name | Type | Type | Description | Allocation | 1542 | | | Extension | | Policy | 1543 +----------+------+-----------+--------------------+----------------+ 1544 | SMF_TYPE | TBD2 | 0-255 | Specifies relay | Section 11.4.4 | 1545 | | | | algorithm | | 1546 | | | | supported by the | | 1547 | | | | SMF router | | 1548 | | | | corresponding to | | 1549 | | | | the advertised | | 1550 | | | | address, according | | 1551 | | | | to Section 11.4.4. | | 1552 +----------+------+-----------+--------------------+----------------+ 1554 Table 12: SMF_NBR_TYPE Address Block TLV Type Extension Registry 1556 11.4.4. SMF Relay Algorithm ID Registry 1558 Types for the Type Extension Registries for the SMF_TYPE Message TLV 1559 and the SMF_NBR_TYPE Address Block TLV are unified in this single SMF 1560 Relay Algorithm ID Registry, defined in this section. 1562 IANA is requested to create a registry for recording Relay Algorithm 1563 Identifiers, with initial assignments and allocation policies as 1564 specified in Table 13. 1566 +---------+---------+-------------+-------------------+ 1567 | Name | Value | Description | Allocation Policy | 1568 +---------+---------+-------------+-------------------+ 1569 | CF | 0 | | | 1570 | S-MPR | 1 | Appendix B | | 1571 | E-CDS | 2 | Appendix A | | 1572 | MPR-CDS | 3 | Appendix C | | 1573 | | 4-127 | Unassigned | Expert Review | 1574 | | 128-255 | Unassigned | Experimental Use | 1575 +---------+---------+-------------+-------------------+ 1577 Table 13: Relay Set Algorithm Type Values 1579 A specification requestion an allocation from the 4-127 range from 1580 the SMF Relay Algorithm ID Registry MUST specify the interpretation 1581 of the field (if any). 1583 12. Acknowledgments 1585 Many of the concepts and mechanisms used and adopted by SMF resulted 1586 over several years of discussion and related work within the MANET 1587 working group since the late 1990s. There are obviously many 1588 contributors to past discussions and related draft documents within 1589 the working group that have influenced the development of SMF 1590 concepts and they deserve acknowledgment. In particular, the 1591 document is largely a direct product of the earlier SMF design team 1592 within the IETF MANET working group and borrows text and 1593 implementation ideas from the related individuals and activities. 1594 Some of the direct contributors who have been involved in design, 1595 content editing, prototype implementation, major commenting, and core 1596 discussions are listed below in alphabetical order. We appreciate 1597 all the input and feedback from the many community members and early 1598 implementation users we have heard from that are not on this list as 1599 well. 1601 Some core contributors/authors in alphabetical order: 1602 Brian Adamson 1603 Teco Boot 1604 Ian Chakeres 1605 Thomas Clausen 1606 Justin Dean 1607 Brian Haberman 1608 Ulrich Herberg 1609 Charles Perkins 1610 Pedro Ruiz 1611 Fred Templin 1612 Maoyu Wang 1614 The RFC text was produced using Marshall Rose's xml2rfc tool and Bill 1615 Fenner's XMLmind add-ons. 1617 13. References 1619 13.1. Normative References 1621 [MPR-CDS] Adjih, C., Jacquet, P., and L. Viennot, "Computing 1622 Connected Dominating Sets with Multipoint Relays", Ad Hoc 1623 and Sensor Wireless Networks , January 2005. 1625 [RFC0791] Postel, J., "Internet Protocol", STD 5, RFC 791, 1626 September 1981. 1628 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 1629 Requirement Levels", BCP 14, RFC 2119, March 1997. 1631 [RFC2460] Deering, S. and R. Hinden, "Internet Protocol, Version 6 1632 (IPv6) Specification", RFC 2460, December 1998. 1634 [RFC2644] Senie, D., "Changing the Default for Directed Broadcasts 1635 in Routers", BCP 34, RFC 2644, August 1999. 1637 [RFC2780] Bradner, S., "IANA Allocation Guidelines For Values In the 1638 Internet Protocol and Related Headers", RFC 2780, 1639 March 2000. 1641 [RFC3174] Eastlake, D. and P. Jones, "US Secure Hash Algorithm 1 1642 (SHA1)", RFC 3174, September 2001. 1644 [RFC3626] Clausen, T. and P. Jacquet, "Optimized Link State Routing 1645 Protocol", RFC 3626, 2003. 1647 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 1648 Architecture", RFC 4291, February 2006. 1650 [RFC4302] Kent, S., "IP Authentication Header", RFC 4302, 1651 December 2005. 1653 [RFC5444] Clausen, T., Adjih, C., Dearlove, C., and J. Dean, 1654 "Generalized MANET Packet/Message Format", RFC 5444, 1655 February 2009. 1657 [RFC5614] Ogier, R. and P. Spagnolo, "Mobile Ad Hoc Network (MANET) 1658 Extension of OSPF Using Connected Dominating Set (CDS) 1659 Flooding", RFC 5614, August 2009. 1661 [RFC5771] Cotton, M., Vegoda, L., and D. Meyer, "IANA Guidelines for 1662 IPv4 Multicast Address Assignment", RFC 5771, March 2010. 1664 [RFC6130] Clausen, T., Dearlove, C., and J. Dean, "Mobile Ad Hoc 1665 Network (MANET) Neighborhood Discovery Protocol (NHDP)", 1666 RFC 6130, March 2011. 1668 13.2. Informative References 1670 [CDHM07] Chakeres, I., Danilov, C., and T. Henderson, "Connecting 1671 MANET Multicast", IEEE MILCOM 2007 Proceedings , 2007. 1673 [DHG09] Danilov, C., Henderson, T., and T. Goff, "Experiment and 1674 field demonstration of a 802.11-based ground-UAV mobile 1675 ad-hoc network", Proceedings of the 28th IEEE conference 1676 on Military Communications , 2009. 1678 [DHS08] Danilov, C., Henderson, T., and T. Spagnolo, "MANET 1679 Multicast with Multiple Gateways", IEEE MILCOM 2008 1680 Proceedings , 2008. 1682 [GM99] Garcia-Luna-Aceves, JJ. and E. Madruga, "The core-assisted 1683 mesh protocol", Selected Areas in Communications, IEEE 1684 Journal on Volume 17, Issue 8, August 1999. 1686 [JLMV02] Jacquet, P., Laouiti, V., Minet, P., and L. Viennot, 1687 "Performance of multipoint relaying in ad hoc mobile 1688 routing protocols", Networking , 2002. 1690 [MDC04] Macker, J., Dean, J., and W. Chao, "Simplified Multicast 1691 Forwarding in Mobile Ad hoc Networks", IEEE MILCOM 2004 1692 Proceedings , 2004. 1694 [MDDA07] Macker, J., Downard, I., Dean, J., and R. Adamson, 1695 "Evaluation of distributed cover set algorithms in mobile 1696 ad hoc network for simplified multicast forwarding", ACM 1697 SIGMOBILE Mobile Computing and Communications Review 1698 Volume 11 , Issue 3, July 2007. 1700 [MGL04] Mohapatra, P., Gui, C., and J. Li, "Group Communications 1701 in Mobile Ad hoc Networks", IEEE Computer Vol. 37, No. 2, 1702 February 2004. 1704 [NTSC99] Ni, S., Tseng, Y., Chen, Y., and J. Sheu, "The Broadcast 1705 Storm Problem in Mobile Ad hoc Networks", Proceedings Of 1706 ACM Mobicom 99 , 1999. 1708 [RFC2501] Macker, JP. and MS. Corson, "Mobile Ad hoc Networking 1709 (MANET): Routing Protocol Performance Issues and 1710 Evaluation Considerations", RFC 2501, January 1999. 1712 [RFC3684] Ogier, R., Templin, F., and M. Lewis, "Topology 1713 Dissemination Based on Reverse-Path Forwarding", RFC 3684, 1714 February 2004. 1716 [RFC3973] Adams, A., Nicholas, J., and W. Siadak, "Protocol 1717 Independent Multicast - Dense Mode (PIM-DM): Protocol 1718 Specification (Revised)", RFC 3973, January 2005. 1720 [RFC4601] Fenner, B., Handley, M., Holbrook, H., and I. Kouvelas, 1721 "Protocol Independent Multicast - Sparse Mode (PIM-SM): 1722 Protocol Specification (Revised)", RFC 4601, August 2006. 1724 Appendix A. Essential Connecting Dominating Set (E-CDS) Algorithm 1726 The "Essential Connected Dominating Set" (E-CDS) algorithm [RFC5614] 1727 forms a single CDS mesh for the SMF operating region. It allows 1728 routers to use 2-hop neighborhood topology information to dynamically 1729 perform relay self election to form a CDS. Its packet forwarding 1730 rules are not dependent upon previous hop knowledge. Additionally, 1731 E-CDS SMF forwarders can be easily mixed without problems with CF SMF 1732 forwarders, even those not participating in NHDP. Another benefit is 1733 that packets opportunistically received from non-symmetric neighbors 1734 may be forwarded without compromising flooding efficiency or 1735 correctness. Furthermore, multicast sources not participating in 1736 NHDP may freely inject their traffic and any neighboring E-CDS relays 1737 will properly forward the traffic. The E-CDS based relay set 1738 selection algorithm is based upon [RFC5614]. E-CDS was originally 1739 discussed in the context of forming partial adjacencies and efficient 1740 flooding for MANET OSPF extensions work and the core algorithm is 1741 applied here for SMF. 1743 It is RECOMMENDED that the SMF_TYPE:E-CDS Message TLV be included in 1744 NHDP_HELLO messages that are generated by routers conducting E-CDS 1745 SMF operation. It is also RECOMMENDED that the SMF_NBR_TYPE:E-CDS 1746 address block TLV be used to advertise neighbor routers that are also 1747 conducting E-CDS SMF operation. 1749 A.1. E-CDS Relay Set Selection Overview 1751 The E-CDS relay set selection requires 2-hop neighborhood information 1752 collected through NHDP or another process. Relay routers, in E-CDS 1753 SMF selection, are "self-elected" using a router identifier (Router 1754 ID) and an optional nodal metric, referred to here as "Router 1755 Priority" for all 1-hop and 2-hop neighbors. To ensure proper relay 1756 set self-election, the Router ID and Router Priority MUST be 1757 consistent among participating routers. It is RECOMMENDED that NHDP 1758 be used to share Router ID and Router Priority through the use of 1759 SMF_TYPE:E-CDS TLVs as described in this appendix. The Router ID is 1760 a logical identification that MUST be consistent across 1761 interoperating SMF neighborhoods and it is RECOMMENDED to be chosen 1762 as the numerically largest address contained in a routers "Neighbor 1763 Address List" as defined in NHDP. The E-CDS self-election process 1764 can be summarized as follows: 1766 1. If an SMF router has a higher ordinal (Router Priority, Router 1767 ID) than all of its symmetric neighbors, it elects itself to act 1768 as a forwarder for all received multicast packets, 1769 2. Else, if there does not exist a path from the neighbor with 1770 largest (Router Priority, Router ID) to any other neighbor, via 1771 neighbors with larger values of (Router Priority, Router ID), 1772 then it elects itself to the relay set. 1774 The basic form of E-CDS described and applied within this 1775 specification does not provide for redundant relay set election 1776 (e.g., bi-connected) but such capability is supported by the basic 1777 E-CDS design. 1779 A.2. E-CDS Forwarding Rules 1781 With E-CDS, any SMF router that has selected itself as a relay 1782 performs DPD and forwards all non-duplicative multicast traffic 1783 allowed by the present forwarding policy. Packet previous hop 1784 knowledge is not needed for forwarding decisions when using E-CDS. 1786 1. Upon packet reception, DPD is performed. Note E-CDS requires a 1787 single duplicate table for the set of interfaces associated with 1788 the relay set selection. 1789 2. If the packet is a duplicate, no further action is taken. 1790 3. If the packet is non-duplicative: 1791 A. A DPD entry is made for the packet identifier 1792 B. The packet is forwarded out all interfaces associated with 1793 the relay set selection 1795 As previously mentioned, even packets sourced (or relayed) by routers 1796 not participating in NHDP and/or the E-CDS relay set selection may be 1797 forwarded by E-CDS forwarders without problem. A particular 1798 deployment MAY choose to not forward packets from previous hop 1799 routers that have been not explicitly identified via NHDP or other 1800 means as operating as part of a different relay set algorithm (e.g. 1801 S-MPR) to allow coexistent deployments to operate correctly. Also, 1802 E-CDS relay set selection may be configured to be influenced by 1803 statically-configured CF relays that are identified via NHDP or other 1804 means. 1806 A.3. E-CDS Neighborhood Discovery Requirements 1808 It is possible to perform E-CDS relay set selection without 1809 modification of NHDP, basing the self-election process exclusively on 1810 the "Neighbor Address List" of participating SMF routers. For 1811 example by setting the "Router Priority" to a default value and 1812 selecting the "Router ID" as the numerically largest address 1813 contained in the "Neighbor Address List". However steps MUST be 1814 taken to insure that all NHDP enabled routers not using SMF_TYPE:E- 1815 CDS full type Message TLVs are in fact running SMF E-CDS with the 1816 same methods for selecting "Router Priority" and "Router ID", 1817 otherwise incorrect forwarding may occur. Note that SMF routers with 1818 higher "Router Priority" values will be favored as relays over 1819 routers with lower "Router Priority". Thus, preferred relays MAY be 1820 administratively configured to be selected when possible. 1821 Additionally, other metrics (e.g. nodal degree, energy capacity, etc) 1822 may also be taken into account in constructing a "Router Priority" 1823 value. When using "Router Priority" with multiple interfaces all 1824 interfaces on a router MUST use and advertise a common "Router 1825 Priority" value. A routers "Router Priority" value may be 1826 administratively or algorithmically selected. The method of 1827 selection does not need to be the same among different routers. 1829 E-CDS relay set selection may be configured to be influenced by 1830 statically configured CF relays that are identified via NHDP or other 1831 means. Nodes advertising CF through NHDP may be considered E-CDS SMF 1832 routers with maximal "Router Priority". 1834 To share a router's "Router Priority" with its 1-hop neighbors the 1835 SMF_TYPE:E-CDS Message TLV's field is defined as shown in 1836 Table 14. 1838 +---------------+---------+-----------------+ 1839 | Length(bytes) | Value | Router Priority | 1840 +---------------+---------+-----------------+ 1841 | 0 | N/A | 64 | 1842 | 1 | | 0-127 | 1843 +---------------+---------+-----------------+ 1845 Table 14: E-CDS Message TLV Values 1847 Where is a one octet long bit field which is defined as: 1849 bit 0: the leftmost bit is reserved and SHOULD be set to 0. 1851 bit 1-7: contain the unsigned "Router Priority" value, 0-127, which 1852 is associated with the "Neighbor Address List". 1854 Combinations of value field lengths and values other than specified 1855 here are NOT permitted and SHOULD be ignored. Figure 5 shows an 1856 example SMF_TYPE:E-CDS Message TLV 1857 0 1 2 3 1858 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 1859 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1860 ... | SMF_TYPE |1|0|0|1|0|0| | 1861 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1862 | E-CDS |0|0|0|0|0|0|0|1|R| priority | ... | 1863 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1865 Figure 5: E-CDS Message TLV Example 1867 To convey "Router Priority" values among 2-hop neighborhoods the 1868 SMF_NBR_TYPE:E-CDS address block TLV's field is used. Multi- 1869 index and multi-value TLV layouts as defined in [RFC5444] are 1870 supported. SMF_NBR_TYPE:E-CDS value fields are defined thus: 1872 +---------------+--------+----------+-------------------------------+ 1873 | Length(bytes) | # Addr | Value | Router Priority | 1874 +---------------+--------+----------+-------------------------------+ 1875 | 0 | Any | N/A | 64 | 1876 | 1 | Any | | is for all addresses | 1877 | N | N | * | Each address gets its own | 1878 | | | | | 1879 +---------------+--------+----------+-------------------------------+ 1881 Table 15: E-CDS Address Block TLV Values 1883 Where is a one byte bit field which is defined as: 1885 bit 0: the leftmost bit is reserved and SHOULD be set to 0. 1887 bit 1-7: contain the unsigned "Router Priority" value, 0-127, which 1888 is associated with the appropriate address(es). 1890 Combinations of value field lengths and # of addresses other than 1891 specified here are NOT permitted and SHOULD be ignored. A default 1892 technique of using nodal degree (i.e. count of 1-hop neighbors) is 1893 RECOMMENDED for the value field of these TLV types. Below are two 1894 example SMF_NBR_TYPE:E-CDS address block TLVs. 1896 0 1 2 3 1897 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 1898 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1899 ... | SMF_NBR_TYPE |1|0|0|1|0|0| | 1900 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1901 | E-CDS |0|0|0|0|0|0|0|1|R| priority | ... | 1902 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1904 Figure 6: E-CDS Address Block TLV Example 1 1906 The single value example TLV, depicted in Figure 6 , specifies that 1907 all address(es) contained in the address block are running SMF using 1908 the E-CDS algorithm and all address(es) share the value field and 1909 therefore the same "Router Priority". 1910 0 1 2 3 1911 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 1912 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1913 ... | SMF_NBR_TYPE |1|0|1|1|0|1| | 1914 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1915 | E-CDS | index-start | index-end | length | 1916 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1917 |R| priority0 |R| priority1 | ... |R| priorityN | 1918 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 1920 Figure 7: E-CDS Address Block TLV Example 2 1922 The example multivalued TLV, depicted in Figure 7, specifies that 1923 address(es) contained in the address block from index-start to index- 1924 end inclusive are running SMF using the E-CDS algorithm. Each 1925 address is associated with its own value byte and therefore its own 1926 "Router Priority". 1928 A.4. E-CDS Selection Algorithm 1930 This section describes an algorithm for E-CDS relay selection (self- 1931 election). The algorithm described uses 2-hop information. Note it 1932 is possible to extend this algorithm to use k-hop information with 1933 added computational complexity and mechanisms for sharing k-hop 1934 topology information that are not described in this document or 1935 within the NHDP specification. It should also be noted that this 1936 algorithm does not impose the "hop limit" bound described in 1937 [RFC5614] when performing the path search that is used for relay 1938 selection. However, the algorithm below could be easily augmented to 1939 accommodate this additional criterion. It is not expected that the 1940 "hop limit" bound will provide significant benefit to the algorithm 1941 defined in this appendix. 1943 The tuple of "Router Priority" and "Router ID" is used in E-CDS relay 1944 set selection. Precedence is given to the "Router Priority" portion 1945 and the "Router ID" value is used as a tie-breaker. The evaluation 1946 of this tuple is referred to as "RtrPri(n)" in the description below 1947 where "n" references a specific router. Note it is possible that the 1948 "Router Priority" portion may be optional and the evaluation of 1949 "RtrPri()" be solely based upon the unique "Router ID". Since there 1950 MUST NOT be any duplicate "Router ID" values among SMF routers, a 1951 comparison of RtrPri(n) between any two routers will always be an 1952 inequality. The use of nodal degree for calculating "Router 1953 Priority" is RECOMMENDED as default and the largest IP address in the 1954 "Neighbor Address List" as advertised by NHDP MUST be used as the 1955 "Router ID". NHDP provides all interface address throughout the 1956 2-hop neighborhood through HELLO messages, so explicitly conveying a 1957 "Router ID" is not necessary. The following steps describe a basic 1958 algorithm for conducting E-CDS relay selection for a router "n0": 1959 1. Initialize the set "N1" with tuples ("Router Priority", "Router 1960 ID", "Neighbor Address List") for each 1-hop neighbor of "n0". 1961 2. If "N1" has less than 2 tuples, then "n0" does not elect itself 1962 as a relay and no further steps are taken. 1963 3. Initialize the set "N2" with tuples ("Router Priority", "Router 1964 ID", "2-hop address") for each "2-hop address" of "n0", where 1965 "2-hop address" is defined in NHDP. 1966 4. If "RtrPri(n0)" is greater than that of all tuples in the union 1967 of "N1" and "N2", then "n0" selects itself as a relay and no 1968 further steps are taken. 1969 5. Initialize all tuples in the union of "N1" and "N2" as 1970 "unvisited". 1971 6. Find the tuple "n1_Max" that has the largest "RtrPri()" of all 1972 tuples in "N1" 1973 7. Initialize queue "Q" to contain "n1_Max", marking "n1_Max" as 1974 "visited" 1975 8. While router queue "Q" is not empty, remove router "x" from the 1976 head of "Q", and for each 1-hop neighbor "n" of router "x" 1977 (excluding "n0") that is not marked "visited" 1978 A. Mark router "n" as "visited" 1979 B. If "RtrPri(n)" is greater than "RtrPri(n0), append "n" to "Q" 1980 9. If any tuple in "N1" remains "unvisited", then "n0" selects 1981 itself as a relay. Otherwise "n0" does not act as a relay. 1982 Note these steps are re-evaluated upon neighborhood status changes. 1983 Steps 5 through 8 of this procedure describe an approach to a path 1984 search. The purpose of this path search is to determine if paths 1985 exist from the 1-hop neighbor with maximum "RtrPri()" to all other 1986 1-hop neighbors without traversing an intermediate router with a 1987 "RtrPri()" value less than "RtrPri(n0)". These steps comprise a 1988 breadth-first traversal that evaluates only paths that meet that 1989 criteria. If all 1-hop neighbors of "n0" are "visited" during this 1990 traversal, then the path search has succeeded and router "n0" does 1991 not need to provide relay. It can be assumed that other routers will 1992 provide relay operation to ensure SMF connectivity. 1994 It is possible to extend this algorithm to consider neighboring SMF 1995 routers that are known to be statically configured for CF (always 1996 relaying). The modification to the above algorithm is to process 1997 such routers as having a maximum possible "Router Priority" value. 1998 It is expected that routers configured for CF and participating in 1999 NHDP would indicate this with use of the SMF_TYPE:CF and 2000 SMF_NBR_TYPE:CF TLV types in their NHDP_HELLO message and address 2001 blocks, respectively. 2003 Appendix B. Source-based Multipoint Relay (S-MPR) 2005 The source-based multipoint relay (S-MPR) set selection algorithm 2006 enables individual routers, using two-hop topology information, to 2007 select relays from their set of neighboring routers. Relays are 2008 selected so that forwarding to the router's complete two-hop neighbor 2009 set is covered. This distributed relay set selection technique has 2010 been shown to approximate a minimal connected dominating set (MCDS) 2011 in [JLMV02]. Individual routers must collect two-hop neighborhood 2012 information from neighbors, determine an appropriate current relay 2013 set, and inform selected neighbors of their relay status. Note that 2014 since each router picks its neighboring relays independently, S-MPR 2015 forwarders depend upon previous hop information (e.g, source MAC 2016 address) to operate correctly. The Optimized Link State Routing 2017 (OLSR) protocol has used this algorithm and protocol for relay of 2018 link state updates and other control information [RFC3626] and it has 2019 been demonstrated operationally in dynamic network environments. 2021 It is RECOMMENDED that the SMF_TYPE:S-MPR Message TLV be included in 2022 NHDP_HELLO messages that are generated by routers conducting S-MPR 2023 SMF operation. It is also RECOMMENDED that the SMF_NBR_TYPE:S-MPR 2024 address block TLV be used to specify which neighbor routers are 2025 conducting S-MPR SMF operation. 2027 B.1. S-MPR Relay Set Selection Overview 2029 The S-MPR algorithm uses bi-directional 1-hop and 2-hop neighborhood 2030 information collected via NHDP to select, from a router's 1-hop 2031 neighbors, a set of relays that will cover the router's entire 2-hop 2032 neighbor set upon forwarding. The algorithm described uses a 2033 "greedy" heuristic of first picking the 1-hop neighbor who will cover 2034 the most 2-hop neighbors. Then, excluding those 2-hop neighbors that 2035 have been covered, additional relays from its 1-hop neighbor set are 2036 iteratively selected until the entire 2-hop neighborhood is covered. 2037 Note that 1-hop neighbors also identified as 2-hop neighbors are 2038 considered as 1-hop neighbors only. 2040 NHDP HELLO messages supporting S-MPR forwarding operation SHOULD use 2041 the TLVs defined in Section 8.1 using the S-MPR extended type. The 2042 value field of an address block TLV which has a full type value of 2043 SMF_NBR_TYPE:S-MPR is defined in Table 17 such that signaling of MPR 2044 selections to 1-hop neighbors is possible. The value field of a 2045 message block TLV which has a full type value of SMF_TYPE:S-MPR is 2046 defined in Table 16 such that signaling of "Router Priority" 2047 (described as "WILLINGNESS" in [RFC3626]) to 1-hop neighbors is 2048 possible. It is important to note that S-MPR forwarding is dependent 2049 upon the previous hop of an incoming packet. An S-MPR router MUST 2050 forward packets only for neighbors which have explicitly selected it 2051 as a multi-point relay (i.e., its "selectors"). There are also some 2052 additional requirements for duplicate packet detection to support 2053 S-MPR SMF operation that are described below. 2055 For multiple interface operation, MPR selection SHOULD be conducted 2056 on a per-interface basis. However, it is possible to economize MPR 2057 selection among multiple interfaces by selecting common MPRs to the 2058 extent possible. 2060 B.2. S-MPR Forwarding Rules 2062 An S-MPR SMF router MUST only forward packets for neighbors that have 2063 explicitly selected it as an MPR. The source-based forwarding 2064 technique also stipulates some additional duplicate packet detection 2065 operations. For multiple network interfaces, independent DPD state 2066 MUST be maintained for each separate interface. The following 2067 provides the procedure for S-MPR packet forwarding given the arrival 2068 of a packet on a given interface, denoted . There are 2069 three possible actions, depending upon the previous-hop transmitter: 2071 1. If the previous-hop transmitter has selected the current router 2072 as an MPR, 2073 A. The packet identifier is checked against the DPD state for 2074 each possible outbound interface, including the . 2075 B. If the packet is not a duplicate for an outbound interface, 2076 the packet is forwarded on that interface and a DPD entry is 2077 made for the given packet identifier for the interface. 2078 C. If the packet is a duplicate, no action is taken for that 2079 interface. 2080 2. Else, if the previous-hop transmitter is a 1-hop symmetric 2081 neighbor, 2082 A. A DPD entry is added for that packet for the , but 2083 the packet is not forwarded. 2084 3. Otherwise, no action is taken. 2086 Case number two in the above table is non-intuitive, but important to 2087 ensure correctness of S-MPR SMF operation. The selection of source- 2088 based relays does not result in a common set among neighboring 2089 routers, so relays MUST mark in their DPD state, packets received 2090 from non-selector, symmetric, one-hop neighbors (for a given 2091 interface) and not forward subsequent duplicates of that packet if 2092 received on that interface. Deviation here can result in 2093 unnecessary, repeated packet forwarding throughout the network, or 2094 incomplete flooding. 2096 Nodes not participating in neighborhood discovery and relay set 2097 selection will not be able to source multicast packets into the area 2098 and have SMF forward them, unlike E-CDS or MPR-CDS where forwarding 2099 may occur dependent on topology. Correct S-MPR relay behavior will 2100 occur with the introduction of repeaters (non-NHDP/SMF participants 2101 that relay multicast packets using duplicate detection and CF) but 2102 the repeaters will not efficiently contribute to S-MPR forwarding as 2103 these routers will not be identified as neighbors (symmetric or 2104 otherwise) in the S-MPR forwarding process. NHDP/SMF participants 2105 MUST NOT provide extra forwarding, forwarding packets which are not 2106 selected by the algorithm, as this can disrupt network-wide S-MPR 2107 flooding, resulting in incomplete or inefficient flooding. The 2108 result is that non S-MPR SMF routers will be unable to source 2109 multicast packets and have them forwarded by other S-MPR SMF routers. 2111 B.3. S-MPR Neighborhood Discovery Requirements 2113 Nodes may optionally signal a "Router Priority" value to their one 2114 hop neighbors by using the SMF_TYPE:S-MPR message block TLV value 2115 field. If the value field is omitted, a default "Router Priority" 2116 value of 64 is to be assumed. This is summarized here: 2118 +---------------+---------+-----------------+ 2119 | Length(bytes) | Value | Router Priority | 2120 +---------------+---------+-----------------+ 2121 | 0 | N/A | 64 | 2122 | 1 | | 0-127 | 2123 +---------------+---------+-----------------+ 2125 Table 16: S-MPR Message TLV Values 2127 Where is a one octet long bit field defined as: 2129 bit 0: the leftmost bit is reserved and SHOULD be set to 0. 2131 bit 1-7: contain the "Router Priority" value, 0-127, which is 2132 associated with the "Neighbor Address List". 2134 "Router Priority" values for S-MPR are interpreted in the same 2135 fashion as "WILLINGNESS" ([RFC3626])with value 0 indicating a router 2136 will NEVER forward and value 127 indicating a router will ALWAYS 2137 forward. Values 1-126 indicate how likely a S-MPR SMF router will be 2138 selected as an MPR by a neighboring SMF router, with higher values 2139 increasing the likelihood. Combinations of value field lengths and 2140 values other than specified here are NOT permitted and SHOULD be 2141 ignored. Below is an example SMF_TYPE:S-MPR Message TLV. 2143 0 1 2 3 2144 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2145 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2146 ... | SMF_TYPE |1|0|0|1|0|0| | 2147 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2148 | S-MPR |0|0|0|0|0|0|0|1|R| priority | ... | 2149 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2151 Figure 8: S-MPR Message TLV Example 2153 S-MPR election operation requires 2-hop neighbor knowledge as 2154 provided by NHDP [RFC6130] or from external sources. MPRs are 2155 dynamically selected by each router and selections MUST be advertised 2156 and dynamically updated within NHDP or an equivalent protocol or 2157 mechanism. For NHDP use, the SMF_NBR_TYPE:S-MPR address block TLV 2158 value field is defined as such: 2160 +---------------+--------+----------+-------------------------------+ 2161 | Length(bytes) | # Addr | Value | Meaning | 2162 +---------------+--------+----------+-------------------------------+ 2163 | 0 | Any | N/A | NOT MPRs | 2164 | 1 | Any | | is for all addresses | 2165 | N | N | * | Each address gets its own | 2166 | | | | | 2167 +---------------+--------+----------+-------------------------------+ 2169 Table 17: S-MPR Address Block TLV Values 2171 Where , if present, is a one octet bit field defined as: 2173 bit 0: The leftmost bit is the M bit. When set indicates MPR 2174 selection of the relevant interface, represented by the associated 2175 address(es), by the originator router of the NHDP HELLO message. 2176 When unset, indicates the originator router of the NHDP HELLO message 2177 has not selected the relevant interfaces, represented by the 2178 associated address(es), as its MPR. 2180 bit 1-7: are reserved and SHOULD be set to 0. 2182 Combinations of value field lengths and number of addresses other 2183 than specified here are NOT permitted and SHOULD be ignored. All 2184 bits, excepting the leftmost bit, are RESERVED and SHOULD be set to 2185 0. 2187 0 1 2 3 2188 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2 3 4 5 6 7 8 9 0 1 2189 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2190 ... | SMF_NBR_TYPE |1|1|0|1|0|0| | 2191 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2192 | S-MPR | start-index |0|0|0|0|0|0|0|1|M| reserved | 2193 +-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+ 2195 Figure 9: S-MPR Address Block TLV Example 2197 The single index TLV example, depicted in Figure 9, indicates that 2198 the address specified by the field is running SMF using 2199 S-MPR and has been selected by the originator of the NHDP HELLO 2200 message as an MPR forwarder if the M bit is set. Multivalued TLVs 2201 may also be used to specify MPR selection status of multiple 2202 addresses using only one TLV. See Figure 7 for a similar example on 2203 how this may be done. 2205 B.4. S-MPR Selection Algorithm 2207 This section describes a basic algorithm for the S-MPR selection 2208 process. Note that the selection is with respect to a specific 2209 interface of the router performing selection and other router 2210 interfaces referenced are reachable from this reference router 2211 interface. This is consistent with the S-MPR forwarding rules 2212 described above. When multiple interfaces per router are used, it is 2213 possible to enhance the overall selection process across multiple 2214 interfaces such that common routers are selected as MPRs for each 2215 interface to avoid unnecessary inefficiencies in flooding. The 2216 following steps describe a basic algorithm for conducting S-MPR 2217 selection for a router interface "n0": 2219 1. Initialize the set "MPR" to empty. 2220 2. Initialize the set "N1" to include all 1-hop neighbors of "n0". 2221 3. Initialize the set "N2" to include all 2-hop neighbors, excluding 2222 "n0" and any routers in "N1". Nodes which are only reachable via 2223 "N1" routers with router priority values of NEVER are also 2224 excluded. 2225 4. For each interface "y" in "N1", initialize a set "N2(y)" to 2226 include any interfaces in "N2" that are 1-hop neighbors of "y". 2227 5. For each interface "x" in "N1" with a router priority value of 2228 "ALWAYS" (or using CF relay algorithm), select "x" as a MPR: 2229 A. Add "x" to the set "MPR" and remove "x" from "N1". 2230 B. For each interface "z" in "N2(x)", remove "z" from "N2" 2231 C. For each interface "y" in "N1", remove any interfaces in 2232 "N2(x)" from "N2(y)" 2234 6. For each interface "z" in "N2", initialize the set "N1(z)" to 2235 include any interfaces in "N1" that are 1-hop neighbors of "z". 2236 7. For each interface "x" in "N2" where "N1(x)" has only one member, 2237 select "x" as a MPR: 2238 A. Add "x" to the set "MPR" and remove "x" from "N1". 2239 B. For each interface "z" in "N2(x)", remove "z" from "N2" and 2240 delete "N1(z)" 2241 C. For each interface "y" in "N1", remove any interfaces in 2242 "N2(x)" from "N2(y)" 2243 8. While "N2" is not empty, select the interface "x" in "N1" with 2244 the largest router priority which has the number of members in 2245 "N_2(x)" as a MPR: 2246 A. Add "x" to the set "MPR" and remove "x" from "N1". 2247 B. For each interface "z" in "N2(x)", remove "z" from "N2" 2248 C. For each interface "y" in "N1", remove any interfaces in 2249 "N2(x)" from "N2(y)" 2251 After the set of routers "MPR" is selected, router "n_0" must signal 2252 its selections to its neighbors. With NHDP, this is done by using 2253 the MPR address block TLV to mark selected neighbor addresses in 2254 NHDP_HELLO messages. Neighbors MUST record their MPR selection 2255 status and the previous hop address (e.g., link or MAC layer) of the 2256 selector. Note these steps are re-evaluated upon neighborhood status 2257 changes. 2259 Appendix C. Multipoint Relay Connected Dominating Set (MPR-CDS) 2260 Algorithm 2262 The MPR-CDS algorithm is an extension to the basic S-MPR election 2263 algorithm that results in a shared (non source-specific) SMF CDS. 2264 Thus its forwarding rules are not dependent upon previous hop 2265 information similar to E-CDS. An overview of the MPR-CDS selection 2266 algorithm is provided in [MPR-CDS]. 2268 It is RECOMMENDED that the SMF_TYPE Message TLV be included in 2269 NHDP_HELLO messages that are generated by routers conducting MPR-CDS 2270 SMF operation. 2272 C.1. MPR-CDS Relay Set Selection Overview 2274 The MPR-CDS relay set selection process is based upon the MPR 2275 selection process of the S-MPR algorithm with the added refinement of 2276 a distributed technique for subsequently down-selecting to a common 2277 reduced, shared relay set. A router ordering (or "prioritization") 2278 metric is used as part of this down-selection process like the E-CDS 2279 algorithm, this metric can be based upon router address(es) or some 2280 other unique router identifier (e.g. "Router ID" based on largest 2281 address contained within the "Neighbor Address List") as well as an 2282 additional "Router Priority" measure, if desired. The process for 2283 MPR-CDS relay selection is as follows: 2284 1. First, MPR selection per the S-MPR algorithm is conducted, with 2285 selectors informing their MPRs (via NHDP) of their selection. 2286 2. Then, the following rules are used on a distributed basis by 2287 selected routers to possibly deselect themselves and thus jointly 2288 establish a common set of shared SMF relays: 2289 A. If a selected router has a larger "RtrPri()" than all of its 2290 1-hop symmetric neighbors, then it acts as a relay for all 2291 multicast traffic, regardless of the previous hop 2292 B. Else, if the 1-hop symmetric neighbor with the largest 2293 "RtrPri()" value has selected the router, then it also acts 2294 as a relay for all multicast traffic, regardless of the 2295 previous hop. 2296 C. Otherwise, it deselects itself as a relay and does not 2297 forward any traffic unless changes occur that require re- 2298 evaluation of the above steps. 2300 This technique shares many of the desirable properties of the E-CDS 2301 technique with regards to compatibility with multicast sources not 2302 participating in NHDP and the opportunity for statically-configure CF 2303 routers to be present, regardless of their participation in NHDP. 2305 C.2. MPR-CDS Forwarding Rules 2307 The forwarding rules for MPR-CDS are common with those of E-CDS. Any 2308 SMF router that has selected itself as a relay performs DPD and 2309 forwards all non-duplicative multicast traffic allowed by the present 2310 forwarding policy. Packet previous hop knowledge is not needed for 2311 forwarding decisions when using MPR-CDS. 2313 1. Upon packet reception, DPD is performed. Note MPR-CDS require 2314 one duplicate table for the set of interfaces associated with the 2315 relay set selection. 2316 2. If the packet is a duplicate, no further action is taken. 2317 3. If the packet is non-duplicative: 2318 A. A DPD entry is added for the packet identifier 2319 B. The packet is forwarded out all interfaces associated with 2320 the relay set selection 2322 As previously mentioned, even packets sourced (or relayed) by routers 2323 not participating in NHDP and/or the MPR-CDS relay set selection may 2324 be forwarded by MPR-CDS forwarders without problem. A particular 2325 deployment MAY choose to not forward packets from sources or relays 2326 that have been explicitly identified via NHDP or other means as 2327 operating as part of a different relay set algorithm (e.g. S-MPR) to 2328 allow coexistent deployments to operate correctly. 2330 C.3. MPR-CDS Neighborhood Discovery Requirements 2332 The neighborhood discovery requirements for MPR-CDS have commonality 2333 with both the S-MPR and E-CDS algorithms. MPR-CDS selection 2334 operation requires 2-hop neighbor knowledge as provided by NHDP 2335 [RFC6130] or from external sources. Unlike S-MPR operation, there is 2336 no need for associating link-layer address information with 1-hop 2337 neighbors since MPR-CDS forwarding is independent of the previous hop 2338 similar to E-CDS forwarding. 2340 To advertise an optional "Router Priority" value or "WILLINGNESS" an 2341 originating router may use the Message TLV of type SMF_TYPE:MPR-CDS 2342 which shares a common format with both SMF_TYPE:E-CDS 2343 Table 14 and SMF_TYPE:S-MPR Table 16. 2345 MPR-CDS only requires 1-hop knowledge of "Router Priority" for 2346 correct operation. In the S-MPR phase of MPR-CDS selection, MPRs are 2347 dynamically determined by each router and selections MUST be 2348 advertised and dynamically updated using NHDP or an equivalent 2349 protocol or mechanism. The field of the SMF_NBR_TYPE:MPR-CDS 2350 type TLV shares a common format with SMF_NBR_TYPE:S-MPR Table 17 to 2351 convey MPR selection. 2353 C.4. MPR-CDS Selection Algorithm 2355 This section describes an algorithm for the MPR-CDS selection 2356 process. Note that the selection described is with respect to a 2357 specific interface of the router performing selection and other 2358 router interfaces referenced are reachable from this reference router 2359 interface. An ordered tuple of "Router Priority" and "Router ID" is 2360 used in MPR-CDS relay set selection. The "Router ID" value should be 2361 set to the largest advertised address of a given router, this 2362 information is provided to one hop neighbors via NHDP by default. 2363 Precedence is given to the "Router Priority" portion and the "Router 2364 ID" value is used as a tie-breaker. The evaluation of this tuple is 2365 referred to as "RtrPri(n)" in the description below where "n" 2366 references a specific router. Note it is possible that the "Router 2367 Priority" portion may be optional and the evaluation of "RtrPri()" be 2368 solely based upon the unique "Router ID". Since there MUST NOT be 2369 any duplicate address values among SMF routers, a comparison of 2370 RtrPri(n) between any two routers will always be an inequality. The 2371 following steps, repeated upon any changes detected within the 1-hop 2372 and 2-hop neighborhood, describe a basic algorithm for conducting 2373 MPR-CDS selection for a router interface "n0": 2375 1. Perform steps 1-8 of Appendix B.4 to select MPRs from the set of 2376 1-hop neighbors of "n0" and notify/update neighbors of 2377 selections. 2379 2. Upon being selected as an MPR (or any change in the set of 2380 routers selecting "n0" as an MPR): 2381 A. If no neighbors have selected "n0" as an MPR, "n0" does not 2382 act as a relay and no further steps are taken until a change 2383 in neighborhood topology or selection status occurs. 2384 B. Determine the router "n1_max" that has the maximum "RtrPri()" 2385 of all 1-hop neighbors. 2386 C. If "RtrPri(n0)" is greater than "RtrPri(n1_max)", then "n0" 2387 selects itself as a relay for all multicast packets, 2388 D. Else, if "n1_max" has selected "n0" as an MPR, then "0" 2389 selects itself as a relay for all multicast packets. 2390 E. Otherwise, "n0" does not act as a relay. 2392 It is possible to extend this algorithm to consider neighboring SMF 2393 routers that are known to be statically configured for CF (always 2394 relaying). The modification to the above algorithm is to process 2395 such routers as having a maximum possible "Router Priority" value. 2396 This is the same as the case for participating routers that have been 2397 configured with a S-MPR "WILLINGNESS" value of "WILL_ALWAYS". It is 2398 expected that routers configured for CF and participating in NHDP 2399 would indicate their status with use of the SMF_TYPE TLV type in 2400 their NHDP_HELLO message TLV block. It is important to note however 2401 that CF routers will not select MPR routers and therefore cannot 2402 guarantee connectedness. 2404 Author's Address 2406 Joseph Macker 2407 NRL 2408 Washington, DC 20375 2409 USA 2411 Email: macker@itd.nrl.navy.mil