idnits 2.17.1 draft-ietf-hip-rfc4423-bis-17.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, but was first submitted on or after 10 November 2008. The disclaimer is usually necessary only for documents that revise or obsolete older RFCs, and that take significant amounts of text from those RFCs. If you can contact all authors of the source material and they are willing to grant the BCP78 rights to the IETF Trust, 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 (August 7, 2017) is 2447 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Outdated reference: A later version (-24) exists of draft-ietf-hip-dex-05 == Outdated reference: A later version (-33) exists of draft-ietf-hip-native-nat-traversal-16 -- Obsolete informational reference (is this intentional?): RFC 2535 (Obsoleted by RFC 4033, RFC 4034, RFC 4035) -- Obsolete informational reference (is this intentional?): RFC 2766 (Obsoleted by RFC 4966) -- Obsolete informational reference (is this intentional?): RFC 4306 (Obsoleted by RFC 5996) -- Obsolete informational reference (is this intentional?): RFC 4423 (Obsoleted by RFC 9063) Summary: 0 errors (**), 0 flaws (~~), 4 warnings (==), 5 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Network Working Group R. Moskowitz, Ed. 3 Internet-Draft HTT Consulting 4 Obsoletes: 4423 (if approved) M. Komu 5 Intended status: Informational Ericsson 6 Expires: February 8, 2018 August 7, 2017 8 Host Identity Protocol Architecture 9 draft-ietf-hip-rfc4423-bis-17 11 Abstract 13 This memo describes a new namespace, the Host Identity namespace, and 14 a new protocol layer, the Host Identity Protocol, between the 15 internetworking and transport layers. Herein are presented the 16 basics of the current namespaces, their strengths and weaknesses, and 17 how a new namespace will add completeness to them. The roles of this 18 new namespace in the protocols are defined. 20 This document obsoletes RFC 4423 and addresses the concerns raised by 21 the IESG, particularly that of crypto agility. It incorporates 22 lessons learned from the implementations of RFC 5201 and goes further 23 to explain how HIP works as a secure signaling channel. 25 Status of This Memo 27 This Internet-Draft is submitted in full conformance with the 28 provisions of BCP 78 and BCP 79. 30 Internet-Drafts are working documents of the Internet Engineering 31 Task Force (IETF). Note that other groups may also distribute 32 working documents as Internet-Drafts. The list of current Internet- 33 Drafts is at http://datatracker.ietf.org/drafts/current/. 35 Internet-Drafts are draft documents valid for a maximum of six months 36 and may be updated, replaced, or obsoleted by other documents at any 37 time. It is inappropriate to use Internet-Drafts as reference 38 material or to cite them other than as "work in progress." 40 This Internet-Draft will expire on February 8, 2018. 42 Copyright Notice 44 Copyright (c) 2017 IETF Trust and the persons identified as the 45 document authors. All rights reserved. 47 This document is subject to BCP 78 and the IETF Trust's Legal 48 Provisions Relating to IETF Documents 49 (http://trustee.ietf.org/license-info) in effect on the date of 50 publication of this document. Please review these documents 51 carefully, as they describe your rights and restrictions with respect 52 to this document. Code Components extracted from this document must 53 include Simplified BSD License text as described in Section 4.e of 54 the Trust Legal Provisions and are provided without warranty as 55 described in the Simplified BSD License. 57 This document may contain material from IETF Documents or IETF 58 Contributions published or made publicly available before November 59 10, 2008. The person(s) controlling the copyright in some of this 60 material may not have granted the IETF Trust the right to allow 61 modifications of such material outside the IETF Standards Process. 62 Without obtaining an adequate license from the person(s) controlling 63 the copyright in such materials, this document may not be modified 64 outside the IETF Standards Process, and derivative works of it may 65 not be created outside the IETF Standards Process, except to format 66 it for publication as an RFC or to translate it into languages other 67 than English. 69 Table of Contents 71 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 72 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 4 73 2.1. Terms common to other documents . . . . . . . . . . . . . . 4 74 2.2. Terms specific to this and other HIP documents . . . . . . 5 75 3. Background . . . . . . . . . . . . . . . . . . . . . . . . . 6 76 3.1. A desire for a namespace for computing platforms . . . . . 7 77 4. Host Identity namespace . . . . . . . . . . . . . . . . . . . 9 78 4.1. Host Identifiers . . . . . . . . . . . . . . . . . . . . . 10 79 4.2. Host Identity Hash (HIH) . . . . . . . . . . . . . . . . . 10 80 4.3. Host Identity Tag (HIT) . . . . . . . . . . . . . . . . . . 11 81 4.4. Local Scope Identifier (LSI) . . . . . . . . . . . . . . . 11 82 4.5. Storing Host Identifiers in directories . . . . . . . . . . 12 83 5. New stack architecture . . . . . . . . . . . . . . . . . . . 13 84 5.1. On the multiplicity of identities . . . . . . . . . . . . . 14 85 6. Control plane . . . . . . . . . . . . . . . . . . . . . . . . 15 86 6.1. Base exchange . . . . . . . . . . . . . . . . . . . . . . . 15 87 6.2. End-host mobility and multi-homing . . . . . . . . . . . . 16 88 6.3. Rendezvous mechanism . . . . . . . . . . . . . . . . . . . 16 89 6.4. Relay mechanism . . . . . . . . . . . . . . . . . . . . . . 17 90 6.5. Termination of the control plane . . . . . . . . . . . . . 17 91 7. Data plane . . . . . . . . . . . . . . . . . . . . . . . . . 17 92 8. HIP and NATs . . . . . . . . . . . . . . . . . . . . . . . . 18 93 8.1. HIP and Upper-layer checksums . . . . . . . . . . . . . . . 19 94 9. Multicast . . . . . . . . . . . . . . . . . . . . . . . . . . 19 95 10. HIP policies . . . . . . . . . . . . . . . . . . . . . . . . 19 96 11. Design considerations . . . . . . . . . . . . . . . . . . . . 20 97 11.1. Benefits of HIP . . . . . . . . . . . . . . . . . . . . . 20 98 11.2. Drawbacks of HIP . . . . . . . . . . . . . . . . . . . . . 23 99 11.3. Deployment and adoption considerations . . . . . . . . . . 24 100 11.3.1. Deployment analysis . . . . . . . . . . . . . . . . . . 24 101 11.3.2. HIP in 802.15.4 networks . . . . . . . . . . . . . . . . 25 102 11.3.3. HIP and Internet of Things . . . . . . . . . . . . . . . 26 103 11.4. Answers to NSRG questions . . . . . . . . . . . . . . . . 27 104 12. Security considerations . . . . . . . . . . . . . . . . . . . 29 105 12.1. MiTM Attacks . . . . . . . . . . . . . . . . . . . . . . . 29 106 12.2. Protection against flooding attacks . . . . . . . . . . . 30 107 12.3. HITs used in ACLs . . . . . . . . . . . . . . . . . . . . 31 108 12.4. Alternative HI considerations . . . . . . . . . . . . . . 32 109 13. IANA considerations . . . . . . . . . . . . . . . . . . . . . 33 110 14. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 33 111 15. Changes from RFC 4423 . . . . . . . . . . . . . . . . . . . . 33 112 16. References . . . . . . . . . . . . . . . . . . . . . . . . . 34 113 16.1. Normative References . . . . . . . . . . . . . . . . . . . 34 114 16.2. Informative references . . . . . . . . . . . . . . . . . . 35 115 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 42 117 1. Introduction 119 The Internet has two important global namespaces: Internet Protocol 120 (IP) addresses and Domain Name Service (DNS) names. These two 121 namespaces have a set of features and abstractions that have powered 122 the Internet to what it is today. They also have a number of 123 weaknesses. Basically, since they are all we have, we try and do too 124 much with them. Semantic overloading and functionality extensions 125 have greatly complicated these namespaces. 127 The proposed Host Identity namespace fills an important gap between 128 the IP and DNS namespaces. A Host Identity conceptually refers to a 129 computing platform, and there may be multiple such Host Identities 130 per computing platform (because the platform may wish to present a 131 different identity to different communicating peers). The Host 132 Identity namespace consists of Host Identifiers (HI). There is 133 exactly one Host Identifier for each Host Identity (although there 134 may be transient periods of time such as key replacement when more 135 than one identifier may be active). While this text later talks 136 about non-cryptographic Host Identifiers, the architecture focuses on 137 the case in which Host Identifiers are cryptographic in nature. 138 Specifically, the Host Identifier is the public key of an asymmetric 139 key-pair. Each Host Identity uniquely identifies a single host, 140 i.e., no two hosts have the same Host Identity. If two or more 141 computing platforms have the same Host Identifier, then they are 142 instantiating a distributed host. The Host Identifier can either be 143 public (e.g. published in the DNS), or unpublished. Client systems 144 will tend to have both public and unpublished Host Identifiers. 146 There is a subtle but important difference between Host Identities 147 and Host Identifiers. An Identity refers to the abstract entity that 148 is identified. An Identifier, on the other hand, refers to the 149 concrete bit pattern that is used in the identification process. 151 Although the Host Identifiers could be used in many authentication 152 systems, such as IKEv2 [RFC4306], the presented architecture 153 introduces a new protocol, called the Host Identity Protocol (HIP), 154 and a cryptographic exchange, called the HIP base exchange; see also 155 Section 6. HIP provides for limited forms of trust between systems, 156 enhance mobility, multi-homing and dynamic IP renumbering, aid in 157 protocol translation / transition, and reduce certain types of 158 denial-of-service (DoS) attacks. 160 When HIP is used, the actual payload traffic between two HIP hosts is 161 typically, but not necessarily, protected with ESP [RFC7402]. The 162 Host Identities are used to create the needed ESP Security 163 Associations (SAs) and to authenticate the hosts. When ESP is used, 164 the actual payload IP packets do not differ in any way from standard 165 ESP protected IP packets. 167 Much has been learned about HIP [RFC6538] since [RFC4423] was 168 published. This document expands Host Identities beyond use to 169 enable IP connectivity and security to general interhost secure 170 signalling at any protocol layer. The signal may establish a 171 security association between the hosts, or simply pass information 172 within the channel. 174 2. Terminology 176 2.1. Terms common to other documents 177 +---------------+---------------------------------------------------+ 178 | Term | Explanation | 179 +---------------+---------------------------------------------------+ 180 | Public key | The public key of an asymmetric cryptographic key | 181 | | pair. Used as a publicly known identifier for | 182 | | cryptographic identity authentication. Public is | 183 | | a a relative term here, ranging from "known to | 184 | | peers only" to "known to the world." | 185 | Private key | The private or secret key of an asymmetric | 186 | | cryptographic key pair. Assumed to be known only | 187 | | to the party identified by the corresponding | 188 | | public key. Used by the identified party to | 189 | | authenticate its identity to other parties. | 190 | Public key | An asymmetric cryptographic key pair consisting | 191 | pair | of public and private keys. For example, Rivest- | 192 | | Shamir-Adleman (RSA), Digital Signature Algorithm | 193 | | (DSA) and Elliptic Curve DSA (ECDSA) key pairs | 194 | | are such key pairs. | 195 | End-point | A communicating entity. For historical reasons, | 196 | | the term 'computing platform' is used in this | 197 | | document as a (rough) synonym for end-point. | 198 +---------------+---------------------------------------------------+ 200 2.2. Terms specific to this and other HIP documents 202 It should be noted that many of the terms defined herein are 203 tautologous, self-referential or defined through circular reference 204 to other terms. This is due to the succinct nature of the 205 definitions. See the text elsewhere in this document and the base 206 specification [RFC7401] for more elaborate explanations. 208 +---------------+---------------------------------------------------+ 209 | Term | Explanation | 210 +---------------+---------------------------------------------------+ 211 | Computing | An entity capable of communicating and computing, | 212 | platform | for example, a computer. See the definition of | 213 | | 'End-point', above. | 214 | HIP base | A cryptographic protocol; see also Section 6. | 215 | exchange | | 216 | HIP packet | An IP packet that carries a 'Host Identity | 217 | | Protocol' message. | 218 | Host Identity | An abstract concept assigned to a 'computing | 219 | | platform'. See 'Host Identifier', below. | 220 | Host Identity | A name space formed by all possible Host | 221 | namespace | Identifiers. | 222 | Host Identity | A protocol used to carry and authenticate Host | 223 | Protocol | Identifiers and other information. | 224 | Host Identity | The cryptographic hash used in creating the Host | 225 | Hash | Identity Tag from the Host Identity. | 226 | Host Identity | A 128-bit datum created by taking a cryptographic | 227 | Tag | hash over a Host Identifier plus bits to identify | 228 | | which hash used. | 229 | Host | A public key used as a name for a Host Identity. | 230 | Identifier | | 231 | Local Scope | A 32-bit datum denoting a Host Identity. | 232 | Identifier | | 233 | Public Host | A published or publicly known Host Identifier | 234 | Identifier | used as a public name for a Host Identity, and | 235 | and Identity | the corresponding Identity. | 236 | Unpublished | A Host Identifier that is not placed in any | 237 | Host | public directory, and the corresponding Host | 238 | Identifier | Identity. Unpublished Host Identities are | 239 | and Identity | typically short lived in nature, being often | 240 | | replaced and possibly used just once. | 241 | Rendezvous | A mechanism used to locate mobile hosts based on | 242 | Mechanism | their HIT. | 243 +---------------+---------------------------------------------------+ 245 3. Background 247 The Internet is built from three principal components: computing 248 platforms (end-points), packet transport (i.e., internetworking) 249 infrastructure, and services (applications). The Internet exists to 250 service two principal components: people and robotic services 251 (silicon-based people, if you will). All these components need to be 252 named in order to interact in a scalable manner. Here we concentrate 253 on naming computing platforms and packet transport elements. 255 There are two principal namespaces in use in the Internet for these 256 components: IP addresses, and Domain Names. Domain Names provide 257 hierarchically assigned names for some computing platforms and some 258 services. Each hierarchy is delegated from the level above; there is 259 no anonymity in Domain Names. Email, HTTP, and SIP addresses all 260 reference Domain Names. 262 The IP addressing namespace has been overloaded to name both 263 interfaces (at layer-3) and endpoints (for the endpoint-specific part 264 of layer-3, and for layer-4). In their role as interface names, IP 265 addresses are sometimes called "locators" and serve as an endpoint 266 within a routing topology. 268 IP addresses are numbers that name networking interfaces, and 269 typically only when the interface is connected to the network. 270 Originally, IP addresses had long-term significance. Today, the vast 271 number of interfaces use ephemeral and/or non-unique IP addresses. 272 That is, every time an interface is connected to the network, it is 273 assigned an IP address. 275 In the current Internet, the transport layers are coupled to the IP 276 addresses. Neither can evolve separately from the other. IPng 277 deliberations were strongly shaped by the decision that a 278 corresponding TCPng would not be created. 280 There are three critical deficiencies with the current namespaces. 281 Firstly, dynamic readdressing cannot be directly managed. Secondly, 282 confidentiality is not provided in a consistent, trustable manner. 283 Finally, authentication for systems and datagrams is not provided. 284 All of these deficiencies arise because computing platforms are not 285 well named with the current namespaces. 287 3.1. A desire for a namespace for computing platforms 289 An independent namespace for computing platforms could be used in 290 end-to-end operations independent of the evolution of the 291 internetworking layer and across the many internetworking layers. 292 This could support rapid readdressing of the internetworking layer 293 because of mobility, rehoming, or renumbering. 295 If the namespace for computing platforms is based on public-key 296 cryptography, it can also provide authentication services. If this 297 namespace is locally created without requiring registration, it can 298 provide anonymity. 300 Such a namespace (for computing platforms) and the names in it should 301 have the following characteristics: 303 o The namespace should be applied to the IP 'kernel' or stack. The 304 IP stack is the 'component' between applications and the packet 305 transport infrastructure. 307 o The namespace should fully decouple the internetworking layer from 308 the higher layers. The names should replace all occurrences of IP 309 addresses within applications (like in the Transport Control 310 Block, TCB). This replacement can be handled transparently for 311 legacy applications as the LSIs and HITs are compatible with IPv4 312 and IPv6 addresses [RFC5338]. However, HIP-aware applications 313 require some modifications from the developers, who may employ 314 networking API extensions for HIP [RFC6317]. 316 o The introduction of the namespace should not mandate any 317 administrative infrastructure. Deployment must come from the 318 bottom up, in a pairwise deployment. 320 o The names should have a fixed length representation, for easy 321 inclusion in datagram headers and existing programming interfaces 322 (e.g the TCB). 324 o Using the namespace should be affordable when used in protocols. 325 This is primarily a packet size issue. There is also a 326 computational concern in affordability. 328 o Name collisions should be avoided as much as possible. The 329 mathematics of the birthday paradox can be used to estimate the 330 chance of a collision in a given population and hash space. In 331 general, for a random hash space of size n bits, we would expect 332 to obtain a collision after approximately 1.2*sqrt(2**n) hashes 333 were obtained. For 64 bits, this number is roughly 4 billion. A 334 hash size of 64 bits may be too small to avoid collisions in a 335 large population; for example, there is a 1% chance of collision 336 in a population of 640M. For 100 bits (or more), we would not 337 expect a collision until approximately 2**50 (1 quadrillion) 338 hashes were generated. 340 o The names should have a localized abstraction so that they can be 341 used in existing protocols and APIs. 343 o It must be possible to create names locally. When such names are 344 not published, this can provide anonymity at the cost of making 345 resolvability very difficult. 347 o The namespace should provide authentication services. 349 o The names should be long lived, but replaceable at any time. This 350 impacts access control lists; short lifetimes will tend to result 351 in tedious list maintenance or require a namespace infrastructure 352 for central control of access lists. 354 In this document, a new namespace approaching these ideas is called 355 the Host Identity namespace. Using Host Identities requires its own 356 protocol layer, the Host Identity Protocol, between the 357 internetworking and transport layers. The names are based on public- 358 key cryptography to supply authentication services. Properly 359 designed, it can deliver all of the above stated requirements. 361 4. Host Identity namespace 363 A name in the Host Identity namespace, a Host Identifier (HI), 364 represents a statistically globally unique name for naming any system 365 with an IP stack. This identity is normally associated with, but not 366 limited to, an IP stack. A system can have multiple identities, some 367 'well known', some unpublished or 'anonymous'. A system may self- 368 assert its own identity, or may use a third-party authenticator like 369 DNSSEC [RFC2535], PGP, or X.509 to 'notarize' the identity assertion 370 to another namespace. It is expected that the Host Identifiers will 371 initially be authenticated with DNSSEC and that all implementations 372 will support DNSSEC as a minimal baseline. 374 In theory, any name that can claim to be 'statistically globally 375 unique' may serve as a Host Identifier. In the HIP architecture, the 376 public key of a private-public key pair has been chosen as the Host 377 Identifier because it can be self managed and it is computationally 378 difficult to forge. As specified in the Host Identity Protocol 379 [RFC7401] specification, a public-key-based HI can authenticate the 380 HIP packets and protect them from man-in-the-middle attacks. Since 381 authenticated datagrams are mandatory to provide much of HIP's 382 denial-of-service protection, the Diffie-Hellman exchange in HIP base 383 exchange has to be authenticated. Thus, only public-key HI and 384 authenticated HIP messages are supported in practice. 386 In this document, the non-cryptographic forms of HI and HIP are 387 presented to complete the theory of HI, but they should not be 388 implemented as they could produce worse denial-of-service attacks 389 than the Internet has without Host Identity. There has been past 390 research in challenge puzzles to use non-cryptographic HI, for Radio 391 Frequency IDentification (RFID), in an HIP exchange tailored to the 392 workings of such challenges (as described further in [urien-rfid] and 393 [urien-rfid-draft]). 395 4.1. Host Identifiers 397 Host Identity adds two main features to Internet protocols. The 398 first is a decoupling of the internetworking and transport layers; 399 see Section 5. This decoupling will allow for independent evolution 400 of the two layers. Additionally, it can provide end-to-end services 401 over multiple internetworking realms. The second feature is host 402 authentication. Because the Host Identifier is a public key, this 403 key can be used for authentication in security protocols like ESP. 405 The only completely defined structure of the Host Identity is that of 406 a public/private key pair. In this case, the Host Identity is 407 referred to by its public component, the public key. Thus, the name 408 representing a Host Identity in the Host Identity namespace, i.e., 409 the Host Identifier, is the public key. In a way, the possession of 410 the private key defines the Identity itself. If the private key is 411 possessed by more than one node, the Identity can be considered to be 412 a distributed one. 414 Architecturally, any other Internet naming convention might form a 415 usable base for Host Identifiers. However, non-cryptographic names 416 should only be used in situations of high trust - low risk. That is 417 any place where host authentication is not needed (no risk of host 418 spoofing) and no use of ESP. However, at least for interconnected 419 networks spanning several operational domains, the set of 420 environments where the risk of host spoofing allowed by non- 421 cryptographic Host Identifiers is acceptable is the null set. Hence, 422 the current HIP documents do not specify how to use any other types 423 of Host Identifiers but public keys. For instance, Back-to-My-Mac 424 [RFC6281] from Apple comes pretty close to the functionality of HIP, 425 but unlike HIP, it is based on non-cryptographic identifiers. 427 The actual Host Identifiers are never directly used at the transport 428 or network layers. The corresponding Host Identifiers (public keys) 429 may be stored in various DNS or other directories as identified 430 elsewhere in this document, and they are passed in the HIP base 431 exchange. A Host Identity Tag (HIT) is used in other protocols to 432 represent the Host Identity. Another representation of the Host 433 Identities, the Local Scope Identifier (LSI), can also be used in 434 protocols and APIs. 436 4.2. Host Identity Hash (HIH) 438 The Host Identity Hash is the cryptographic hash algorithm used in 439 producing the HIT from the HI. It is also the hash used throughout 440 the HIP protocol for consistency and simplicity. It is possible to 441 for the two hosts in the HIP exchange to use different hash 442 algorithms. 444 Multiple HIHs within HIP are needed to address the moving target of 445 creation and eventual compromise of cryptographic hashes. This 446 significantly complicates HIP and offers an attacker an additional 447 downgrade attack that is mitigated in the HIP protocol [RFC7401]. 449 4.3. Host Identity Tag (HIT) 451 A Host Identity Tag is a 128-bit representation for a Host Identity. 452 It is created from an HIH, an IPv6 prefix [RFC7343] and a hash 453 identifier. There are two advantages of using the HIT over using the 454 Host Identifier in protocols. Firstly, its fixed length makes for 455 easier protocol coding and also better manages the packet size cost 456 of this technology. Secondly, it presents the identity in a 457 consistent format to the protocol independent of the cryptographic 458 algorithms used. 460 In essence, the HIT is a hash over the public key. As such, two 461 algorithms affect the generation of a HIT: the public-key algorithm 462 of the HI and the used HIH. The two algorithms are encoded in the 463 bit presentation of the HIT. As the two communicating parties may 464 support different algorithms, [RFC7401] defines the minimum set for 465 interoperability. For further interoperability, the responder may 466 store its keys in DNS records, and thus the initiator may have to 467 couple destination HITs with appropriate source HITs according to 468 matching HIH. 470 In the HIP packets, the HITs identify the sender and recipient of a 471 packet. Consequently, a HIT should be unique in the whole IP 472 universe as long as it is being used. In the extremely rare case of 473 a single HIT mapping to more than one Host Identity, the Host 474 Identifiers (public keys) will make the final difference. If there 475 is more than one public key for a given node, the HIT acts as a hint 476 for the correct public key to use. 478 4.4. Local Scope Identifier (LSI) 480 An LSI is a 32-bit localized representation for a Host Identity. The 481 purpose of an LSI is to facilitate using Host Identities in existing 482 APIs for IPv4-based applications. Besides facilitating HIP-based 483 connectivity for legacy IPv4 applications, the LSIs are beneficial in 484 two other scenarios [RFC6538]. 486 In the first scenario, two IPv4-only applications are residing on two 487 separate hosts connected by IPv6-only network. With HIP-based 488 connectivity, the two applications are able to communicate despite of 489 the mismatch in the protocol families of the applications and the 490 underlying network. The reason is that the HIP layer translates the 491 LSIs originating from the upper layers into routable IPv6 locators 492 before delivering the packets on the wire. 494 The second scenario is the same as the first one, but with the 495 difference that one of the applications supports only IPv6. Now two 496 obstacles hinder the communication between the application: the 497 addressing families of the two applications differ, and the 498 application residing at the IPv4-only side is again unable to 499 communicate because of the mismatch between addressing families of 500 the application (IPv4) and network (IPv6). With HIP-based 501 connectivity for applications, this scenario works; the HIP layer can 502 choose whether to translate the locator of an incoming packet into an 503 LSI or HIT. 505 Effectively, LSIs improve IPv6 interoperability at the network layer 506 as described in the first scenario and at the application layer as 507 depicted in the second example. The interoperability mechanism 508 should not be used to avoid transition to IPv6; the authors firmly 509 believe in IPv6 adoption and encourage developers to port existing 510 IPv4-only applications to use IPv6. However, some proprietary, 511 closed-source, IPv4-only applications may never see the daylight of 512 IPv6, and the LSI mechanism is suitable for extending the lifetime of 513 such applications even in IPv6-only networks. 515 The main disadvantage of an LSI is its local scope. Applications may 516 violate layering principles and pass LSIs to each other in 517 application-layer protocols. As the LSIs are valid only in the 518 context of the local host, they may represent an entirely different 519 host when passed to another host. However, it should be emphasized 520 here that the LSI concept is effectively a host-based NAT and does 521 not introduce any more issues than the prevalent middlebox based NATs 522 for IPv4. In other words, the applications violating layering 523 principles are already broken by the NAT boxes that are ubiquitously 524 deployed. 526 4.5. Storing Host Identifiers in directories 528 The public Host Identifiers should be stored in DNS; the unpublished 529 Host Identifiers should not be stored anywhere (besides the 530 communicating hosts themselves). The (public) HI along with the 531 supported HIHs are stored in a new RR type. This RR type is defined 532 in HIP DNS Extension [RFC8005]. 534 Alternatively, or in addition to storing Host Identifiers in the DNS, 535 they may be stored in various other directories. For instance, a 536 directory based on the Lightweight Directory Access Protocol (LDAP) 537 or a Public Key Infrastructure (PKI) [I-D.ietf-hip-rfc6253-bis] may 538 be used. Alternatively, Distributed Hash Tables (DHTs) [RFC6537] 539 have successfully been utilized [RFC6538]. Such a practice may allow 540 them to be used for purposes other than pure host identification. 542 Some types of applications may cache and use Host Identifiers 543 directly, while others may indirectly discover them through symbolic 544 host name (such as FQDN) look up from a directory. Even though Host 545 Identities can have a substantially longer lifetime associated with 546 them than routable IP addresses, directories may be a better approach 547 to manage the lifespan of Host Identities. For example, an LDAP- 548 based directory or DHT can be used for locally published identities 549 whereas DNS can be more suitable for public advertisement. 551 5. New stack architecture 553 One way to characterize Host Identity is to compare the proposed new 554 architecture with the current one. Using the terminology from the 555 IRTF Name Space Research Group Report [nsrg-report] and, e.g., the 556 unpublished Internet-Draft Endpoints and Endpoint Names 557 [chiappa-endpoints], the IP addresses currently embody the dual role 558 of locators and end-point identifiers. That is, each IP address 559 names a topological location in the Internet, thereby acting as a 560 routing direction vector, or locator. At the same time, the IP 561 address names the physical network interface currently located at the 562 point-of-attachment, thereby acting as a end-point name. 564 In the HIP architecture, the end-point names and locators are 565 separated from each other. IP addresses continue to act as locators. 566 The Host Identifiers take the role of end-point identifiers. It is 567 important to understand that the end-point names based on Host 568 Identities are slightly different from interface names; a Host 569 Identity can be simultaneously reachable through several interfaces. 571 The difference between the bindings of the logical entities are 572 illustrated in Figure 1. Left side illustrates the current TCP/IP 573 architecture and right side the HIP-based architecture. 575 Transport ---- Socket Transport ------ Socket 576 association | association | 577 | | 578 | | 579 | | 580 End-point | End-point --- Host Identity 581 \ | | 582 \ | | 583 \ | | 584 \ | | 585 Location --- IP address Location --- IP address 587 Figure 1 589 Architecturally, HIP provides for a different binding of transport- 590 layer protocols. That is, the transport-layer associations, i.e., 591 TCP connections and UDP associations, are no longer bound to IP 592 addresses but rather to Host Identities. In practice, the Host 593 Identities are exposed as LSIs and HITs for legacy applications and 594 the transport layer to facilitate backward compatibility with 595 existing networking APIs and stacks. 597 5.1. On the multiplicity of identities 599 A host may have multiple identities both at the client and server 600 side. This raises some additional concerns that are addressed in 601 this section. 603 For security reasons, it may be a bad idea to duplicate the same Host 604 Identity on multiple hosts because the compromise of a single host 605 taints the identities of the other hosts. Management of machines 606 with identical Host Identities may also present other challenges and, 607 therefore, it is advisable to have a unique identity for each host. 609 Instead of duplicating identities, HIP opportunistic mode can be 610 employed, where the initiator leaves out the identifier of the 611 responder when initiating the key exchange and learns it upon the 612 completion of the exchange. The tradeoffs are related to lowered 613 security guarantees, but a benefit of the approach is to avoid 614 publishing of Host Identifiers in any directories [komu-leap]. The 615 approach could also be used for load balancing purposes at the HIP 616 layer because the identity of the responder can be decided 617 dynamically during the key exchange. Thus, the approach has the 618 potential to be used as a HIP-layer "anycast", either directly 619 between two hosts or indirectly through the rendezvous service 620 [komu-diss]. 622 At the client side, a host may have multiple Host Identities, for 623 instance, for privacy purposes. Another reason can be that the 624 person utilizing the host employs different identities for different 625 administrative domains as an extra security measure. If a HIP-aware 626 middlebox, such as a HIP-based firewall, is on the path between the 627 client and server, the user or the underlying system should carefully 628 choose the correct identity to avoid the firewall to unnecessarily 629 drop HIP-base connectivity [komu-diss]. 631 Similarly, a server may have multiple Host Identities. For instance, 632 a single web server may serve multiple different administrative 633 domains. Typically, the distinction is accomplished based on the DNS 634 name, but also the Host Identity could be used for this purpose. 635 However, a more compelling reason to employ multiple identities are 636 HIP-aware firewalls that are unable see the HTTP traffic inside the 637 encrypted IPsec tunnel. In such a case, each service could be 638 configured with a separate identity, thus allowing the firewall to 639 segregate the different services of the single web server from each 640 other [lindqvist-enterprise]. 642 6. Control plane 644 HIP decouples control and data plane from each other. Two end-hosts 645 initialize the control plane using a key exchange procedure called 646 the base exchange. The procedure can be assisted by new 647 infrastructural intermediaries called rendezvous or relay servers. 648 In the event of IP address changes, the end-hosts sustain control 649 plane connectivity with mobility and multihoming extensions. 650 Eventually, the end-hosts terminate the control plane and remove the 651 associated state. 653 6.1. Base exchange 655 The base exchange is a key exchange procedure that authenticates the 656 initiator and responder to each other using their public keys. 657 Typically, the initiator is the client-side host and the responder is 658 the server-side host. The roles are used by the state machine of a 659 HIP implementation, but discarded upon successful completion. 661 The exchange consists of four messages during which the hosts also 662 create symmetric keys to protect the control plane with Hash-based 663 message authentication codes (HMACs). The keys can be also used to 664 protect the data plane, and IPsec ESP [RFC7402] is typically used as 665 the data-plane protocol, albeit HIP can also accommodate others. 666 Both the control and data plane are terminated using a closing 667 procedure consisting of two messages. 669 In addition, the base exchange also includes a computational puzzle 670 [RFC7401] that the initiator must solve. The responder chooses the 671 difficulty of the puzzle which permits the responder to delay new 672 incoming initiators according to local policies, for instance, when 673 the responder is under heavy load. The puzzle can offer some 674 resiliency against DoS attacks because the design of the puzzle 675 mechanism allows the responder to remain stateless until the very end 676 of the base exchange [aura-dos]. HIP puzzles have also been studied 677 under steady-state DDoS attacks [beal-dos], on multiple adversary 678 models with varying puzzle difficulties [tritilanunt-dos] and with 679 ephemeral Host Identities [komu-mitigation]. 681 6.2. End-host mobility and multi-homing 683 HIP decouples the transport from the internetworking layer, and binds 684 the transport associations to the Host Identities (actually through 685 either the HIT or LSI). After the initial key exchange, the HIP 686 layer maintains transport-layer connectivity and data flows using its 687 mobility [RFC8046] and multihoming [RFC8047] extensions. 688 Consequently, HIP can provide for a degree of internetworking 689 mobility and multi-homing at a low infrastructure cost. HIP mobility 690 includes IP address changes (via any method) to either party. Thus, 691 a system is considered mobile if its IP address can change 692 dynamically for any reason like PPP, DHCP, IPv6 prefix reassignments, 693 or a NAT device remapping its translation. Likewise, a system is 694 considered multi-homed if it has more than one globally routable IP 695 address at the same time. HIP links IP addresses together, when 696 multiple IP addresses correspond to the same Host Identity. If one 697 address becomes unusable, or a more preferred address becomes 698 available, existing transport associations can easily be moved to 699 another address. 701 When a node moves while communication is already on-going, address 702 changes are rather straightforward. The peer of the mobile node can 703 just accept a HIP or an integrity protected ESP packet from any 704 address and ignore the source address. However, as discussed in 705 Section 12.2 below, a mobile node must send a HIP UPDATE packet to 706 inform the peer of the new address(es), and the peer must verify that 707 the mobile node is reachable through these addresses. This is 708 especially helpful for those situations where the peer node is 709 sending data periodically to the mobile node (that is, re-starting a 710 connection after the initial connection). 712 6.3. Rendezvous mechanism 714 Establishing a contact to a mobile, moving node is slightly more 715 involved. In order to start the HIP exchange, the initiator node has 716 to know how to reach the mobile node. For instance, the mobile node 717 can employ Dynamic DNS [RFC2136] to update its reachability 718 information in the DNS. To avoid the dependency to DNS, HIP provides 719 its own HIP-specific alternative: the HIP rendezvous mechanism as 720 defined in HIP Rendezvous specifications [RFC8004]. 722 Using the HIP rendezvous extensions, the mobile node keeps the 723 rendezvous infrastructure continuously updated with its current IP 724 address(es). The mobile nodes trusts the rendezvous mechanism in 725 order to properly maintain their HIT and IP address mappings. 727 The rendezvous mechanism is especially useful in scenarios where both 728 of the nodes are expected to change their address at the same time. 729 In such a case, the HIP UPDATE packets will cross each other in the 730 network and never reach the peer node. 732 6.4. Relay mechanism 734 The HIP relay mechanism [I-D.ietf-hip-native-nat-traversal] is an 735 alternative to the HIP rendezvous mechanism. The HIP relay mechanism 736 is more suitable for IPv4 networks with NATs because a HIP relay can 737 forward all control and data plane communications in order to 738 guarantee successful NAT traversal. 740 6.5. Termination of the control plane 742 The control plane between two hosts is terminated using a secure two 743 message exchange as specified in base exchange specification 744 [RFC7401]. The related state (i.e. host associations) should be 745 removed upon successful termination. 747 7. Data plane 749 The encapsulation format for the data plane used for carrying the 750 application-layer traffic can be dynamically negotiated during the 751 key exchange. For instance, HICCUPS extensions [RFC6078] define one 752 way to transport application-layer datagrams directly over the HIP 753 control plane, protected by asymmetric key cryptography. Also, S-RTP 754 has been considered as the data encapsulation protocol [hip-srtp]. 755 However, the most widely implemented method is the Encapsulated 756 Security Payload (ESP) [RFC7402] that is protected by symmetric keys 757 derived during the key exchange. ESP Security Associations (SAs) 758 offer both confidentiality and integrity protection, of which the 759 former can be disabled during the key exchange. In the future, other 760 ways of transporting application-layer data may be defined. 762 The ESP SAs are established and terminated between the initiator and 763 the responder hosts. Usually, the hosts create at least two SAs, one 764 in each direction (initiator-to-responder SA and responder-to- 765 initiator SA). If the IP addresses of either host changes, the HIP 766 mobility extensions can be used to re-negotiate the corresponding 767 SAs. 769 On the wire, the difference in the use of identifiers between the HIP 770 control and data plane is that the HITs are included in all control 771 packets, but not in the data plane when ESP is employed. Instead, 772 the ESP employs SPI numbers that act as compressed HITs. Any HIP- 773 aware middlebox (for instance, a HIP-aware firewall) interested in 774 the ESP based data plane should keep track between the control and 775 data plane identifiers in order to associate them with each other. 777 Since HIP does not negotiate any SA lifetimes, all lifetimes are 778 subject to local policy. The only lifetimes a HIP implementation 779 must support are sequence number rollover (for replay protection), 780 and SA timeout. An SA times out if no packets are received using 781 that SA. Implementations may support lifetimes for the various ESP 782 transforms and other data-plane protocols. 784 8. HIP and NATs 786 Passing packets between different IP addressing realms requires 787 changing IP addresses in the packet header. This may occur, for 788 example, when a packet is passed between the public Internet and a 789 private address space, or between IPv4 and IPv6 networks. The 790 address translation is usually implemented as Network Address 791 Translation (NAT) [RFC3022] or NAT Protocol translation (NAT-PT) 792 [RFC2766]. 794 In a network environment where identification is based on the IP 795 addresses, identifying the communicating nodes is difficult when NATs 796 are employed because private address spaces are overlapping. In 797 other words, two hosts cannot be distinguished from each other solely 798 based on their IP address. With HIP, the transport-layer end-points 799 (i.e. applications) are bound to unique Host Identities rather than 800 overlapping private addresses. This allows two end-points to 801 distinguish one other even when they are located in different private 802 address realms. Thus, the IP addresses are used only for routing 803 purposes and can be changed freely by NATs when a packet between two 804 HIP capable hosts traverses through multiple private address realms. 806 NAT traversal extensions for HIP [I-D.ietf-hip-native-nat-traversal] 807 can be used to realize the actual end-to-end connectivity through NAT 808 devices. To support basic backward compatibility with legacy NATs, 809 the extensions encapsulate both HIP control and data plane in UDP. 810 The extensions define mechanisms for forwarding the two planes 811 through an intermediary host called HIP relay and procedures to 812 establish direct end-to-end connectivity by penetrating NATs. 814 Besides this "native" NAT traversal mode for HIP, other NAT traversal 815 mechanisms have been successfully utilized, such as Teredo 816 [varjonen-split]. 818 Besides legacy NATs, a HIP-aware NAT has been designed and 819 implemented [ylitalo-spinat]. For a HIP-based flow, a HIP-aware NAT 820 or NAT-PT system tracks the mapping of HITs, and the corresponding 821 ESP SPIs, to an IP address. The NAT system has to learn mappings 822 both from HITs and from SPIs to IP addresses. Many HITs (and SPIs) 823 can map to a single IP address on a NAT, simplifying connections on 824 address poor NAT interfaces. The NAT can gain much of its knowledge 825 from the HIP packets themselves; however, some NAT configuration may 826 be necessary. 828 8.1. HIP and Upper-layer checksums 830 There is no way for a host to know if any of the IP addresses in an 831 IP header are the addresses used to calculate the TCP checksum. That 832 is, it is not feasible to calculate the TCP checksum using the actual 833 IP addresses in the pseudo header; the addresses received in the 834 incoming packet are not necessarily the same as they were on the 835 sending host. Furthermore, it is not possible to recompute the 836 upper-layer checksums in the NAT/NAT-PT system, since the traffic is 837 ESP protected. Consequently, the TCP and UDP checksums are 838 calculated using the HITs in the place of the IP addresses in the 839 pseudo header. Furthermore, only the IPv6 pseudo header format is 840 used. This provides for IPv4 / IPv6 protocol translation. 842 9. Multicast 844 A number of studies investigating HIP-based multicast have been 845 published (including [shields-hip], [xueyong-hip], [xueyong-hip], 846 [amir-hip], [kovacshazi-host] and [xueyong-secure]). In particular, 847 so-called Bloom filters, that allow compressing of multiple labels 848 into small data structures, may be a promising way forward 849 [sarela-bloom]. However, the different schemes have not been adopted 850 by the HIP working group (nor the HIP research group in IRTF), so the 851 details are not further elaborated here. 853 10. HIP policies 855 There are a number of variables that influence the HIP exchange that 856 each host must support. All HIP implementations should support at 857 least 2 HIs, one to publish in DNS or similar directory service and 858 an unpublished one for anonymous usage. Although unpublished HIs 859 will be rarely used as responder HIs, they are likely to be common 860 for initiators. Support for multiple HIs is recommended. This 861 provides new challenges for systems or users to decide which type of 862 HI to expose when they start a new session. 864 Opportunistic mode (where the initiator starts a HIP exchange without 865 prior knowledge of the responder's HI) presents a security tradeoff. 866 At the expense of being subject to MITM attacks, the opportunistic 867 mode allows the initiator to learn the identity of the responder 868 during communication rather than from an external directory. 869 Opportunistic mode can be used for registration to HIP-based services 870 [RFC8003] (i.e. utilized by HIP for its own internal purposes) or by 871 the application layer [komu-leap]. For security reasons, especially 872 the latter requires some involvement from the user to accept the 873 identity of the responder similar to how SSH prompts the user when 874 connecting to a server for the first time [pham-leap]. In practice, 875 this can be realized in end-host based firewalls in the case of 876 legacy applications [karvonen-usable] or with native APIs for HIP 877 APIs [RFC6317] in the case of HIP-aware applications. 879 Many initiators would want to use a different HI for different 880 responders. The implementations should provide for a policy mapping 881 of initiator HITs to responder HITs. This policy should also include 882 preferred transforms and local lifetimes. 884 Responders would need a similar policy, describing the hosts allowed 885 to participate in HIP exchanges, and the preferred transforms and 886 local lifetimes. 888 11. Design considerations 890 11.1. Benefits of HIP 892 In the beginning, the network layer protocol (i.e., IP) had the 893 following four "classic" invariants: 895 1. Non-mutable: The address sent is the address received. 897 2. Non-mobile: The address doesn't change during the course of an 898 "association". 900 3. Reversible: A return header can always be formed by reversing the 901 source and destination addresses. 903 4. Omniscient: Each host knows what address a partner host can use 904 to send packets to it. 906 Actually, the fourth can be inferred from 1 and 3, but it is worth 907 mentioning explicitly for reasons that will be obvious soon if not 908 already. 910 In the current "post-classic" world, we are intentionally trying to 911 get rid of the second invariant (both for mobility and for multi- 912 homing), and we have been forced to give up the first and the fourth. 913 Realm Specific IP [RFC3102] is an attempt to reinstate the fourth 914 invariant without the first invariant. IPv6 is an attempt to 915 reinstate the first invariant. 917 Few client-side systems on the Internet have DNS names that are 918 meaningful. That is, if they have a Fully Qualified Domain Name 919 (FQDN), that name typically belongs to a NAT device or a dial-up 920 server, and does not really identify the system itself but its 921 current connectivity. FQDNs (and their extensions as email names) 922 are application-layer names; more frequently naming services than 923 particular systems. This is why many systems on the Internet are not 924 registered in the DNS; they do not have services of interest to other 925 Internet hosts. 927 DNS names are references to IP addresses. This only demonstrates the 928 interrelationship of the networking and application layers. DNS, as 929 the Internet's only deployed and distributed database, is also the 930 repository of other namespaces, due in part to DNSSEC and application 931 specific key records. Although each namespace can be stretched (IP 932 with v6, DNS with KEY records), neither can adequately provide for 933 host authentication or act as a separation between internetworking 934 and transport layers. 936 The Host Identity (HI) namespace fills an important gap between the 937 IP and DNS namespaces. An interesting thing about the HI is that it 938 actually allows a host to give up all but the 3rd network-layer 939 invariant. That is to say, as long as the source and destination 940 addresses in the network-layer protocol are reversible, HIP takes 941 care of host identification, and reversibility allows a local host to 942 receive a packet back from a remote host. The address changes 943 occurring during NAT transit (non-mutable) or host movement (non- 944 omniscient or non-mobile) can be managed by the HIP layer. 946 With the exception of High-Performance Computing applications, the 947 Sockets API is the most common way to develop network applications. 948 Applications use the Sockets API either directly or indirectly 949 through some libraries or frameworks. However, the Sockets API is 950 based on the assumption of static IP addresses, and DNS with its 951 lifetime values was invented at later stages during the evolution of 952 the Internet. Hence, the Sockets API does not deal with the lifetime 953 of addresses [RFC6250]. As the majority of the end-user equipment is 954 mobile today, their addresses are effectively ephemeral, but the 955 Sockets API still gives a fallacious illusion of persistent IP 956 addresses to the unwary developer. HIP can be used to solidify this 957 illusion because HIP provides persistent surrogate addresses to the 958 application layer in the form of LSIs and HITs. 960 The persistent identifiers as provided by HIP are useful in multiple 961 scenarios (see, e.g., [ylitalo-diss] or [komu-diss], for a more 962 elaborate discussion): 964 o When a mobile host moves physically between two different WLAN 965 networks and obtains a new address, an application using the 966 identifiers remains isolated regardless of the topology changes 967 while the underlying HIP layer re-establishes connectivity (i.e. a 968 horizontal handoff). 970 o Similarly, the application utilizing the identifiers remains again 971 unaware of the topological changes when the underlying host 972 equipped with WLAN and cellular network interfaces switches 973 between the two different access technologies (i.e. a vertical 974 handoff). 976 o Even when hosts are located in private address realms, 977 applications can uniquely distinguish different hosts from each 978 other based on their identifiers. In other words, it can be 979 stated that HIP improves Internet transparency for the application 980 layer [komu-diss]. 982 o Site renumbering events for services can occur due to corporate 983 mergers or acquisitions, or by changes in Internet Service 984 Provider. They can involve changing the entire network prefix of 985 an organization, which is problematic due to hard-coded addresses 986 in service configuration files or cached IP addresses at the 987 client side [RFC5887]. Considering such human errors, a site 988 employing location-independent identifiers as promoted by HIP may 989 experience less problems while renumbering their network. 991 o More agile IPv6 interoperability can be achieved, as discussed in 992 Section 4.4. IPv6-based applications can communicate using HITs 993 with IPv4-based applications that are using LSIs. Additionally, 994 the underlying network type (IPv4 or IPv6) becomes independent of 995 the addressing family of the application. 997 o HITs (or LSIs) can be used in IP-based access control lists as a 998 more secure replacement for IPv6 addresses. Besides security, HIT 999 based access control has two other benefits. First, the use of 1000 HITs can potentially halve the size of access control lists 1001 because separate rules for IPv4 are not needed [komu-diss]. 1002 Second, HIT-based configuration rules in HIP-aware middleboxes 1003 remain static and independent of topology changes, thus 1004 simplifying administrative efforts particularly for mobile 1005 environments. For instance, the benefits of HIT based access 1006 control have been harnessed in the case of HIP-aware firewalls, 1007 but can be utilized directly at the end-hosts as well [RFC6538]. 1009 While some of these benefits could be and have been redundantly 1010 implemented by individual applications, providing such generic 1011 functionality at the lower layers is useful because it reduces 1012 software development effort and networking software bugs (as the 1013 layer is tested with multiple applications). It also allows the 1014 developer to focus on building the application itself rather than 1015 delving into the intricacies of mobile networking, thus facilitating 1016 separation of concerns. 1018 HIP could also be realized by combining a number of different 1019 protocols, but the complexity of the resulting software may become 1020 substantially larger, and the interaction between multiple possibly 1021 layered protocols may have adverse effects on latency and throughput. 1022 It is also worth noting that virtually nothing prevents realizing the 1023 HIP architecture, for instance, as an application-layer library, 1024 which has been actually implemented in the past [xin-hip-lib]. 1025 However, the tradeoff in moving the HIP layer to the application 1026 layer is that legacy applications may not be supported. 1028 11.2. Drawbacks of HIP 1030 In computer science, many problems can be solved with an extra layer 1031 of indirection. However, the indirection always involves some costs 1032 as there is no such a thing as "free lunch". In the case of HIP, the 1033 main costs could be stated as follows: 1035 o In general, a new layer and a new namespace always involve some 1036 initial effort in terms of implementation, deployment and 1037 maintenance. Some education of developers and administrators may 1038 also be needed. However, the HIP community at the IETF has spent 1039 years in experimenting, exploring, testing, documenting and 1040 implementing HIP to ease the adoption costs. 1042 o HIP decouples identifier and locator roles of IP addresses. 1043 Consequently, a mapping mechanism is needed to associate them 1044 together. A failure to map a HIT to its corresponding locator may 1045 result in failed connectivity because a HIT is "flat" by its 1046 nature and cannot be looked up from the hierarchically organized 1047 DNS. HITs are flat by design due to a security tradeoff. The 1048 more bits are allocated for the hash in the HIT, the less likely 1049 there will be (malicious) collisions. 1051 o From performance viewpoint, HIP control and data plane processing 1052 introduces some overhead in terms of throughput and latency as 1053 elaborated below. 1055 The key exchange introduces some extra latency (two round trips) in 1056 the initial transport layer connection establishment between two 1057 hosts. With TCP, additional delay occurs if the underlying network 1058 stack implementation drops the triggering SYN packet during the key 1059 exchange. The same cost may also occur during HIP handoff 1060 procedures. However, subsequent TCP sessions using the same HIP 1061 association will not bear this cost (within the key lifetime). Both 1062 the key exchange and handoff penalties can be minimized by caching 1063 TCP packets. The latter case can further be optimized with TCP user 1064 timeout extensions [RFC5482] as described in further detail by 1065 Schuetz et al [schuetz-intermittent]. 1067 The most CPU-intensive operations involve the use of the asymmetric 1068 keys and Diffie-Hellman key derivation at the control plane, but this 1069 occurs only during the key exchange, its maintenance (handoffs, 1070 refreshing of key material) and tear down procedures of HIP 1071 associations. The data plane is typically implemented with ESP 1072 because it has a smaller overhead due to symmetric key encryption. 1073 Naturally, even ESP involves some overhead in terms of latency 1074 (processing costs) and throughput (tunneling) (see e.g. 1075 [ylitalo-diss] for a performance evaluation). 1077 11.3. Deployment and adoption considerations 1079 This section describes some deployment and adoption considerations 1080 related to HIP from a technical perspective. 1082 11.3.1. Deployment analysis 1084 HIP has commercially been utilized at Boeing airplane factory for 1085 their internal purposes [paine-hip]. It has been included in a 1086 security product called Tofino to support layer-two Virtual Private 1087 Networks [henderson-vpls] to facilitate, e.g, supervisory control and 1088 data acquisition (SCADA) security. However, HIP has not been a "wild 1089 success" [RFC5218] in the Internet as argued by Levae et al 1090 [leva-barriers]. Here, we briefly highlight some of their findings 1091 based on interviews with 19 experts from the industry and academia. 1093 From a marketing perspective, the demand for HIP has been low and 1094 substitute technologies have been favored. Another identified reason 1095 has been that some technical misconceptions related to the early 1096 stages of HIP specifications still persist. Two identified 1097 misconceptions are that HIP does not support NAT traversal, and that 1098 HIP must be implemented in the OS kernel. Both of these claims are 1099 untrue; HIP does have NAT traversal extensions 1100 [I-D.ietf-hip-native-nat-traversal], and kernel modifications can be 1101 avoided with modern operating systems by diverting packets for 1102 userspace processing. 1104 The analysis by Levae et al clarifies infrastructural requirements 1105 for HIP. In a minimal set up, a client and server machine have to 1106 run HIP software. However, to avoid manual configurations, usually 1107 DNS records for HIP are set up. For instance, the popular DNS server 1108 software Bind9 does not require any changes to accommodate DNS 1109 records for HIP because they can be supported in binary format in its 1110 configuration files [RFC6538]. HIP rendezvous servers and firewalls 1111 are optional. No changes are required to network address points, 1112 NATs, edge routers or core networks. HIP may require holes in legacy 1113 firewalls. 1115 The analysis also clarifies the requirements for the host components 1116 that consist of three parts. First, a HIP control plane component is 1117 required, typically implemented as a userspace daemon. Second, a 1118 data plane component is needed. Most HIP implementations utilize the 1119 so called BEET mode of ESP that has been available since Linux kernel 1120 2.6.27, but is included also as a userspace component in a few of the 1121 implementations. Third, HIP systems usually provide a DNS proxy for 1122 the local host that translates HIP DNS records to LSIs and HITs, and 1123 communicates the corresponding locators to HIP userspace daemon. 1124 While the third component is not mandatory, it is very useful for 1125 avoiding manual configurations. The three components are further 1126 described in the HIP experiment report [RFC6538]. 1128 Based on the interviews, Levae et al suggest further directions to 1129 facilitate HIP deployment. Transitioning the HIP specifications to 1130 the standards track may help, but other measures could be taken. As 1131 a more radical measure, the authors suggest to implement HIP as a 1132 purely application-layer library [xin-hip-lib] or other kind of 1133 middleware. On the other hand, more conservative measures include 1134 focusing on private deployments controlled by a single stakeholder. 1135 As a more concrete example of such a scenario, HIP could be used by a 1136 single service provider to facilitate secure connectivity between its 1137 servers [komu-cloud]. 1139 11.3.2. HIP in 802.15.4 networks 1141 The IEEE 802 standards have been defining MAC layered security. Many 1142 of these standards use EAP [RFC3748] as a Key Management System (KMS) 1143 transport, but some like IEEE 802.15.4 [IEEE.802-15-4.2011] leave the 1144 KMS and its transport as "Out of Scope". 1146 HIP is well suited as a KMS in these environments: 1148 o HIP is independent of IP addressing and can be directly 1149 transported over any network protocol. 1151 o Master Keys in 802 protocols are commonly pair-based with group 1152 keys transported from the group controller using pair-wise keys. 1154 o AdHoc 802 networks can be better served by a peer-to-peer KMS than 1155 the EAP client/server model. 1157 o Some devices are very memory constrained and a common KMS for both 1158 MAC and IP security represents a considerable code savings. 1160 11.3.3. HIP and Internet of Things 1162 HIP requires certain amount computational resources from a device due 1163 to cryptographic processing. HIP scales down to phones and small 1164 system-on-chip devices (such as Raspberry Pis, Intel Edison), but 1165 small sensors operating with small batteries have remained 1166 problematic. Different extensions to the HIP have been developed to 1167 scale HIP down to smaller devices, typically with different security 1168 tradeoffs. For example, the non-cryptographic identifiers have been 1169 proposed in RFID scenarios. The slimfit approach [hummen] proposes a 1170 compression layer for HIP to make it more suitable for constrained 1171 networks. The approach is applied to a light-weight version of HIP 1172 (i.e. "Diet HIP") in order to scale down to small sensors. 1174 The HIP Diet Exchange [I-D.ietf-hip-dex] design aims at reducing the 1175 overhead of the employed cryptographic primitives by omitting public- 1176 key signatures and hash functions. In doing so, the main goal is to 1177 still deliver similar security properties to the Base Exchange (BEX). 1179 DEX is primarily designed for computation or memory- constrained 1180 sensor/actuator devices. Like BEX, it is expected to be used 1181 together with a suitable security protocol such as the Encapsulated 1182 Security Payload (ESP) for the protection of upper layer protocol 1183 data. In addition, DEX can also be used as a keying mechanism for 1184 security primitives at the MAC layer, e.g., for IEEE 802.15.9 1185 networks ([IEEE.802-15-9]. 1187 The main differences between HIP BEX and DEX are: 1189 1. Minimum collection of cryptographic primitives to reduce the 1190 protocol overhead. 1192 * Static Elliptic Curve Diffie-Hellman key pairs for peer 1193 authentication and encryption of the session key. 1195 * AES-CTR for symmetric encryption and AES-CMAC for MACing 1196 function. 1198 * A simple fold function for HIT generation. 1200 2. Forfeit of Perfect Forward Secrecy with the dropping of an 1201 ephemeral Diffie-Hellman key agreement. 1203 3. Forfeit of digital signatures with the removal of a hash 1204 function. Reliance on ECDH derived key used in HIP_MAC to prove 1205 ownership of the private key. 1207 4. Diffie-Hellman derived key ONLY used to protect the HIP packets. 1208 A separate secret exchange within the HIP packets creates the 1209 session key(s). 1211 5. Optional retransmission strategy tailored to handle the 1212 potentially extensive processing time of the employed 1213 cryptographic operations on computationally constrained devices. 1215 11.4. Answers to NSRG questions 1217 The IRTF Name Space Research Group has posed a number of evaluating 1218 questions in their report [nsrg-report]. In this section, we provide 1219 answers to these questions. 1221 1. How would a stack name improve the overall functionality of the 1222 Internet? 1224 HIP decouples the internetworking layer from the transport 1225 layer, allowing each to evolve separately. The decoupling 1226 makes end-host mobility and multi-homing easier, also across 1227 IPv4 and IPv6 networks. HIs make network renumbering easier, 1228 and they also make process migration and clustered servers 1229 easier to implement. Furthermore, being cryptographic in 1230 nature, they provide the basis for solving the security 1231 problems related to end-host mobility and multi-homing. 1233 2. What does a stack name look like? 1235 A HI is a cryptographic public key. However, instead of using 1236 the keys directly, most protocols use a fixed size hash of the 1237 public key. 1239 3. What is its lifetime? 1241 HIP provides both stable and temporary Host Identifiers. 1242 Stable HIs are typically long lived, with a lifetime of years 1243 or more. The lifetime of temporary HIs depends on how long 1244 the upper-layer connections and applications need them, and 1245 can range from a few seconds to years. 1247 4. Where does it live in the stack? 1249 The HIs live between the transport and internetworking layers. 1251 5. How is it used on the end points? 1253 The Host Identifiers may be used directly or indirectly (in 1254 the form of HITs or LSIs) by applications when they access 1255 network services. Additionally, the Host Identifiers, as 1256 public keys, are used in the built in key agreement protocol, 1257 called the HIP base exchange, to authenticate the hosts to 1258 each other. 1260 6. What administrative infrastructure is needed to support it? 1262 In some environments, it is possible to use HIP 1263 opportunistically, without any infrastructure. However, to 1264 gain full benefit from HIP, the HIs must be stored in the DNS 1265 or a PKI, and a new rendezvous mechanism is needed [RFC8005]. 1267 7. If we add an additional layer would it make the address list in 1268 SCTP unnecessary? 1270 Yes 1272 8. What additional security benefits would a new naming scheme 1273 offer? 1275 HIP reduces dependency on IP addresses, making the so called 1276 address ownership [Nik2001] problems easier to solve. In 1277 practice, HIP provides security for end-host mobility and 1278 multi-homing. Furthermore, since HIP Host Identifiers are 1279 public keys, standard public key certificate infrastructures 1280 can be applied on the top of HIP. 1282 9. What would the resolution mechanisms be, or what characteristics 1283 of a resolution mechanisms would be required? 1285 For most purposes, an approach where DNS names are resolved 1286 simultaneously to HIs and IP addresses is sufficient. 1287 However, if it becomes necessary to resolve HIs into IP 1288 addresses or back to DNS names, a flat resolution 1289 infrastructure is needed. Such an infrastructure could be 1290 based on the ideas of Distributed Hash Tables, but would 1291 require significant new development and deployment. 1293 12. Security considerations 1295 This section includes discussion on some issues and solutions related 1296 to security in the HIP architecture. 1298 12.1. MiTM Attacks 1300 HIP takes advantage of the new Host Identity paradigm to provide 1301 secure authentication of hosts and to provide a fast key exchange for 1302 ESP. HIP also attempts to limit the exposure of the host to various 1303 denial-of-service (DoS) and man-in-the-middle (MitM) attacks. In so 1304 doing, HIP itself is subject to its own DoS and MitM attacks that 1305 potentially could be more damaging to a host's ability to conduct 1306 business as usual. 1308 Resource exhausting denial-of-service attacks take advantage of the 1309 cost of setting up a state for a protocol on the responder compared 1310 to the 'cheapness' on the initiator. HIP allows a responder to 1311 increase the cost of the start of state on the initiator and makes an 1312 effort to reduce the cost to the responder. This is done by having 1313 the responder start the authenticated Diffie-Hellman exchange instead 1314 of the initiator, making the HIP base exchange 4 packets long. The 1315 first packet sent by the responder can be prebuilt to further 1316 mitigate the costs. This packet also includes a computational puzzle 1317 that can optionally be used to further delay the initiator, for 1318 instance, when the responder is overloaded. The details are 1319 explained in the base exchange specification [RFC7401]. 1321 Man-in-the-middle (MitM) attacks are difficult to defend against, 1322 without third-party authentication. A skillful MitM could easily 1323 handle all parts of the HIP base exchange, but HIP indirectly 1324 provides the following protection from a MitM attack. If the 1325 responder's HI is retrieved from a signed DNS zone or securely 1326 obtained by some other means, the initiator can use this to 1327 authenticate the signed HIP packets. Likewise, if the initiator's HI 1328 is in a secure DNS zone, the responder can retrieve it and validate 1329 the signed HIP packets. However, since an initiator may choose to 1330 use an unpublished HI, it knowingly risks a MitM attack. The 1331 responder may choose not to accept a HIP exchange with an initiator 1332 using an unknown HI. 1334 Other types of MitM attacks against HIP can be mounted using ICMP 1335 messages that can be used to signal about problems. As a overall 1336 guideline, the ICMP messages should be considered as unreliable 1337 "hints" and should be acted upon only after timeouts. The exact 1338 attack scenarios and countermeasures are described in full detail the 1339 base exchange specification [RFC7401]. 1341 The need to support multiple hashes for generating the HIT from the 1342 HI affords the MitM to mount a potentially powerful downgrade attack 1343 due to the a-priori need of the HIT in the HIP base exchange. The 1344 base exchange has been augmented to deal with such an attack by 1345 restarting on detecting the attack. At worst this would only lead to 1346 a situation in which the base exchange would never finish (or would 1347 be aborted after some retries). As a drawback, this leads to an 1348 6-way base exchange which may seem bad at first. However, since this 1349 only occurs in an attack scenario and since the attack can be handled 1350 (so it is not interesting to mount anymore), we assume the subsequent 1351 messages do not represent a security threat. Since the MitM cannot 1352 be successful with a downgrade attack, these sorts of attacks will 1353 only occur as 'nuisance' attacks. So, the base exchange would still 1354 be usually just four packets even though implementations must be 1355 prepared to protect themselves against the downgrade attack. 1357 In HIP, the Security Association for ESP is indexed by the SPI; the 1358 source address is always ignored, and the destination address may be 1359 ignored as well. Therefore, HIP-enabled Encapsulated Security 1360 Payload (ESP) is IP address independent. This might seem to make 1361 attacking easier, but ESP with replay protection is already as well 1362 protected as possible, and the removal of the IP address as a check 1363 should not increase the exposure of ESP to DoS attacks. 1365 12.2. Protection against flooding attacks 1367 Although the idea of informing about address changes by simply 1368 sending packets with a new source address appears appealing, it is 1369 not secure enough. That is, even if HIP does not rely on the source 1370 address for anything (once the base exchange has been completed), it 1371 appears to be necessary to check a mobile node's reachability at the 1372 new address before actually sending any larger amounts of traffic to 1373 the new address. 1375 Blindly accepting new addresses would potentially lead to flooding 1376 Denial-of-Service attacks against third parties [RFC4225]. In a 1377 distributed flooding attack an attacker opens high volume HIP 1378 connections with a large number of hosts (using unpublished HIs), and 1379 then claims to all of these hosts that it has moved to a target 1380 node's IP address. If the peer hosts were to simply accept the move, 1381 the result would be a packet flood to the target node's address. To 1382 prevent this type of attack, HIP mobility extensions include a return 1383 routability check procedure where the reachability of a node is 1384 separately checked at each address before using the address for 1385 larger amounts of traffic. 1387 A credit-based authorization approach for host mobility with the Host 1388 Identity Protocol [RFC8046] can be used between hosts for sending 1389 data prior to completing the address tests. Otherwise, if HIP is 1390 used between two hosts that fully trust each other, the hosts may 1391 optionally decide to skip the address tests. However, such 1392 performance optimization must be restricted to peers that are known 1393 to be trustworthy and capable of protecting themselves from malicious 1394 software. 1396 12.3. HITs used in ACLs 1398 At end-hosts, HITs can be used in IP-based access control lists at 1399 the application and network layers. At middleboxes, HIP-aware 1400 firewalls [lindqvist-enterprise] can use HITs or public keys to 1401 control both ingress and egress access to networks or individual 1402 hosts, even in the presence of mobile devices because the HITs and 1403 public keys are topologically independent. As discussed earlier in 1404 Section 7, once a HIP session has been established, the SPI value in 1405 an ESP packet may be used as an index, indicating the HITs. In 1406 practice, firewalls can inspect HIP packets to learn of the bindings 1407 between HITs, SPI values, and IP addresses. They can even explicitly 1408 control ESP usage, dynamically opening ESP only for specific SPI 1409 values and IP addresses. The signatures in HIP packets allow a 1410 capable firewall to ensure that the HIP exchange is indeed occurring 1411 between two known hosts. This may increase firewall security. 1413 A potential drawback of HITs in ACLs is their 'flatness' means they 1414 cannot be aggregated, and this could potentially result in larger 1415 table searches in HIP-aware firewalls. A way to optimize this could 1416 be to utilize Bloom filters for grouping of HITs [sarela-bloom]. 1417 However, it should be noted that it is also easier to exclude 1418 individual, misbehaving hosts out when the firewall rules concern 1419 individual HITs rather than groups. 1421 There has been considerable bad experience with distributed ACLs that 1422 contain public key related material, for example, with SSH. If the 1423 owner of a key needs to revoke it for any reason, the task of finding 1424 all locations where the key is held in an ACL may be impossible. If 1425 the reason for the revocation is due to private key theft, this could 1426 be a serious issue. 1428 A host can keep track of all of its partners that might use its HIT 1429 in an ACL by logging all remote HITs. It should only be necessary to 1430 log responder hosts. With this information, the host can notify the 1431 various hosts about the change to the HIT. There have been attempts 1432 to develop a secure method to issue the HIT revocation notice 1433 [zhang-revocation]. 1435 Some of the HIP-aware middleboxes, such as firewalls 1436 [lindqvist-enterprise] or NATs [ylitalo-spinat], may observe the on- 1437 path traffic passively. Such middleboxes are transparent by their 1438 nature and may not get a notification when a host moves to a 1439 different network. Thus, such middleboxes should maintain soft state 1440 and timeout when the control and data plane between two HIP end-hosts 1441 has been idle too long. Correspondingly, the two end-hosts may send 1442 periodically keepalives, such as UPDATE packets or ICMP messages 1443 inside the ESP tunnel, to sustain state at the on-path middleboxes. 1445 One general limitation related to end-to-end encryption is that 1446 middleboxes may not be able to participate to the protection of data 1447 flows. While the issue may affect also other protocols, Heer at al 1448 [heer-end-host] have analyzed the problem in the context of HIP. 1449 More specifically, when ESP is used as the data-plane protocol for 1450 HIP, the association between the control and data plane is weak and 1451 can be exploited under certain assumptions. In the scenario, the 1452 attacker has already gained access to the target network protected by 1453 a HIP-aware firewall, but wants to circumvent the HIP-based firewall. 1454 To achieve this, the attacker passively observes a base exchange 1455 between two HIP hosts and later replays it. This way, the attacker 1456 manages to penetrate the firewall and can use a fake ESP tunnel to 1457 transport its own data. This is possible because the firewall cannot 1458 distinguish when the ESP tunnel is valid. As a solution, HIP-aware 1459 middleboxes may participate to the control plane interaction by 1460 adding random nonce parameters to the control traffic, which the end- 1461 hosts have to sign to guarantee the freshness of the control traffic 1462 [heer-midauth]. As an alternative, extensions for transporting data 1463 plane directly over the control plane can be used [RFC6078]. 1465 12.4. Alternative HI considerations 1467 The definition of the Host Identifier states that the HI need not be 1468 a public key. It implies that the HI could be any value; for example 1469 a FQDN. This document does not describe how to support such a non- 1470 cryptographic HI, but examples of such protocol variants do exist 1471 ([urien-rfid], [urien-rfid-draft]). A non-cryptographic HI would 1472 still offer the services of the HIT or LSI for NAT traversal. It 1473 would be possible to carry HITs in HIP packets that had neither 1474 privacy nor authentication. Such schemes may be employed for 1475 resource constrained devices, such as small sensors operating on 1476 battery power, but are not further analyzed here. 1478 If it is desirable to use HIP in a low security situation where 1479 public key computations are considered expensive, HIP can be used 1480 with very short Diffie-Hellman and Host Identity keys. Such use 1481 makes the participating hosts vulnerable to MitM and connection 1482 hijacking attacks. However, it does not cause flooding dangers, 1483 since the address check mechanism relies on the routing system and 1484 not on cryptographic strength. 1486 13. IANA considerations 1488 This document has no actions for IANA. 1490 14. Acknowledgments 1492 For the people historically involved in the early stages of HIP, see 1493 the Acknowledgments section in the Host Identity Protocol 1494 specification. 1496 During the later stages of this document, when the editing baton was 1497 transferred to Pekka Nikander, the comments from the early 1498 implementers and others, including Jari Arkko, Tom Henderson, Petri 1499 Jokela, Miika Komu, Mika Kousa, Andrew McGregor, Jan Melen, Tim 1500 Shepard, Jukka Ylitalo, Sasu Tarkoma, and Jorma Wall, were 1501 invaluable. Also, the comments from Lars Eggert, Spencer Dawkins and 1502 Dave Crocker were also useful. 1504 The authors want to express their special thanks to Tom Henderson, 1505 who took the burden of editing the document in response to IESG 1506 comments at the time when both of the authors were busy doing other 1507 things. Without his perseverance original document might have never 1508 made it as RFC4423. 1510 This main effort to update and move HIP forward within the IETF 1511 process owes its impetuous to a number of HIP development teams. The 1512 authors are grateful for Boeing, Helsinki Institute for Information 1513 Technology (HIIT), NomadicLab of Ericsson, and the three 1514 universities: RWTH Aachen, Aalto and University of Helsinki, for 1515 their efforts. Without their collective efforts HIP would have 1516 withered as on the IETF vine as a nice concept. 1518 Thanks also for Suvi Koskinen for her help with proofreading and with 1519 the reference jungle. 1521 15. Changes from RFC 4423 1523 In a nutshell, the changes from RFC 4423 [RFC4423] are mostly 1524 editorial, including clarifications on topics described in a 1525 difficult way and omitting some of the non-architectural 1526 (implementation) details that are already described in other 1527 documents. A number of missing references to the literature were 1528 also added. New topics include the drawbacks of HIP, discussion on 1529 802.15.4 and MAC security, HIP for IoT scenarios, deployment 1530 considerations and description of the base exchange. 1532 16. References 1534 16.1. Normative References 1536 [I-D.ietf-hip-dex] 1537 Moskowitz, R. and R. Hummen, "HIP Diet EXchange (DEX)", 1538 draft-ietf-hip-dex-05 (work in progress), February 2017. 1540 [I-D.ietf-hip-native-nat-traversal] 1541 Keranen, A., Melen, J., and M. Komu, "Native NAT Traversal 1542 Mode for the Host Identity Protocol", draft-ietf-hip- 1543 native-nat-traversal-16 (work in progress), February 2017. 1545 [I-D.ietf-hip-rfc6253-bis] 1546 Heer, T. and S. Varjonen, "Host Identity Protocol 1547 Certificates", draft-ietf-hip-rfc6253-bis-09 (work in 1548 progress), July 2016. 1550 [RFC5482] Eggert, L. and F. Gont, "TCP User Timeout Option", 1551 RFC 5482, DOI 10.17487/RFC5482, March 2009, 1552 . 1554 [RFC7343] Laganier, J. and F. Dupont, "An IPv6 Prefix for Overlay 1555 Routable Cryptographic Hash Identifiers Version 2 1556 (ORCHIDv2)", RFC 7343, DOI 10.17487/RFC7343, September 1557 2014, . 1559 [RFC7401] Moskowitz, R., Ed., Heer, T., Jokela, P., and T. 1560 Henderson, "Host Identity Protocol Version 2 (HIPv2)", 1561 RFC 7401, DOI 10.17487/RFC7401, April 2015, 1562 . 1564 [RFC7402] Jokela, P., Moskowitz, R., and J. Melen, "Using the 1565 Encapsulating Security Payload (ESP) Transport Format with 1566 the Host Identity Protocol (HIP)", RFC 7402, 1567 DOI 10.17487/RFC7402, April 2015, 1568 . 1570 [RFC8003] Laganier, J. and L. Eggert, "Host Identity Protocol (HIP) 1571 Registration Extension", RFC 8003, DOI 10.17487/RFC8003, 1572 October 2016, . 1574 [RFC8004] Laganier, J. and L. Eggert, "Host Identity Protocol (HIP) 1575 Rendezvous Extension", RFC 8004, DOI 10.17487/RFC8004, 1576 October 2016, . 1578 [RFC8005] Laganier, J., "Host Identity Protocol (HIP) Domain Name 1579 System (DNS) Extension", RFC 8005, DOI 10.17487/RFC8005, 1580 October 2016, . 1582 [RFC8046] Henderson, T., Ed., Vogt, C., and J. Arkko, "Host Mobility 1583 with the Host Identity Protocol", RFC 8046, 1584 DOI 10.17487/RFC8046, February 2017, 1585 . 1587 [RFC8047] Henderson, T., Ed., Vogt, C., and J. Arkko, "Host 1588 Multihoming with the Host Identity Protocol", RFC 8047, 1589 DOI 10.17487/RFC8047, February 2017, 1590 . 1592 16.2. Informative references 1594 [amir-hip] 1595 Amir, K., Forsgren, H., Grahn, K., Karvi, T., and G. 1596 Pulkkis, "Security and Trust of Public Key Cryptography 1597 for HIP and HIP Multicast", International Journal of 1598 Dependable and Trustworthy Information Systems (IJDTIS), 1599 2(3), 17-35, DOI: 10.4018/jdtis.2011070102, 2013. 1601 [aura-dos] 1602 Aura, T., Nikander, P., and J. Leiwo, "DOS-resistant 1603 Authentication with Client Puzzles", 8th International 1604 Workshop on Security Protocols, pages 170-177. Springer, , 1605 April 2001. 1607 [beal-dos] 1608 Beal, J. and T. Shephard, "Deamplification of DoS Attacks 1609 via Puzzles", , October 2004. 1611 [chiappa-endpoints] 1612 Chiappa, J., "Endpoints and Endpoint Names: A Proposed 1613 Enhancement to the Internet Architecture", 1614 URL http://www.chiappa.net/~jnc/tech/endpoints.txt, 1999. 1616 [heer-end-host] 1617 Heer, T., Hummen, R., Komu, M., Goetz, S., and K. Wehre, 1618 "End-host Authentication and Authorization for Middleboxes 1619 based on a Cryptographic Namespace", ICC2009 Communication 1620 and Information Systems Security Symposium, , 2009. 1622 [heer-midauth] 1623 Heer, T. and M. Komu, "End-Host Authentication for HIP 1624 Middleboxes", Working draft draft-heer-hip-middle-auth-02, 1625 September 2009. 1627 [henderson-vpls] 1628 Henderson, T. and D. Mattes, "HIP-based Virtual Private 1629 LAN Service (HIPLS)", Working draft draft-henderson-hip- 1630 vpls-07, Dec 2013. 1632 [hip-srtp] 1633 Tschofenig, H., Muenz, F., and M. Shanmugam, "Using SRTP 1634 transport format with HIP", Working draft draft- 1635 tschofenig-hiprg-hip-srtp-01, October 2005. 1637 [hummen] Hummen, R., Hiller, J., Henze, M., and K. Wehrle, "Slimfit 1638 - A HIP DEX Compression Layer for the IP-based Internet of 1639 Things", Wireless and Mobile Computing, Networking and 1640 Communications (WiMob), 2013 IEEE 9th International 1641 Conference on , page 259-266. DOI: 1642 10.1109/WiMOB.2013.6673370, October 2013. 1644 [IEEE.802-15-4.2011] 1645 "Information technology - Telecommunications and 1646 information exchange between systems - Local and 1647 metropolitan area networks - Specific requirements - Part 1648 15.4: Wireless Medium Access Control (MAC) and Physical 1649 Layer (PHY) Specifications for Low-Rate Wireless Personal 1650 Area Networks (WPANs)", IEEE Standard 802.15.4, September 1651 2011, . 1654 [IEEE.802-15-9] 1655 "IEEE Draft Recommended Practice for Transort of Key 1656 Management Protocol (KMP) Datagrams", IEEE P802.15.9/D04, 1657 May 2015. 1659 [karvonen-usable] 1660 Karvonen, K., Komu, M., and A. Gurtov, "Usable Security 1661 Management with Host Identity Protocol", 7th ACS/IEEE 1662 International Conference on Computer Systems and 1663 Applications, (AICCSA-2009), 2009. 1665 [komu-cloud] 1666 Komu, M., Sethi, M., Mallavarapu, R., Oirola, H., Khan, 1667 R., and S. Tarkoma, "Secure Networking for Virtual 1668 Machines in the Cloud", International Workshop on Power 1669 and QoS Aware Computing (PQoSCom2012), IEEE, ISBN: 1670 978-1-4244-8567-3, September 2012. 1672 [komu-diss] 1673 Komu, M., "A Consolidated Namespace for Network 1674 Applications, Developers, Administrators and Users", 1675 Dissertation, Aalto University, Espoo, Finland ISBN: 1676 978-952-60-4904-5 (printed), ISBN: 978-952-60-4905-2 1677 (electronic). , December 2012. 1679 [komu-leap] 1680 Komu, M. and J. Lindqvist, "Leap-of-Faith Security is 1681 Enough for IP Mobility", 6th Annual IEEE Consumer 1682 Communications and Networking Conference IEEE CCNC 2009, 1683 Las Vegas, Nevada, , January 2009. 1685 [komu-mitigation] 1686 Komu, M., Tarkoma, S., and A. Lukyanenko, "Mitigation of 1687 Unsolicited Traffic Across Domains with Host Identities 1688 and Puzzles", 15th Nordic Conference on Secure IT Systems 1689 (NordSec 2010), Springer Lecture Notes in Computer 1690 Science, Volume 7127, pp. 33-48, ISBN: 978-3-642-27936-2, 1691 October 2010. 1693 [kovacshazi-host] 1694 Kovacshazi, Z. and R. Vida, "Host Identity Specific 1695 Multicast", International conference on Networking and 1696 Services (ICNS'06), IEEE Computer Society, Los Alamitos, 1697 CA, USA, http://doi.ieeecomputersociety.org/10.1109/ 1698 ICNS.2007.66, 2007. 1700 [leva-barriers] 1701 Levae, A., Komu, M., and S. Luukkainen, "Adoption Barriers 1702 of Network-layer Protocols: the Case of Host Identity 1703 Protocol", The International Journal of Computer and 1704 Telecommunications Networking, ISSN: 1389-1286, March 1705 2013. 1707 [lindqvist-enterprise] 1708 Lindqvist, J., Vehmersalo, E., Manner, J., and M. Komu, 1709 "Enterprise Network Packet Filtering for Mobile 1710 Cryptographic Identities", International Journal of 1711 Handheld Computing Research, 1 (1), 79-94, , January-March 1712 2010. 1714 [Nik2001] Nikander, P., "Denial-of-Service, Address Ownership, and 1715 Early Authentication in the IPv6 World", in Proceesings 1716 of Security Protocols, 9th International Workshop, 1717 Cambridge, UK, April 25-27 2001, LNCS 2467, pp. 12-26, 1718 Springer, 2002. 1720 [nsrg-report] 1721 Lear, E. and R. Droms, "What's In A Name:Thoughts from the 1722 NSRG", draft-irtf-nsrg-report-10 (work in progress), 1723 September 2003. 1725 [paine-hip] 1726 Paine, R., "Beyond HIP: The End to Hacking As We Know It", 1727 BookSurge Publishing, ISBN: 1439256047, 9781439256046, 1728 2009. 1730 [pham-leap] 1731 Pham, V. and T. Aura, "Security Analysis of Leap-of-Faith 1732 Protocols", Seventh ICST International Conference on 1733 Security and Privacy for Communication Networks, , 1734 September 2011. 1736 [RFC2136] Vixie, P., Ed., Thomson, S., Rekhter, Y., and J. Bound, 1737 "Dynamic Updates in the Domain Name System (DNS UPDATE)", 1738 RFC 2136, DOI 10.17487/RFC2136, April 1997, 1739 . 1741 [RFC2535] Eastlake 3rd, D., "Domain Name System Security 1742 Extensions", RFC 2535, DOI 10.17487/RFC2535, March 1999, 1743 . 1745 [RFC2766] Tsirtsis, G. and P. Srisuresh, "Network Address 1746 Translation - Protocol Translation (NAT-PT)", RFC 2766, 1747 DOI 10.17487/RFC2766, February 2000, 1748 . 1750 [RFC3022] Srisuresh, P. and K. Egevang, "Traditional IP Network 1751 Address Translator (Traditional NAT)", RFC 3022, 1752 DOI 10.17487/RFC3022, January 2001, 1753 . 1755 [RFC3102] Borella, M., Lo, J., Grabelsky, D., and G. Montenegro, 1756 "Realm Specific IP: Framework", RFC 3102, 1757 DOI 10.17487/RFC3102, October 2001, 1758 . 1760 [RFC3748] Aboba, B., Blunk, L., Vollbrecht, J., Carlson, J., and H. 1761 Levkowetz, Ed., "Extensible Authentication Protocol 1762 (EAP)", RFC 3748, DOI 10.17487/RFC3748, June 2004, 1763 . 1765 [RFC4225] Nikander, P., Arkko, J., Aura, T., Montenegro, G., and E. 1766 Nordmark, "Mobile IP Version 6 Route Optimization Security 1767 Design Background", RFC 4225, DOI 10.17487/RFC4225, 1768 December 2005, . 1770 [RFC4306] Kaufman, C., Ed., "Internet Key Exchange (IKEv2) 1771 Protocol", RFC 4306, DOI 10.17487/RFC4306, December 2005, 1772 . 1774 [RFC4423] Moskowitz, R. and P. Nikander, "Host Identity Protocol 1775 (HIP) Architecture", RFC 4423, DOI 10.17487/RFC4423, May 1776 2006, . 1778 [RFC5218] Thaler, D. and B. Aboba, "What Makes for a Successful 1779 Protocol?", RFC 5218, DOI 10.17487/RFC5218, July 2008, 1780 . 1782 [RFC5338] Henderson, T., Nikander, P., and M. Komu, "Using the Host 1783 Identity Protocol with Legacy Applications", RFC 5338, 1784 DOI 10.17487/RFC5338, September 2008, 1785 . 1787 [RFC5887] Carpenter, B., Atkinson, R., and H. Flinck, "Renumbering 1788 Still Needs Work", RFC 5887, DOI 10.17487/RFC5887, May 1789 2010, . 1791 [RFC6078] Camarillo, G. and J. Melen, "Host Identity Protocol (HIP) 1792 Immediate Carriage and Conveyance of Upper-Layer Protocol 1793 Signaling (HICCUPS)", RFC 6078, DOI 10.17487/RFC6078, 1794 January 2011, . 1796 [RFC6250] Thaler, D., "Evolution of the IP Model", RFC 6250, 1797 DOI 10.17487/RFC6250, May 2011, 1798 . 1800 [RFC6281] Cheshire, S., Zhu, Z., Wakikawa, R., and L. Zhang, 1801 "Understanding Apple's Back to My Mac (BTMM) Service", 1802 RFC 6281, DOI 10.17487/RFC6281, June 2011, 1803 . 1805 [RFC6317] Komu, M. and T. Henderson, "Basic Socket Interface 1806 Extensions for the Host Identity Protocol (HIP)", 1807 RFC 6317, DOI 10.17487/RFC6317, July 2011, 1808 . 1810 [RFC6537] Ahrenholz, J., "Host Identity Protocol Distributed Hash 1811 Table Interface", RFC 6537, DOI 10.17487/RFC6537, February 1812 2012, . 1814 [RFC6538] Henderson, T. and A. Gurtov, "The Host Identity Protocol 1815 (HIP) Experiment Report", RFC 6538, DOI 10.17487/RFC6538, 1816 March 2012, . 1818 [sarela-bloom] 1819 Saerelae, M., Esteve Rothenberg, C., Zahemszky, A., 1820 Nikander, P., and J. Ott, "BloomCasting: Security in Bloom 1821 filter based multicast", , Lecture Notes in Computer 1822 Science 2012, , pages 1-16, Springer Berlin Heidelberg, 1823 2012. 1825 [schuetz-intermittent] 1826 Schuetz, S., Eggert, L., Schmid, S., and M. Brunner, 1827 "Protocol enhancements for intermittently connected 1828 hosts", SIGCOMM Comput. Commun. Rev., 35(3):5-18, , July 1829 2005. 1831 [shields-hip] 1832 Shields, C. and J. Garcia-Luna-Aceves, "The HIP protocol 1833 for hierarchical multicast routing", Proceedings of the 1834 seventeenth annual ACM symposium on Principles of 1835 distributed computing, pages 257-266. ACM, New York, NY, 1836 USA, ISBN: 0-89791-977-7, DOI: 10.1145/277697.277744, 1837 1998. 1839 [tritilanunt-dos] 1840 Tritilanunt, S., Boyd, C., Foo, E., and J. Nieto, 1841 "Examining the DoS Resistance of HIP", OTM Workshops (1), 1842 volume 4277 of Lecture Notes in Computer Science, pages 1843 616-625,Springer , 2006. 1845 [urien-rfid] 1846 Urien, P., Chabanne, H., Bouet, M., de Cunha, D., Guyot, 1847 V., Pujolle, G., Paradinas, P., Gressier, E., and J. 1848 Susini, "HIP-based RFID Networking Architecture", IFIP 1849 International Conference on Wireless and Optical 1850 Communications Networks, DOI: 10.1109/WOCN.2007.4284140, 1851 July 2007. 1853 [urien-rfid-draft] 1854 Urien, P., Lee, G., and G. Pujolle, "HIP support for 1855 RFIDs", IRTF Working draft draft-irtf-hiprg-rfid-07, April 1856 2013. 1858 [varjonen-split] 1859 Varjonen, S., Komu, M., and A. Gurtov, "Secure and 1860 Efficient IPv4/IPv6 Handovers Using Host-Based Identifier- 1861 Location Split", Journal of Communications Software and 1862 Systems, 6(1), 2010, ISSN: 18456421, 2010. 1864 [xin-hip-lib] 1865 Xin, G., "Host Identity Protocol Version 2.5", Master's 1866 Thesis, Aalto University, Espoo, Finland, , June 2012. 1868 [xueyong-hip] 1869 Xueyong, Z., Zhiguo, D., and W. Xinling, "A Multicast 1870 Routing Algorithm Applied to HIP-Multicast Model", 1871 Proceedings of the 2011 International Conference on 1872 Network Computing and Information Security - Volume 01 1873 (NCIS '11), Vol. 1. IEEE Computer Society, Washington, DC, 1874 USA, pages 169-174, DOI: 10.1109/NCIS.2011.42, 2011. 1876 [xueyong-secure] 1877 Xueyong, Z. and J. Atwood, "A Secure Multicast Model for 1878 Peer-to-Peer and Access Networks Using the Host Identity 1879 Protocol", Consumer Communications and Networking 1880 Conference. CCNC 2007. 4th IEEE, pages 1098,1102, DOI: 1881 10.1109/CCNC.2007.221, January 2007. 1883 [ylitalo-diss] 1884 Ylitalo, J., "Secure Mobility at Multiple Granularity 1885 Levels over Heterogeneous Datacom Networks", Dissertation, 1886 Helsinki University of Technology, Espoo, Finland ISBN 1887 978-951-22-9531-9, 2008. 1889 [ylitalo-spinat] 1890 Ylitalo, J., Salmela, P., and H. Tschofenig, "SPINAT: 1891 Integrating IPsec into overlay routing", Proceedings of 1892 the First International Conference on Security and Privacy 1893 for Emerging Areas in Communication Networks (SecureComm 1894 2005). Athens, Greece. IEEE Computer Society, pages 1895 315-326, ISBN: 0-7695-2369-2, September 2005. 1897 [zhang-revocation] 1898 Zhang, D., Kuptsov, D., and S. Shen, "Host Identifier 1899 Revocation in HIP", IRTF Working draft draft-irtf-hiprg- 1900 revocation-05, Mar 2012. 1902 Authors' Addresses 1904 Robert Moskowitz (editor) 1905 HTT Consulting 1906 Oak Park 1907 Michigan 1908 USA 1910 Email: rgm@labs.htt-consult.com 1912 Miika Komu 1913 Ericsson 1914 Hirsalantie 11 1915 02420 Jorvas 1916 Finland 1918 Email: miika.komu@ericsson.com