idnits 2.17.1 draft-ietf-hip-rfc4423-bis-06.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 == Line 1663 has weird spacing: '...ncement to th...' == 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 (November 7, 2013) is 3824 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Outdated reference: A later version (-12) exists of draft-ietf-hip-multihoming-03 == Outdated reference: A later version (-33) exists of draft-ietf-hip-native-nat-traversal-05 == Outdated reference: A later version (-20) exists of draft-ietf-hip-rfc5201-bis-14 == Outdated reference: A later version (-07) exists of draft-ietf-hip-rfc5202-bis-04 == Outdated reference: A later version (-11) exists of draft-ietf-hip-rfc5203-bis-02 == Outdated reference: A later version (-08) exists of draft-ietf-hip-rfc5204-bis-02 == Outdated reference: A later version (-10) exists of draft-ietf-hip-rfc5205-bis-02 == Outdated reference: A later version (-14) exists of draft-ietf-hip-rfc5206-bis-06 == Outdated reference: A later version (-09) exists of draft-ietf-hip-rfc6253-bis-01 -- Obsolete informational reference (is this intentional?): RFC 2535 (Obsoleted by RFC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 (~~), 12 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 Verizon 4 Obsoletes: 4423 (if approved) M. Komu 5 Intended status: Informational Aalto 6 Expires: May 11, 2014 November 7, 2013 8 Host Identity Protocol Architecture 9 draft-ietf-hip-rfc4423-bis-06 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 May 11, 2014. 42 Copyright Notice 44 Copyright (c) 2013 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 . . . . . . . . . . . . . . . . . . . . . . 4 72 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . 5 73 2.1. Terms common to other documents . . . . . . . . . . . . . 5 74 2.2. Terms specific to this and other HIP documents . . . . . 5 75 3. Background . . . . . . . . . . . . . . . . . . . . . . . 7 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) . . . . . . . . . . . . . . . . 11 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 . . . . . . . . . . . . . . . . . . . . . . 20 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 . . . . . . . . . . . . . . . . . . . 25 101 11.3.2. HIP in 802.15.4 networks . . . . . . . . . . . . . . . . 26 102 11.4. Answers to NSRG questions . . . . . . . . . . . . . . . . 26 103 12. Security considerations . . . . . . . . . . . . . . . . . 28 104 12.1. MiTM Attacks . . . . . . . . . . . . . . . . . . . . . . 28 105 12.2. Protection against flooding attacks . . . . . . . . . . . 29 106 12.3. HITs used in ACLs . . . . . . . . . . . . . . . . . . . . 30 107 12.4. Alternative HI considerations . . . . . . . . . . . . . . 31 108 13. IANA considerations . . . . . . . . . . . . . . . . . . . 32 109 14. Acknowledgments . . . . . . . . . . . . . . . . . . . . . 32 110 15. Changes from RFC 4423 . . . . . . . . . . . . . . . . . . 33 111 16. References . . . . . . . . . . . . . . . . . . . . . . . 33 112 16.1. Normative References . . . . . . . . . . . . . . . . . . 33 113 16.2. Informative references . . . . . . . . . . . . . . . . . 34 114 Authors' Addresses . . . . . . . . . . . . . . . . . . . 40 116 1. Introduction 118 The Internet has two important global namespaces: Internet Protocol 119 (IP) addresses and Domain Name Service (DNS) names. These two 120 namespaces have a set of features and abstractions that have powered 121 the Internet to what it is today. They also have a number of 122 weaknesses. Basically, since they are all we have, we try and do too 123 much with them. Semantic overloading and functionality extensions 124 have greatly complicated these namespaces. 126 The proposed Host Identity namespace fills an important gap between 127 the IP and DNS namespaces. A Host Identity conceptually refers to a 128 computing platform, and there may be multiple such Host Identities 129 per computing platform (because the platform may wish to present a 130 different identity to different communicating peers). The Host 131 Identity namespace consists of Host Identifiers (HI). There is 132 exactly one Host Identifier for each Host Identity. While this text 133 later talks about non-cryptographic Host Identifiers, the 134 architecture focuses on the case in which Host Identifiers are 135 cryptographic in nature. Specifically, the Host Identifier is the 136 public key of an asymmetric key-pair. Each Host Identity uniquely 137 identifies a single host, i.e., no two hosts have the same Host 138 Identity. If two or more computing platforms have the same Host 139 Identifier, then they are instantiating a distributed host. The Host 140 Identifier can either be public (e.g. published in the DNS), or 141 unpublished. Client systems will tend to have both public and 142 unpublished Host Identifiers. 144 There is a subtle but important difference between Host Identities 145 and Host Identifiers. An Identity refers to the abstract entity that 146 is identified. An Identifier, on the other hand, refers to the 147 concrete bit pattern that is used in the identification process. 149 Although the Host Identifiers could be used in many authentication 150 systems, such as IKEv2 [RFC4306], the presented architecture 151 introduces a new protocol, called the Host Identity Protocol (HIP), 152 and a cryptographic exchange, called the HIP base exchange; see also 153 Section 7. HIP provides for limited forms of trust between systems, 154 enhance mobility, multi-homing and dynamic IP renumbering, aid in 155 protocol translation / transition, and reduce certain types of 156 denial-of-service (DoS) attacks. 158 When HIP is used, the actual payload traffic between two HIP hosts is 159 typically, but not necessarily, protected with ESP. The Host 160 Identities are used to create the needed ESP Security Associations 161 (SAs) and to authenticate the hosts. When ESP is used, the actual 162 payload IP packets do not differ in any way from standard ESP 163 protected IP packets. 165 Much has been learned about HIP [RFC6538] since [RFC4423] was 166 published. This document expands Host Identities beyond use to 167 enable IP connectivity and security to general interhost secure 168 signalling at any protocol layer. The signal may establish a 169 security association between the hosts, or simply pass information 170 within the channel. 172 2. Terminology 174 2.1. Terms common to other documents 176 +---------------+---------------------------------------------------+ 177 | Term | Explanation | 178 +---------------+---------------------------------------------------+ 179 | Public key | The public key of an asymmetric cryptographic key | 180 | | pair. Used as a publicly known identifier for | 181 | | cryptographic identity authentication. Public is | 182 | | a relative term here, ranging from known to peers | 183 | | only to known to the World. | 184 | | | 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 | | | 191 | Public key | An asymmetric cryptographic key pair consisting | 192 | pair | of public and private keys. For example, | 193 | | Rivest-Shamir-Adelman (RSA), Digital Signature | 194 | | Algorithm (DSA) and Elliptic Curve DSA (ECDSA) | 195 | | key pairs are such key pairs. | 196 | | | 197 | End-point | A communicating entity. For historical reasons, | 198 | | the term 'computing platform' is used in this | 199 | | document as a (rough) synonym for end-point. | 200 +---------------+---------------------------------------------------+ 202 2.2. Terms specific to this and other HIP documents 204 It should be noted that many of the terms defined herein are 205 tautologous, self-referential or defined through circular reference 206 to other terms. This is due to the succinct nature of the 207 definitions. See the text elsewhere in this document and in RFC 5201 208 [I-D.ietf-hip-rfc5201-bis] for more elaborate explanations. 210 +---------------+---------------------------------------------------+ 211 | Term | Explanation | 212 +---------------+---------------------------------------------------+ 213 | Computing | An entity capable of communicating and computing, | 214 | platform | for example, a computer. See the definition of | 215 | | 'End-point', above. | 216 | | | 217 | HIP base | A cryptographic protocol; see also Section 7. | 218 | exchange | | 219 | | | 220 | HIP packet | An IP packet that carries a 'Host Identity | 221 | | Protocol' message. | 222 | | | 223 | Host Identity | An abstract concept assigned to a 'computing | 224 | | platform'. See 'Host Identifier', below. | 225 | | | 226 | Host Identity | A name space formed by all possible Host | 227 | namespace | Identifiers. | 228 | | | 229 | Host Identity | A protocol used to carry and authenticate Host | 230 | Protocol | Identifiers and other information. | 231 | | | 232 | Host Identity | The cryptographic hash used in creating the Host | 233 | Hash | Identity Tag from the Host Identity. | 234 | | | 235 | Host Identity | A 128-bit datum created by taking a cryptographic | 236 | Tag | hash over a Host Identifier plus bits to identify | 237 | | which hash used. | 238 | | | 239 | Host | A public key used as a name for a Host Identity. | 240 | Identifier | | 241 | | | 242 | Local Scope | A 32-bit datum denoting a Host Identity. | 243 | Identifier | | 244 | | | 245 | Public Host | A published or publicly known Host Identifier | 246 | Identifier | used as a public name for a Host Identity, and | 247 | and Identity | the corresponding Identity. | 248 | | | 249 | Unpublished | A Host Identifier that is not placed in any | 250 | Host | public directory, and the corresponding Host | 251 | Identifier | Identity. Unpublished Host Identities are | 252 | and Identity | typically short lived in nature, being often | 253 | | replaced and possibly used just once. | 254 | | | 255 | Rendezvous | A mechanism used to locate mobile hosts based on | 256 | Mechanism | their HIT. | 257 +---------------+---------------------------------------------------+ 259 3. Background 261 The Internet is built from three principal components: computing 262 platforms (end-points), packet transport (i.e., internetworking) 263 infrastructure, and services (applications). The Internet exists to 264 service two principal components: people and robotic services 265 (silicon-based people, if you will). All these components need to be 266 named in order to interact in a scalable manner. Here we concentrate 267 on naming computing platforms and packet transport elements. 269 There are two principal namespaces in use in the Internet for these 270 components: IP addresses, and Domain Names. Domain Names provide 271 hierarchically assigned names for some computing platforms and some 272 services. Each hierarchy is delegated from the level above; there is 273 no anonymity in Domain Names. Email, HTTP, and SIP addresses all 274 reference Domain Names. 276 The IP addressing namespace has been overloaded to name both 277 interfaces (at layer-3) and endpoints (for the endpoint-specific part 278 of layer-3, and for layer-4). In their role as interface names, IP 279 addresses are sometimes called "locators" and serve as an endpoint 280 within a routing topology. 282 IP addresses are numbers that name networking interfaces, and 283 typically only when the interface is connected to the network. 284 Originally, IP addresses had long-term significance. Today, the vast 285 number of interfaces use ephemeral and/or non-unique IP addresses. 286 That is, every time an interface is connected to the network, it is 287 assigned an IP address. 289 In the current Internet, the transport layers are coupled to the IP 290 addresses. Neither can evolve separately from the other. IPng 291 deliberations were strongly shaped by the decision that a 292 corresponding TCPng would not be created. 294 There are three critical deficiencies with the current namespaces. 295 Firstly, dynamic readdressing cannot be directly managed. Secondly, 296 confidentiality is not provided in a consistent, trustable manner. 297 Finally, authentication for systems and datagrams is not provided. 298 All of these deficiencies arise because computing platforms are not 299 well named with the current namespaces. 301 3.1. A desire for a namespace for computing platforms 303 An independent namespace for computing platforms could be used in 304 end-to-end operations independent of the evolution of the 305 internetworking layer and across the many internetworking layers. 306 This could support rapid readdressing of the internetworking layer 307 because of mobility, rehoming, or renumbering. 309 If the namespace for computing platforms is based on public-key 310 cryptography, it can also provide authentication services. If this 311 namespace is locally created without requiring registration, it can 312 provide anonymity. 314 Such a namespace (for computing platforms) and the names in it should 315 have the following characteristics: 317 o The namespace should be applied to the IP 'kernel' or stack. The 318 IP stack is the 'component' between applications and the packet 319 transport infrastructure. 321 o The namespace should fully decouple the internetworking layer from 322 the higher layers. The names should replace all occurrences of IP 323 addresses within applications (like in the Transport Control 324 Block, TCB). This replacement can be handled transparently for 325 legacy applications as the LSIs and HITs are compatible with IPv4 326 and IPv6 addresses [RFC5338]. However, HIP-aware applications 327 require some modifications from the developers, who may employ 328 networking API extensions for HIP [RFC6317]. 330 o The introduction of the namespace should not mandate any 331 administrative infrastructure. Deployment must come from the 332 bottom up, in a pairwise deployment. 334 o The names should have a fixed length representation, for easy 335 inclusion in datagram headers and existing programming interfaces 336 (e.g the TCB). 338 o Using the namespace should be affordable when used in protocols. 339 This is primarily a packet size issue. There is also a 340 computational concern in affordability. 342 o Name collisions should be avoided as much as possible. The 343 mathematics of the birthday paradox can be used to estimate the 344 chance of a collision in a given population and hash space. In 345 general, for a random hash space of size n bits, we would expect 346 to obtain a collision after approximately 1.2*sqrt(2**n) hashes 347 were obtained. For 64 bits, this number is roughly 4 billion. A 348 hash size of 64 bits may be too small to avoid collisions in a 349 large population; for example, there is a 1% chance of collision 350 in a population of 640M. For 100 bits (or more), we would not 351 expect a collision until approximately 2**50 (1 quadrillion) 352 hashes were generated. 354 o The names should have a localized abstraction so that it can be 355 used in existing protocols and APIs. 357 o It must be possible to create names locally. When such names are 358 not published, this can provide anonymity at the cost of making 359 resolvability very difficult. 361 o The namespace should provide authentication services. 363 o The names should be long lived, but replaceable at any time. This 364 impacts access control lists; short lifetimes will tend to result 365 in tedious list maintenance or require a namespace infrastructure 366 for central control of access lists. 368 In this document, a new namespace approaching these ideas is called 369 the Host Identity namespace. Using Host Identities requires its own 370 protocol layer, the Host Identity Protocol, between the 371 internetworking and transport layers. The names are based on public- 372 key cryptography to supply authentication services. Properly 373 designed, it can deliver all of the above stated requirements. 375 4. Host Identity namespace 377 A name in the Host Identity namespace, a Host Identifier (HI), 378 represents a statistically globally unique name for naming any system 379 with an IP stack. This identity is normally associated with, but not 380 limited to, an IP stack. A system can have multiple identities, some 381 'well known', some unpublished or 'anonymous'. A system may self- 382 assert its own identity, or may use a third-party authenticator like 383 DNSSEC [RFC2535], PGP, or X.509 to 'notarize' the identity assertion 384 to another namespace. It is expected that the Host Identifiers will 385 initially be authenticated with DNSSEC and that all implementations 386 will support DNSSEC as a minimal baseline. 388 In theory, any name that can claim to be 'statistically globally 389 unique' may serve as a Host Identifier. In the HIP architecture, the 390 public key of a private-public key pair has been chosen as the Host 391 Identifier because it can be self managed and it is computationally 392 difficult to forge. As specified in the Host Identity Protocol 393 [I-D.ietf-hip-rfc5201-bis] specification, a public-key-based HI can 394 authenticate the HIP packets and protect them for man-in-the-middle 395 attacks. Since authenticated datagrams are mandatory to provide much 396 of HIP's denial-of-service protection, the Diffie-Hellman exchange in 397 HIP BEX has to be authenticated. Thus, only public-key HI and 398 authenticated HIP messages are supported in practice. 400 In this document, the non-cryptographic forms of HI and HIP are 401 presented to complete the theory of HI, but they should not be 402 implemented as they could produce worse denial-of-service attacks 403 than the Internet has without Host Identity. There has been past 404 research in challenge puzzles to use non-cryptographic HI, for Radio 405 Frequency IDentification (RFID), in an HIP exchange tailored to the 406 workings of such challenges (as described further in [urien-rfid] and 407 [urien-rfid-draft]). 409 4.1. Host Identifiers 411 Host Identity adds two main features to Internet protocols. The 412 first is a decoupling of the internetworking and transport layers; 413 see Section 5. This decoupling will allow for independent evolution 414 of the two layers. Additionally, it can provide end-to-end services 415 over multiple internetworking realms. The second feature is host 416 authentication. Because the Host Identifier is a public key, this 417 key can be used for authentication in security protocols like ESP. 419 The only completely defined structure of the Host Identity is that of 420 a public/private key pair. In this case, the Host Identity is 421 referred to by its public component, the public key. Thus, the name 422 representing a Host Identity in the Host Identity namespace, i.e., 423 the Host Identifier, is the public key. In a way, the possession of 424 the private key defines the Identity itself. If the private key is 425 possessed by more than one node, the Identity can be considered to be 426 a distributed one. 428 Architecturally, any other Internet naming convention might form a 429 usable base for Host Identifiers. However, non-cryptographic names 430 should only be used in situations of high trust - low risk. That is 431 any place where host authentication is not needed (no risk of host 432 spoofing) and no use of ESP. However, at least for interconnected 433 networks spanning several operational domains, the set of 434 environments where the risk of host spoofing allowed by non- 435 cryptographic Host Identifiers is acceptable is the null set. Hence, 436 the current HIP documents do not specify how to use any other types 437 of Host Identifiers but public keys. For instance, Back-to-My-Mac 438 [RFC6281] from Apple comes pretty close to the functionality of HIP, 439 but unlike HIP, it is based on non-cryptographic identifiers. 441 The actual Host Identifiers are never directly used at the transport 442 or network layers. The corresponding Host Identifiers (public keys) 443 may be stored in various DNS or other directories as identified 444 elsewhere in this document, and they are passed in the HIP base 445 exchange. A Host Identity Tag (HIT) is used in other protocols to 446 represent the Host Identity. Another representation of the Host 447 Identities, the Local Scope Identifier (LSI), can also be used in 448 protocols and APIs. 450 4.2. Host Identity Hash (HIH) 452 The Host Identity Hash is the cryptographic hash algorithm used in 453 producing the HIT from the HI. It is also the hash used through out 454 the HIP protocol for consistency and simplicity. It is possible to 455 for the two hosts in the HIP exchange to use different hash 456 algorithms. 458 Multiple HIHs within HIP are needed to address the moving target of 459 creation and eventual compromise of cryptographic hashes. This 460 significantly complicates HIP and offers an attacker an additional 461 downgrade attack that is mitigated in the HIP protocol 462 [I-D.ietf-hip-rfc5201-bis]. 464 4.3. Host Identity Tag (HIT) 466 A Host Identity Tag is a 128-bit representation for a Host Identity. 467 It is created from an HIH and other information, like an IPv6 prefix 468 and a hash identifier. There are two advantages of using the HIT 469 over using the Host Identifier in protocols. Firstly, its fixed 470 length makes for easier protocol coding and also better manages the 471 packet size cost of this technology. Secondly, it presents the 472 identity in a consistent format to the protocol independent of the 473 cryptographic algorithms used. 475 In essence, the HIT is a hash over the public key. As such, two 476 algorithms affect the generation of a HIT: the public-key algorithm 477 of the HI and the used HIH. The two algorithms are encoded in the 478 bit presentation of the HIT. As the two communicating parties may 479 support different algorithms, [I-D.ietf-hip-rfc5201-bis] defines the 480 minimum set for interoperability. For further interoperability, the 481 responder may store its keys in DNS records, and thus the initiator 482 may have to couple destination HITs with appropriate source HIts 483 according to matching HIH. 485 In the HIP packets, the HITs identify the sender and recipient of a 486 packet. Consequently, a HIT should be unique in the whole IP 487 universe as long as it is being used. In the extremely rare case of 488 a single HIT mapping to more than one Host Identity, the Host 489 Identifiers (public keys) will make the final difference. If there 490 is more than one public key for a given node, the HIT acts as a hint 491 for the correct public key to use. 493 4.4. Local Scope Identifier (LSI) 495 An LSI is a 32-bit localized representation for a Host Identity. The 496 purpose of an LSI is to facilitate using Host Identities in existing 497 APIs for IPv4-based applications. Besides facilitating HIP-based 498 connectivity for legacy IPv4 applications, the LSIs are beneficial in 499 two other scenarios [RFC6538]. 501 In the first scenario, two IPv4-only applications are residing on two 502 separate hosts connected by IPv6-only network. With HIP-based 503 connectivity, the two applications are able to communicate despite of 504 the mismatch in the protocol families of the applications and the 505 underlying network. The reason is that the HIP layer translates the 506 LSIs originating from the upper layers into routable IPv6 locators 507 before delivering the packets on the wire. 509 The second scenario is the same as the first one, but with the 510 difference that one of the applications supports only IPv6. Now two 511 obstacles hinder the communication between the application: the 512 addressing families of the two applications differ, and the 513 application residing at the IPv4-only side is again unable to 514 communicate because of the mismatch between addressing families of 515 the application (IPv4) and network (IPv6). With HIP-based 516 connectivity for applications, this scenario works; the HIP layer can 517 choose whether to translate the locator of an incoming packet into an 518 LSI or HIT. 520 Effectively, LSIs improve IPv6 interoperability at the network layer 521 as described in the first scenario and at the application layer as 522 depicted in the second example. The interoperability mechanism 523 should not be used to avoid transition to IPv6; the authors firmly 524 believe in IPv6 adoption and encourage developers to port existing 525 IPv4-only applications to use IPv6. However, some proprietary, 526 closed-source, IPv4-only applications may never see the daylight of 527 IPv6, and the LSI mechanism is suitable for extending the lifetime of 528 such applications even in IPv6-only networks. 530 The main disadvantage of an LSI is its local scope. Applications may 531 violate layering principles and pass LSIs to each other in 532 application-layer protocols. As the LSIs are valid only in the 533 context of the local host, they may represent an entirely different 534 host when passed to another host. However, it should be emphasized 535 here that the LSI concept is effectively a host-based NAT and does 536 not introduce any more issues than the prevalent middlebox based NATs 537 for IPv4. In other words, the applications violating layering 538 principles are already broken by the NAT boxes that are ubiquitously 539 deployed. 541 4.5. Storing Host Identifiers in directories 543 The public Host Identifiers should be stored in DNS; the unpublished 544 Host Identifiers should not be stored anywhere (besides the 545 communicating hosts themselves). The (public) HI along with the 546 supported HIHs are stored in a new RR type. This RR type is defined 547 in HIP DNS Extension [I-D.ietf-hip-rfc5205-bis]. 549 Alternatively, or in addition to storing Host Identifiers in the DNS, 550 they may be stored in various other directories. For instance, 551 Light-weight Directory Access Protocol (LDAP) or in a Public Key 552 Infrastructure (PKI) [I-D.ietf-hip-rfc6253-bis]. Alternatively, 553 Distributed Hash Tables (DHTs) [RFC6537] have successfully been 554 utilized [RFC6538]. Such a practice may allow them to be used for 555 purposes other than pure host identification. 557 Some types of application may cache and use Host Identifiers 558 directly, while others may indirectly discover them through symbolic 559 host name (such as FQDN) look up from a directory. Even though Host 560 Identities can have a substantially longer lifetime associate with 561 them than routable IP addresses, directories may be a better approach 562 to manage the lifespan of Host Identities. For example, a LDAP or 563 DHT can be used for locally published identities whereas DNS can be 564 more suitable for public advertisement. 566 5. New stack architecture 568 One way to characterize Host Identity is to compare the proposed new 569 architecture with the current one. As discussed above, the IP 570 addresses can be seen to be a confounding of routing direction 571 vectors and interface names. Using the terminology from the IRTF 572 Name Space Research Group Report [nsrg-report] and, e.g., the 573 unpublished Internet-Draft Endpoints and Endpoint Names 574 [chiappa-endpoints], the IP addresses currently embody the dual role 575 of locators and end-point identifiers. That is, each IP address 576 names a topological location in the Internet, thereby acting as a 577 routing direction vector, or locator. At the same time, the IP 578 address names the physical network interface currently located at the 579 point-of-attachment, thereby acting as a end-point name. 581 In the HIP architecture, the end-point names and locators are 582 separated from each other. IP addresses continue to act as locators. 583 The Host Identifiers take the role of end-point identifiers. It is 584 important to understand that the end-point names based on Host 585 Identities are slightly different from interface names; a Host 586 Identity can be simultaneously reachable through several interfaces. 588 The difference between the bindings of the logical entities are 589 illustrated in Figure 1. Left side illustrates the current TCP/IP 590 architecture and right side the HIP-based architecture. 592 Transport ---- Socket Transport ------ Socket 593 association | association | 594 | | 595 | | 596 | | 597 End-point | End-point --- Host Identity 598 \ | | 599 \ | | 600 \ | | 601 \ | | 602 Location --- IP address Location --- IP address 604 Figure 1 606 Architecturally, HIP provides for a different binding of transport- 607 layer protocols. That is, the transport-layer associations, i.e., 608 TCP connections and UDP associations, are no longer bound to IP 609 addresses but rather to Host Identities. In practice, the Host 610 Identities are exposed as LSIs and HITs for legacy applications and 611 the transport layer to facilitate backward compatibility with 612 existing networking APIs and stacks. 614 5.1. On the multiplicity of identities 616 For security reasons, it may be a bad idea to duplicate the same Host 617 Identity on multiple hosts because the compromise of a single host 618 taints the identities of the other hosts. Management of machines 619 with identical Host Identities may also present other challenges and, 620 therefore, it is advisable to have a unique identity for each host. 622 Instead of duplicating identities, HIP opportunistic mode can be 623 employed, where the initiator leaves out the identifier of the 624 responder when initiating the key exchange and learns it upon the 625 completion of the exchange. The tradeoffs are related to lowered 626 security guarantees, but a benefit of the approach is to avoid 627 publishing of Host Identifiers in any directories [komu-leap]. The 628 approach could also be used for load balancing purposes at the HIP 629 layer because the identity of the responder can be decided 630 dynamically during the key exchange. Thus, the approach has the 631 potential to be used as a HIP-layer "anycast", either directly 632 between two hosts or indirectly through the rendezvous service 633 [komu-diss]. 635 At the client side, a host may have multiple Host Identities, for 636 instance, for privacy purposes. Another reason can be that the 637 person utilizing the host employs different identities for different 638 administrative domains as an extra security measure. If a HIP-aware 639 middlebox, such as a HIP-based firewall, is on the path between the 640 client and server, the user or the underlying system should carefully 641 choose the correct identity to avoid the firewall to unnecessarily 642 drop HIP-base connectivity [komu-diss]. 644 Similarly, a server may have multiple Host Identities. For instance, 645 a single web server may serve multiple different administrative 646 domains. Typically, the distinction is accomplished based on the DNS 647 name, but also the Host Identity could be used for this purpose. 648 However, a more compelling reason to employ multiple identities are 649 HIP-aware firewalls that are unable see the HTTP traffic inside the 650 encrypted IPsec tunnel. In such a case, each service could be 651 configured with a separate identity, thus allowing the firewall to 652 segregate the different services of the single web server from each 653 other [lindqvist-enterprise]. 655 6. Control plane 657 HIP decouples control and data plane from each other. The control 658 plane between two end-hosts is initialized using a key exchange 659 procedure called the base exchange. The procedure can be assisted by 660 new infrastructural intermediaries called rendezvous or relay 661 servers. In the event of IP address changes, the end-hosts sustain 662 control plane connectivity with mobility and multihoming extensions. 663 Eventually, the end-hosts terminate the control plane and remove the 664 associated state. 666 6.1. Base exchange 668 The base exchange is key exchange procedure that authenticates the 669 initiator and responder to each other using their public keys. 670 Typically, the initiator is the client-side host and the responder is 671 the server-side host. The roles are used by the state machine of a 672 HIP implementation, but discarded upon successful completion. 674 The exchange consists of four messages during which the hosts also 675 create symmetric keys to protect the control plane with Hash-based 676 message authentication codes (HMACs). The keys can be also used to 677 protect the data plane, and IPsec ESP [I-D.ietf-hip-rfc5202-bis] is 678 typically used as the data-plane protocol, albeit HIP can also 679 accommodate others. Both the control and data plane are terminated 680 using a closing procedure consisting of two messages. 682 The base exchange also includes a computational puzzle 683 [I-D.ietf-hip-rfc5201-bis] that the initiator must solve. The 684 responder chooses the difficulty of the puzzle which allows the 685 responder to delay new incoming initiators according to local 686 policies, for instance, when the responder is under heavy load. The 687 puzzle can offer some resiliency against DoS attacks because the 688 design of the puzzle mechanism allows the responder to remain 689 stateless until the very end of the base exchange [aura-dos]. HIP 690 puzzles have also been researched under steady-state DDoS attacks 691 [beal-dos], multiple adversary models with varying puzzle 692 difficulties [tritilanunt-dos] and ephemeral Host Identities 693 [komu-mitigation]. 695 6.2. End-host mobility and multi-homing 697 HIP decouples the transport from the internetworking layer, and binds 698 the transport associations to the Host Identities (through actually 699 either the HIT or LSI). After the initial key exchange, the HIP 700 layer maintains transport-layer connectivity and data flows using its 701 mobility [I-D.ietf-hip-rfc5206-bis] and multihoming 702 [I-D.ietf-hip-multihoming] extensions. Consequently, HIP can provide 703 for a degree of internetworking mobility and multi-homing at a low 704 infrastructure cost. HIP mobility includes IP address changes (via 705 any method) to either party. Thus, a system is considered mobile if 706 its IP address can change dynamically for any reason like PPP, DHCP, 707 IPv6 prefix reassignments, or a NAT device remapping its translation. 708 Likewise, a system is considered multi-homed if it has more than one 709 globally routable IP address at the same time. HIP links IP 710 addresses together, when multiple IP addresses correspond to the same 711 Host Identity, and if one address becomes unusable, or a more 712 preferred address becomes available, existing transport associations 713 can easily be moved to another address. 715 When a node moves while communication is already on-going, address 716 changes are rather straightforward. The peer of the mobile node can 717 just accept a HIP or an integrity protected ESP packet from any 718 address and ignore the source address. However, as discussed in 719 Section 12.2 below, a mobile node must send a HIP UPDATE packet to 720 inform the peer of the new address(es), and the peer must verify that 721 the mobile node is reachable through these addresses. This is 722 especially helpful for those situations where the peer node is 723 sending data periodically to the mobile node (that is re-starting a 724 connection after the initial connection). 726 6.3. Rendezvous mechanism 728 Making a contact to a mobile node is slightly more involved. In 729 order to start the HIP exchange, the initiator node has to know how 730 to reach the mobile node. Although infrequently moving HIP nodes 731 could use Dynamic DNS [RFC2136] to update their reachability 732 information in the DNS, an alternative to using DNS in this fashion 733 is to use a piece of new static infrastructure to facilitate 734 rendezvous between HIP nodes. 736 The mobile node keeps the rendezvous infrastructure continuously 737 updated with its current IP address(es). The mobile nodes must trust 738 the rendezvous mechanism to properly maintain their HIT and IP 739 address mappings. 741 The rendezvous mechanism is also needed if both of the nodes happen 742 to change their address at the same time, either because they are 743 mobile and happen to move at the same time, because one of them is 744 off-line for a while, or because of some other reason. In such a 745 case, the HIP UPDATE packets will cross each other in the network and 746 never reach the peer node. 748 The HIP rendezvous mechanism is defined in HIP Rendezvous 749 specifications [I-D.ietf-hip-rfc5204-bis]. 751 6.4. Relay mechanism 753 The HIP relay mechanism [I-D.ietf-hip-native-nat-traversal] is an 754 alternative to the HIP rendezvous mechanism. The HIP relay mechanism 755 is more suitable for IPv4 networks with NATs because a HIP relay can 756 forward all control and data plane communications in order to 757 guarantee successful NAT traversal. 759 6.5. Termination of the control plane 761 The control plane between two hosts can be terminated using a secure 762 two message procotol as specified in (XX FIXME). The related state 763 (i.e. host associations) should be removed upon successful 764 termination. 766 7. Data plane 768 The control and data plane are decoupled in the HIP architecture. 769 This means that the encapsulation format for data plane used for 770 carrying the application-layer traffic is changeable and can is 771 dynamically negotiated during the key exchange. For instance, 772 HICCUPS extensions [RFC6078] define a way to transport application- 773 layer datagrams directly over the HIP control plane, protected by 774 asymmetric key cryptography. Also, S-RTP has been considered as the 775 data encapsulation protocol [hip-srtp]. However, the most widely 776 implemented method is the Encapsulated Security Payload (ESP) 777 [I-D.ietf-hip-rfc5202-bis] that is protected by symmetric keys 778 derived during the key exchange. ESP Security Associations (SAs) 779 offer both confidentiality and integrity protection, of which the 780 former can be disabled during the key exchange. In the future, other 781 ways of transporting application-layer data may be defined. 783 The ESP SAs are established and terminated between the initiator and 784 the responder hosts. Usually, the hosts create at least two SAs, one 785 in each direction (initiator-to-responder SA and responder-to- 786 initiator SA). If the IP addresses of either host are changed, the 787 HIP mobility extensions can be used to re-negotiate the corresponding 788 SAs. 790 On the wire, the difference in the use of identifiers between the HIP 791 control and data plane is that the HITs are included in all control 792 packets, but not in the data plane when ESP is employed. Instead, 793 the ESP employs SPI numbers that act as compressed HITs. Any HIP- 794 aware middlebox (for instance, a HIP-aware firewall) interested in 795 the ESP based data plane should keep track between the control and 796 data plane identifiers in order to associate them with each other. 798 Since HIP does not negotiate any SA lifetimes, all lifetimes are 799 local policy. The only lifetimes a HIP implementation must support 800 are sequence number rollover (for replay protection), and SA timeout. 801 An SA times out if no packets are received using that SA. 802 Implementations may support lifetimes for the various ESP transforms 803 and other data-plane protocols. 805 8. HIP and NATs 807 Passing packets between different IP addressing realms requires 808 changing IP addresses in the packet header. This may occur, for 809 example, when a packet is passed between the public Internet and a 810 private address space, or between IPv4 and IPv6 networks. The 811 address translation is usually implemented as Network Address 812 Translation (NAT) [RFC3022] or NAT Protocol translation (NAT-PT) 813 [RFC2766]. 815 In a network environment where identification is based on the IP 816 addresses, identifying the communicating nodes is difficult when NATs 817 are employed because the private address spaces introduced by NATs 818 are overlapping. In other words, two hosts cannot distinguished from 819 each other solely based on their IP address. With HIP, the 820 transport-layer end-points (i.e. applications) are bound to unique 821 Host Identities rather than overlapping private addresses. This 822 makes it possible for two end-points to distinguish one other even 823 when they are located in private address realms. Thus, the IP 824 addresses used only for routing purposes; they may be changed freely 825 during when a packet between two hosts traverses possibly multiple 826 addressing realm boundaries. 828 NAT traversal extensions for HIP [I-D.ietf-hip-native-nat-traversal] 829 can be used to realize the actual end-to-end connectivity through NAT 830 devices. To support basic backward compatibility with legacy NATs, 831 the extensions encapsulated both HIP control and data plane in UDP. 832 The extensions define mechanisms for forwarding the two planes 833 through an intermediary host called HIP relay and procedures to 834 establish direct end-to-end connectivity by penetrating NATs. 835 Besides this "native" NAT traversal mode for HIP, other NAT traversal 836 mechanisms have been successfully utilized, such as Teredo 837 [varjonen-split]. 839 Besides legacy NATs, a HIP-aware NAT has been designed and 840 implemented [ylitalo-spinat]. For a HIP-based flow, a HIP-aware NAT 841 or NAT-PT system tracks the mapping of HITs, and the corresponding 842 ESP SPIs, to an IP address. The NAT system has to learn mappings 843 both from HITs and from SPIs to IP addresses. Many HITs (and SPIs) 844 can map to a single IP address on a NAT, simplifying connections on 845 address poor NAT interfaces. The NAT can gain much of its knowledge 846 from the HIP packets themselves; however, some NAT configuration may 847 be necessary. 849 8.1. HIP and Upper-layer checksums 851 There is no way for a host to know if any of the IP addresses in an 852 IP header are the addresses used to calculate the TCP checksum. That 853 is, it is not feasible to calculate the TCP checksum using the actual 854 IP addresses in the pseudo header; the addresses received in the 855 incoming packet are not necessarily the same as they were on the 856 sending host. Furthermore, it is not possible to recompute the 857 upper-layer checksums in the NAT/NAT-PT system, since the traffic is 858 ESP protected. Consequently, the TCP and UDP checksums are 859 calculated using the HITs in the place of the IP addresses in the 860 pseudo header. Furthermore, only the IPv6 pseudo header format is 861 used. This provides for IPv4 / IPv6 protocol translation. 863 9. Multicast 865 A number of studies have intestigating HIP-based multicast have been 866 published (including [shields-hip], [xueyong-hip], [xueyong-hip], 867 [amir-hip], [kovacshazi-host] and [xueyong-secure]). Particularly, 868 so called bloom filters, that allow to compressing of multiple labels 869 into small datastructures, may be a promising way forward 870 [sarela-bloom]. However, the different schemes have not been adopted 871 by HIP working group (nor the HIP research group in IRTF), so the 872 details are not further elaborated here. 874 10. HIP policies 876 There are a number of variables that will influence the HIP exchanges 877 that each host must support. All HIP implementations should support 878 at least 2 HIs, one to publish in DNS or similar directory service 879 and an unpublished one for anonymous usage. Although unpublished HIs 880 will be rarely used as responder HIs, they are likely be common for 881 initiators. Support for multiple HIs is recommended. This provides 882 new challenges for systems or users to decide which type of HI to 883 expose when they start a new session. 885 Opportunistic mode (where the initiator starts a HIP exchange without 886 prior knowledge of the responder's HI) presents a security tradeoff. 887 At the expense of being subject to MITM attacks, the opportunistic 888 mode allows the initiator learn the the identity of the responder 889 during communications rather than from an external directory. 890 Opportunistic mode can be used for registering to HIP-based services 891 [I-D.ietf-hip-rfc5203-bis] (i.e. utilized by HIP for its own internal 892 purposes) or by the application layer [komu-leap]. For security 893 reasons, especially the latter requires some involvement from the 894 user to accept the identity of the responder in a similar vain as SSH 895 prompts the user when connecting to a server for the first time 896 [pham-leap]. In practice, this can be realized for with end-host 897 based firewalls in the case of legacy applications [karvonen-usable] 898 or with native APIs for HIP APIs [RFC6317] in the case of HIP-aware 899 applications. 901 Many initiators would want to use a different HI for different 902 responders. The implementations should provide for a policy of 903 initiator HIT to responder HIT. This policy should also include 904 preferred transforms and local lifetimes. 906 Responders would need a similar policy, describing the hosts allowed 907 to participate in HIP exchanges, and the preferred transforms and 908 local lifetimes. 910 11. Design considerations 912 11.1. Benefits of HIP 914 In the beginning, the network layer protocol (i.e., IP) had the 915 following four "classic" invariants: 917 1. Non-mutable: The address sent is the address received. 919 2. Non-mobile: The address doesn't change during the course of an 920 "association". 922 3. Reversible: A return header can always be formed by reversing the 923 source and destination addresses. 925 4. Omniscient: Each host knows what address a partner host can use 926 to send packets to it. 928 Actually, the fourth can be inferred from 1 and 3, but it is worth 929 mentioning for reasons that will be obvious soon if not already. 931 In the current "post-classic" world, we are intentionally trying to 932 get rid of the second invariant (both for mobility and for multi- 933 homing), and we have been forced to give up the first and the fourth. 934 Realm Specific IP [RFC3102] is an attempt to reinstate the fourth 935 invariant without the first invariant. IPv6 is an attempt to 936 reinstate the first invariant. 938 Few client-side systems on the Internet have DNS names that are 939 meaningful. That is, if they have a Fully Qualified Domain Name 940 (FQDN), that name typically belongs to a NAT device or a dial-up 941 server, and does not really identify the system itself but its 942 current connectivity. FQDNs (and their extensions as email names) 943 are application-layer names; more frequently naming services than a 944 particular system. This is why many systems on the Internet are not 945 registered in the DNS; they do not have services of interest to other 946 Internet hosts. 948 DNS names are references to IP addresses. This only demonstrates the 949 interrelationship of the networking and application layers. DNS, as 950 the Internet's only deployed, distributed database is also the 951 repository of other namespaces, due in part to DNSSEC and application 952 specific key records. Although each namespace can be stretched (IP 953 with v6, DNS with KEY records), neither can adequately provide for 954 host authentication or act as a separation between internetworking 955 and transport layers. 957 The Host Identity (HI) namespace fills an important gap between the 958 IP and DNS namespaces. An interesting thing about the HI is that it 959 actually allows one to give up all but the 3rd network-layer 960 invariant. That is to say, as long as the source and destination 961 addresses in the network-layer protocol are reversible, then things 962 work ok because HIP takes care of host identification, and 963 reversibility allows one to receive a packet back to one's partner 964 host. You do not care if the network-layer address changes in 965 transit (mutable) and you don't care what network-layer address the 966 partner is using (non-omniscient). 968 The Host Identity (HI) namespace fills an important gap between the 969 IP and DNS namespaces. An interesting thing about the HI is that it 970 actually allows one to give up all but the 3rd network-layer 971 invariant. That is to say, as long as the source and destination 972 addresses in the network-layer protocol are reversible, then things 973 work ok because HIP takes care of host identification, and 974 reversibility allows one to receive a packet back to one's partner 975 host. You do not care if the network-layer address changes in 976 transit (mutable) and you don't care what network-layer address the 977 partner is using (non-omniscient). 979 The Sockets API is the de-facto API for utilize the TCP/IP stack. 980 Application use the Sockets API either directly or indirectly through 981 some libraries or frameworks. However, the Sockets API was based on 982 the assumption of static IP addresses and DNS with its lifetime 983 values was invented at later stages during the evolution of the 984 Internet. Hence, the Sockets API does not deal with the lifetime of 985 addresses [RFC6250]. As majority of the end-user equipment is mobile 986 today, their addresses are effectively ephemeral, but the Sockets API 987 still gives a fallacious illusion of persistent IP addresses to the 988 unwary developer. HIP can be used to solidify this illusion because 989 HIP provides persistent surrogate addresses to the application layer 990 in the form of LSIs and HITs. 992 The persistent identifiers as provided by HIP are useful in multiple 993 scenarios (as described in more detail in e.g. [ylitalo-diss] or 994 [komu-diss]): 996 o When a mobile host moves physically between two different WLAN 997 networks and obtains a new address, an application using the 998 identifiers remains isolated of the topology changes while the 999 underlying HIP layer re-establishes connectivity (i.e. a 1000 horizontal handoff). 1002 o Similarly, the application utilizing the identifiers remains again 1003 unaware of the topological changes when the underlying host 1004 equipped with WLAN and cellular network interfaces switches 1005 between the two different access technologies (i.e. a vertical 1006 handoff). 1008 o Even when hosts are located in private address realms, 1009 applications can uniquely distinguish different hosts from each 1010 other based on their identifier. In other words, it can be stated 1011 that HIP improves Internet transparency for the application layer 1012 [komu-diss]. 1014 o Site renumbering events for services can occur due to corporate 1015 mergers or acquisitions, or by changes in Internet Service 1016 Provider. They can involve changing the entire network prefix of 1017 an organization, which is problematic due to hard-coded addresses 1018 in service configuration files or cached IP addresses at the 1019 client side [RFC5887]. Considering such human errors, a site 1020 employing location-independent identifiers as promoted by HIP may 1021 experience less problems while renumbering their network. 1023 o More agile IPv6 interoperability as discussed in section 1024 Section 4.4. IPv6-based applications can communicate using HITs 1025 with IPv4-based applications that are using LSIs. Also, the 1026 underlying network type (IPv4 or IPv6) becomes independent of the 1027 addressing family of the application. 1029 o HITs (or LSIs) can be used in IP-based access control lists as a 1030 more secure replacement for IPv6 addresses. Besides security, HIT 1031 based access control has two other benefits. First, the use of 1032 HITs halves the size of access control lists as separate rules for 1033 IPv4 are not needed [komu-diss]. Second, HIT-based configuration 1034 rules in HIP-aware middleboxes remain static and independent of 1035 topology changes, thus simplifying administrative efforts 1036 particularly for mobile environments. For instance, the benefits 1037 of HIT based access control have been harnessed in the case of 1038 HIP-aware firewalls, but can be utilized directly at the end-hosts 1039 as well [RFC6538]. 1041 While some of these benefits could be and have been redundantly 1042 implemented by individual applications, providing such generic 1043 functionality at the lower layers is useful because it reduces 1044 software development efforts and networking software bugs (as the 1045 layer is tested with multiple applications). It also allows the 1046 developer to focus on building the application itself rather than 1047 delving into the intricacies of mobile networking, thus facilitating 1048 separation of concerns. 1050 HIP could also be realized by combining a number of different 1051 protocols, but the complexity of the resulting software may become 1052 substantially larger, and the interaction multiple possibly layered 1053 protocols may have adverse effects on latency and throughput. It is 1054 also worth noting that virtually nothing prevents realizing the HIP 1055 architecture, for instance, as an application-layer library, which 1056 has been actually implemented in the past [xin-hip-lib]. However, 1057 the tradeoff in moving the HIP layer to the application layer is that 1058 legacy applications may not be supported. 1060 11.2. Drawbacks of HIP 1062 In computer science, many problems can be solved with an extra layer 1063 of indirection. However, the indirection always involves some costs 1064 as there no such thing as "free lunch". In the case of HIP, the main 1065 costs could be stated as follows: 1067 o In general, a new layer and a new namespace involves always some 1068 initial effort in terms implementation, deployment and 1069 maintenance. Some education of people may also be needed. 1070 However, the HIP community at the IETF have spent years in 1071 experimenting, exploring, testing, documenting and implementing 1072 HIP curb the amount of efforts required. 1074 o HIP decouples identifier and locator roles of IP addresses. 1075 Consequently, a mapping mechanism is needed to associate them 1076 together. A failure to map a HIT to its corresponding locator may 1077 result in failed connectivity because a HIT is "flat" by its 1078 nature and cannot be looked up from the hierarchically organized 1079 DNS. HITs are flat by design due to a security tradeoff. The 1080 more bits are allocated for the hash in the HIT, the less likely 1081 there will be (malicious) collisions. 1083 o From performance viewpoint, HIP control and data plane processing 1084 introduces some overhead in terms throughput and latency as 1085 elaborated below. 1087 The key exchange introduces some extra latency (two round trips) in 1088 connection establishment. This can further affect TCP traffic 1089 particularly when a TCP application triggers the key exchange and the 1090 triggering SYN packet is dropped instead of being cached. Similarly 1091 as with the key exchange, a similar performance penalty may incur for 1092 TCP during HIP handoff procedures. The penalty can be constrained 1093 with caching TCP packets. Also, TCP user timeout [RFC5482] is 1094 another way to optimize TCP behavior during handoffs 1095 [scultz-intermittent]. 1097 The most CPU-intensive operations involve the use of the asymmetric 1098 keys and Diffie-Hellman key derivation at the control plane, but this 1099 occurs only during the key exchange, its maintenance (handoffs, 1100 refreshing of key material) and tear down procedures of HIP 1101 associations. The data plane is typically implemented with ESP has a 1102 smaller overhead due to symmetric key encryption. Naturally, even 1103 ESP involves some overhead in terms latency (processing costs) and 1104 throughput (tunneling) (see e.g. [ylitalo-diss] for a performance 1105 evaluation). 1107 11.3. Deployment and adoption considerations 1109 This section describes some deployment and adoption considerations 1110 related to HIP from a technical perspective. 1112 11.3.1. Deployment analysis 1114 HIP has commercially been utilized at Boeing airplane factory for 1115 their internal purposes[paine-hip]. It has been included in a 1116 security product called Tofino to support layer-two Virtual Private 1117 Networks [henderson-vpls] to facilitate, e.g, supervisory control and 1118 data acquisition (SCADA) security. However, HIP has not been a "wild 1119 success" [RFC5218] in the Internet as argued by Levae et al 1120 [leva-barriers]. Here, we briefly highligt some of their findings 1121 based on interviews with 19 experts from the industry and academia. 1123 From a marketing perspective, the demand for HIP has been low and 1124 substitute technologies have been favored. Another identified reason 1125 has been that some technical misconceptions related to the early 1126 stages of HIP specifications still persist. Two identified 1127 misconceptions are that HIP does not support NAT traversal, and HIP 1128 must be implemented in the OS kernel. Both of these claims are 1129 untrue; HIP does have NAT traversal extensions 1130 [I-D.ietf-hip-native-nat-traversal], and kernel modifications can be 1131 avoided with modern operating systems by diverting packets for 1132 userspace processing. 1134 The analysis clarifies infrastructural requirements for HIP. In a 1135 minimal set up, a client and server machine have to run HIP software. 1136 However, to avoid manual configurations, usually DNS records for HIP 1137 are set up. For instance, the popular DNS server software Bind9 does 1138 not require any changes to accomodate DNS records for HIP because 1139 they can be supported in binary format in its configuration files 1140 [RFC6538]. HIP rendezvous servers and firewalls are optional. No 1141 changes are required to network address points, NATs, edge routers or 1142 core networks. HIP may require holes in legacy firewalls. 1144 The analysis also clarifies the requirements for the host components 1145 that consist of three parts. First, a HIP control plane component is 1146 required, typically implemented as as userspace daemon. Second, a 1147 data plane component is needed. Most HIP implementations utilize the 1148 so called BEET mode of ESP that has been available since Linux kernel 1149 2.6.27, but is included also as a userspace component in HIPL and 1150 OpenHIP implementations. Third, HIP systems usually provide a DNS 1151 proxy for the local host that translates HIP DNS records to LSIs and 1152 HITs, and communicates the corresponding locators to HIP userspace 1153 daemon. While the third component is not strictly speaking 1154 mandatory, it is very useful for avoiding manual configurations. The 1155 three components are further described in the HIP experiment report 1156 [RFC6538]. 1158 Based on the interviews, Levae et al suggest further directions to 1159 facilitate HIP deployment. Transitioning the HIP specifications to 1160 the standards track may help, but other measures could be taken. As 1161 a more radical measure, the authors suggest to implement HIP as a 1162 purely application-layer library [xin-hip-lib] or other kind of 1163 middleware. On the other hand, more conservative measures include 1164 focusing on private deployments controlled by a single stakeholder. 1165 As an a more concrete example of such a scenario, HIP could be used 1166 by a single service provider to provide interconnectivity between its 1167 servers [komu-cloud]. 1169 11.3.2. HIP in 802.15.4 networks 1171 The IEEE 802 standards have been defining MAC layered security. Many 1172 of these standards use EAP [RFC3748] as a Key Management System (KMS) 1173 transport, but some like IEEE 802.15.4 [IEEE.802-15-4.2011] leave the 1174 KMS and its transport as "Out of Scope". 1176 HIP is well suited as a KMS in these environments: 1178 o HIP is independent of IP addressing and can be directly 1179 transported over any network protocol. 1181 o Master Keys in 802 protocols are strictly pair-based with group 1182 keys transported from the group controller using pair-wise keys. 1184 o AdHoc 802 networks can be better served by a peer-to-peer KMS than 1185 the EAP client/server model. 1187 o Some devices are very memory constrained and a common KMS for both 1188 MAC and IP security represents a considerable code savings. 1190 11.4. Answers to NSRG questions 1192 The IRTF Name Space Research Group has posed a number of evaluating 1193 questions in their report [nsrg-report]. In this section, we provide 1194 answers to these questions. 1196 1. How would a stack name improve the overall functionality of the 1197 Internet? 1199 HIP decouples the internetworking layer from the transport 1200 layer, allowing each to evolve separately. The decoupling 1201 makes end-host mobility and multi-homing easier, also across 1202 IPv4 and IPv6 networks. HIs make network renumbering easier, 1203 and they also make process migration and clustered servers 1204 easier to implement. Furthermore, being cryptographic in 1205 nature, they provide the basis for solving the security 1206 problems related to end-host mobility and multi-homing. 1208 2. What does a stack name look like? 1210 A HI is a cryptographic public key. However, instead of using 1211 the keys directly, most protocols use a fixed size hash of the 1212 public key. 1214 3. What is its lifetime? 1216 HIP provides both stable and temporary Host Identifiers. 1217 Stable HIs are typically long lived, with a lifetime of years 1218 or more. The lifetime of temporary HIs depends on how long 1219 the upper-layer connections and applications need them, and 1220 can range from a few seconds to years. 1222 4. Where does it live in the stack? 1224 The HIs live between the transport and internetworking layers. 1226 5. How is it used on the end points? 1228 The Host Identifiers may be used directly or indirectly (in 1229 the form of HITs or LSIs) by applications when they access 1230 network services. Additionally, the Host Identifiers, as 1231 public keys, are used in the built in key agreement protocol, 1232 called the HIP base exchange, to authenticate the hosts to 1233 each other. 1235 6. What administrative infrastructure is needed to support it? 1237 In some environments, it is possible to use HIP 1238 opportunistically, without any infrastructure. However, to 1239 gain full benefit from HIP, the HIs must be stored in the DNS 1240 or a PKI, and a new rendezvous mechanism is needed 1241 [I-D.ietf-hip-rfc5205-bis]. 1243 7. If we add an additional layer would it make the address list in 1244 SCTP unnecessary? 1246 Yes 1248 8. What additional security benefits would a new naming scheme 1249 offer? 1251 HIP reduces dependency on IP addresses, making the so called 1252 address ownership [Nik2001] problems easier to solve. In 1253 practice, HIP provides security for end-host mobility and 1254 multi-homing. Furthermore, since HIP Host Identifiers are 1255 public keys, standard public key certificate infrastructures 1256 can be applied on the top of HIP. 1258 9. What would the resolution mechanisms be, or what characteristics 1259 of a resolution mechanisms would be required? 1261 For most purposes, an approach where DNS names are resolved 1262 simultaneously to HIs and IP addresses is sufficient. 1263 However, if it becomes necessary to resolve HIs into IP 1264 addresses or back to DNS names, a flat resolution 1265 infrastructure is needed. Such an infrastructure could be 1266 based on the ideas of Distributed Hash Tables, but would 1267 require significant new development and deployment. 1269 12. Security considerations 1271 This section includes discussion on some issues and solutions related 1272 to security in the HIP architecture. 1274 12.1. MiTM Attacks 1276 HIP takes advantage of the new Host Identity paradigm to provide 1277 secure authentication of hosts and to provide a fast key exchange for 1278 ESP. HIP also attempts to limit the exposure of the host to various 1279 denial-of-service (DoS) and man-in-the-middle (MitM) attacks. In so 1280 doing, HIP itself is subject to its own DoS and MitM attacks that 1281 potentially could be more damaging to a host's ability to conduct 1282 business as usual. 1284 Resource exhausting denial-of-service attacks take advantage of the 1285 cost of setting up a state for a protocol on the responder compared 1286 to the 'cheapness' on the initiator. HIP allows a responder to 1287 increase the cost of the start of state on the initiator and makes an 1288 effort to reduce the cost to the responder. This is done by having 1289 the responder start the authenticated Diffie-Hellman exchange instead 1290 of the initiator, making the HIP base exchange 4 packets long. The 1291 first packet sent by the responder can be prebuilt to further 1292 mitigate the costs. This packet also includes a computational puzzle 1293 that can optionally be used to further delay the initiator, for 1294 instance, when the responder is overloaded. The details are 1295 explained in the base exchange specification 1296 [I-D.ietf-hip-rfc5201-bis]. 1298 Man-in-the-middle (MitM) attacks are difficult to defend against, 1299 without third-party authentication. A skillful MitM could easily 1300 handle all parts of the HIP base exchange, but HIP indirectly 1301 provides the following protection from a MitM attack. If the 1302 responder's HI is retrieved from a signed DNS zone or securely 1303 obtained by some other means, the initiator can use this to 1304 authenticate the signed HIP packets. Likewise, if the initiator's HI 1305 is in a secure DNS zone, the responder can retrieve it and validate 1306 the signed HIP packets. However, since an initiator may choose to 1307 use an unpublished HI, it knowingly risks a MitM attack. The 1308 responder may choose not to accept a HIP exchange with an initiator 1309 using an unknown HI. 1311 Other types of MitM attacks against HIP can be mounted using ICMP 1312 messages that can be used to signal about problems. As a overall 1313 guideline, the ICMP messages should be considered as unreliable 1314 "hints" and should be acted upon only after timeouts. The exact 1315 attack scenarios and countermeasures are described in full detail the 1316 base exchange specification [I-D.ietf-hip-rfc5201-bis]. 1318 The need to support multiple hashes for generating the HIT from the 1319 HI affords the MitM to mount a potentially powerful downgrade attack 1320 due to the a-priori need of the HIT in the HIP base exchange. The 1321 base exchange has been augmented to deal with such an attack by 1322 restarting on detecting the attack. At worst this would only lead to 1323 a situation in which the base exchange would never finish (or would 1324 be aborted after some retries). As a drawback, this leads to an 1325 6-way base exchange which may seem bad at first. However, since this 1326 only occurs in an attack scenario and since the attack can be handled 1327 (so it is not interesting to mount anymore), we assume the subsequent 1328 messages do not represent a security threat. Since the MitM cannot 1329 be successful with a downgrade attack, these sorts of attacks will 1330 only occur as 'nuisance' attacks. So, the base exchange would still 1331 be usually just four packets even though implementations must be 1332 prepared to protect themselves against the downgrade attack. 1334 In HIP, the Security Association for ESP is indexed by the SPI; the 1335 source address is always ignored, and the destination address may be 1336 ignored as well. Therefore, HIP-enabled Encapsulated Security 1337 Payload (ESP) is IP address independent. This might seem to make 1338 attacking easier, but ESP with replay protection is already as well 1339 protected as possible, and the removal of the IP address as a check 1340 should not increase the exposure of ESP to DoS attacks. 1342 12.2. Protection against flooding attacks 1344 Although the idea of informing about address changes by simply 1345 sending packets with a new source address appears appealing, it is 1346 not secure enough. That is, even if HIP does not rely on the source 1347 address for anything (once the base exchange has been completed), it 1348 appears to be necessary to check a mobile node's reachability at the 1349 new address before actually sending any larger amounts of traffic to 1350 the new address. 1352 Blindly accepting new addresses would potentially lead to flooding 1353 Denial-of-Service attacks against third parties [RFC4225]. In a 1354 distributed flooding attack an attacker opens high volume HIP 1355 connections with a large number of hosts (using unpublished HIs), and 1356 then claims to all of these hosts that it has moved to a target 1357 node's IP address. If the peer hosts were to simply accept the move, 1358 the result would be a packet flood to the target node's address. To 1359 prevent this type of attack, HIP mobility extensions include a return 1360 routability check procedure where the reachability of a node is 1361 separately checked at each address before using the address for 1362 larger amounts of traffic. 1364 A credit-based authorization approach Host Mobility with the Host 1365 Identity Protocol [I-D.ietf-hip-rfc5206-bis] can be used between 1366 hosts for sending data prior to completing the address tests. 1367 Otherwise, if HIP is used between two hosts that fully trust each 1368 other, the hosts may optionally decide to skip the address tests. 1369 However, such performance optimization must be restricted to peers 1370 that are known to be trustworthy and capable of protecting themselves 1371 from malicious software. 1373 12.3. HITs used in ACLs 1375 At end-hosts, HITs can be used in IP-based access control lists at 1376 the application and network layers". At middleboxes, HIP-aware 1377 firewalls [lindqvist-enterprise] can use HITs or public keys to 1378 control both ingress and egress access to networks or individual 1379 hosts, even in the presence of mobile devices because the HITs and 1380 public keys are topologically independent. As discussed earlier in 1381 Section 7, once a HIP session has been established, the SPI value in 1382 an ESP packet may be used as an index, indicating the HITs. In 1383 practice, firewalls can inspect HIP packets to learn of the bindings 1384 between HITs, SPI values, and IP addresses. They can even explicitly 1385 control ESP usage, dynamically opening ESP only for specific SPI 1386 values and IP addresses. The signatures in HIP packets allow a 1387 capable firewall to ensure that the HIP exchange is indeed occurring 1388 between two known hosts. This may increase firewall security. 1390 A potential drawback of HITs in ACLs is their 'flatness' means they 1391 cannot be aggregated, and this could potentially result in larger 1392 table searches in HIP-aware firewalls. A way to optimize this could 1393 be to utilize bloom filters for grouping of HITs [sarela-bloom]. 1394 However, it should be noted that it is also easier to exclude 1395 individual, misbehaving hosts out when the firewall rules concern 1396 individual HITs rather than groups. 1398 There has been considerable bad experience with distributed ACLs that 1399 contain public key related material, for example, with SSH. If the 1400 owner of a key needs to revoke it for any reason, the task of finding 1401 all locations where the key is held in an ACL may be impossible. If 1402 the reason for the revocation is due to private key theft, this could 1403 be a serious issue. 1405 A host can keep track of all of its partners that might use its HIT 1406 in an ACL by logging all remote HITs. It should only be necessary to 1407 log responder hosts. With this information, the host can notify the 1408 various hosts about the change to the HIT. There has been attempts 1409 to develop a secure method to issue the HIT revocation notice 1410 [zhang-revocation]. 1412 Some of the HIP-aware middleboxes, such as firewalls 1413 [lindqvist-enterprise] or NATs [ylitalo-spinat], may observe the on- 1414 path traffic passively. Such middleboxes are transparent by their 1415 nature and may not get a notification when a host moves to a 1416 different network. Thus, such middleboxes should maintain soft state 1417 and timeout when the control and data plane between two HIP end-hosts 1418 has been idle too long. Correspondingly, the two end-hosts may send 1419 periodically keepalives, such as UPDATE packets or ICMP messages 1420 inside the ESP tunnel, to sustain state at the on-path middleboxes. 1422 Another aspect related to HIP-aware middleboxes is that the 1423 association between the control and data plane, in the case of ESP, 1424 is weak and can be exploited under certain assumptions as described 1425 by Heer et al[heer-end-host]. In the scenario, the attacker has 1426 already gained access to the target network protected by a HIP-aware 1427 firewall, but wants to circumvent the HIP-based firewall. To achieve 1428 this, the attacker passively observes a base exchange between two HIP 1429 hosts and later replays it. This way, the attacker manages to 1430 penetrate the firewall and can use a fake ESP tunnel to transport its 1431 own data. This is possible because the firewall cannot distinguish 1432 when the ESP tunnel is valid. As a solution, HIP-aware middleboxes 1433 may participate to the control plane interaction by adding random 1434 nonce parameters to the control traffic, which the the end-hosts have 1435 to sign to guarantee the freshness of the control traffic 1436 [heer-midauth]. As an alternative, extensions for transporting data 1437 plane directly over the control plane can be used [RFC6078]. 1439 12.4. Alternative HI considerations 1441 The definition of the Host Identifier states that the HI need not be 1442 a public key. It implies that the HI could be any value; for example 1443 a FQDN. This document does not describe how to support such a non- 1444 cryptographic HI, but examples of such protocol variants do exist 1445 ([urien-rfid], [urien-rfid-draft]). A non-cryptographic HI would 1446 still offer the services of the HIT or LSI for NAT traversal. It 1447 would be possible to carry HITs in HIP packets that had neither 1448 privacy nor authentication. Such schemes may be employed for 1449 resource constrained devices, such as small sensors operating on 1450 battery power, but are not further analyzed here. 1452 If it is desirable to use HIP in a low security situation where 1453 public key computations are considered expensive, HIP can be used 1454 with very short Diffie-Hellman and Host Identity keys. Such use 1455 makes the participating hosts vulnerable to MitM and connection 1456 hijacking attacks. However, it does not cause flooding dangers, 1457 since the address check mechanism relies on the routing system and 1458 not on cryptographic strength. 1460 13. IANA considerations 1462 This document has no actions for IANA. 1464 14. Acknowledgments 1466 For the people historically involved in the early stages of HIP, see 1467 the Acknowledgments section in the Host Identity Protocol 1468 specification. 1470 During the later stages of this document, when the editing baton was 1471 transferred to Pekka Nikander, the comments from the early 1472 implementers and others, including Jari Arkko, Tom Henderson, Petri 1473 Jokela, Miika Komu, Mika Kousa, Andrew McGregor, Jan Melen, Tim 1474 Shepard, Jukka Ylitalo, Sasu Tarkoma, and Jorma Wall, were 1475 invaluable. Also, the comments from Lars Eggert, Spencer Dawkins and 1476 Dave Crocker were also useful. 1478 The authors want to express their special thanks to Tom Henderson, 1479 who took the burden of editing the document in response to IESG 1480 comments at the time when both of the authors were busy doing other 1481 things. Without his perseverance original document might have never 1482 made it as RFC4423. 1484 This main effort to update and move HIP forward within the IETF 1485 process owes its impetuous to a number of HIP development teams. The 1486 authors are grateful for Boeing, Helsinki Institute for Information 1487 Technology (HIIT), NomadicLab of Ericsson, and the three 1488 universities: RWTH Aachen, Aalto and University of Helsinki, for 1489 their efforts. Without their collective efforts HIP would have 1490 withered as on the IETF vine as a nice concept. 1492 Thanks also for Suvi Koskinen for her help with proofreading and with 1493 the reference jungle. 1495 15. Changes from RFC 4423 1497 In a nutshell, the changes from RFC 4424 [RFC4423] are mostly 1498 editorial, including clarifications on topics described in a 1499 difficult way and omitting some of the non-architectural 1500 (implementation) details that are already described in other 1501 documents. A number of missing references to the literature were 1502 also added. New topics include the drawbacks of HIP, discussion on 1503 802.15.4 and MAC security, deployment considerations and description 1504 of the base exchange. 1506 16. References 1508 16.1. Normative References 1510 [I-D.ietf-hip-multihoming] 1511 Henderson, T., Vogt, C., and J. Arkko, "Host Multihoming 1512 with the Host Identity Protocol", 1513 draft-ietf-hip-multihoming-03 (work in progress), 1514 July 2013. 1516 [I-D.ietf-hip-native-nat-traversal] 1517 Keranen, A. and J. Melen, "Native NAT Traversal Mode for 1518 the Host Identity Protocol", 1519 draft-ietf-hip-native-nat-traversal-05 (work in progress), 1520 June 2013. 1522 [I-D.ietf-hip-rfc5201-bis] 1523 Moskowitz, R., Heer, T., Jokela, P., and T. Henderson, 1524 "Host Identity Protocol Version 2 (HIPv2)", 1525 draft-ietf-hip-rfc5201-bis-14 (work in progress), 1526 October 2013. 1528 [I-D.ietf-hip-rfc5202-bis] 1529 Jokela, P., Moskowitz, R., and J. Melen, "Using the 1530 Encapsulating Security Payload (ESP) Transport Format with 1531 the Host Identity Protocol (HIP)", 1532 draft-ietf-hip-rfc5202-bis-04 (work in progress), 1533 September 2013. 1535 [I-D.ietf-hip-rfc5203-bis] 1536 Laganier, J. and L. Eggert, "Host Identity Protocol (HIP) 1537 Registration Extension", draft-ietf-hip-rfc5203-bis-02 1538 (work in progress), September 2012. 1540 [I-D.ietf-hip-rfc5204-bis] 1541 Laganier, J. and L. Eggert, "Host Identity Protocol (HIP) 1542 Rendezvous Extension", draft-ietf-hip-rfc5204-bis-02 (work 1543 in progress), September 2012. 1545 [I-D.ietf-hip-rfc5205-bis] 1546 Laganier, J., "Host Identity Protocol (HIP) Domain Name 1547 System (DNS) Extension", draft-ietf-hip-rfc5205-bis-02 1548 (work in progress), September 2012. 1550 [I-D.ietf-hip-rfc5206-bis] 1551 Henderson, T., Vogt, C., and J. Arkko, "Host Mobility with 1552 the Host Identity Protocol", draft-ietf-hip-rfc5206-bis-06 1553 (work in progress), July 2013. 1555 [I-D.ietf-hip-rfc6253-bis] 1556 Heer, T. and S. Varjonen, "Host Identity Protocol 1557 Certificates", draft-ietf-hip-rfc6253-bis-01 (work in 1558 progress), October 2013. 1560 [RFC5482] Eggert, L. and F. Gont, "TCP User Timeout Option", 1561 RFC 5482, March 2009. 1563 16.2. Informative references 1565 [IEEE.802-15-4.2011] 1566 "Information technology - Telecommunications and 1567 information exchange between systems - Local and 1568 metropolitan area networks - Specific requirements - Part 1569 15.4: Wireless Medium Access Control (MAC) and Physical 1570 Layer (PHY) Specifications for Low-Rate Wireless Personal 1571 Area Networks (WPANs)", IEEE Standard 802.15.4, 1572 September 2011, . 1575 [Nik2001] Nikander, P., "Denial-of-Service, Address Ownership, and 1576 Early Authentication in the IPv6 World", in Proceesings 1577 of Security Protocols, 9th International Workshop, 1578 Cambridge, UK, April 25-27 2001, LNCS 2467, pp. 12-26, 1579 Springer, 2002. 1581 [RFC2136] Vixie, P., Thomson, S., Rekhter, Y., and J. Bound, 1582 "Dynamic Updates in the Domain Name System (DNS UPDATE)", 1583 RFC 2136, April 1997. 1585 [RFC2535] Eastlake, D., "Domain Name System Security Extensions", 1586 RFC 2535, March 1999. 1588 [RFC2766] Tsirtsis, G. and P. Srisuresh, "Network Address 1589 Translation - Protocol Translation (NAT-PT)", RFC 2766, 1590 February 2000. 1592 [RFC3022] Srisuresh, P. and K. Egevang, "Traditional IP Network 1593 Address Translator (Traditional NAT)", RFC 3022, 1594 January 2001. 1596 [RFC3102] Borella, M., Lo, J., Grabelsky, D., and G. Montenegro, 1597 "Realm Specific IP: Framework", RFC 3102, October 2001. 1599 [RFC3748] Aboba, B., Blunk, L., Vollbrecht, J., Carlson, J., and H. 1600 Levkowetz, "Extensible Authentication Protocol (EAP)", 1601 RFC 3748, June 2004. 1603 [RFC4225] Nikander, P., Arkko, J., Aura, T., Montenegro, G., and E. 1604 Nordmark, "Mobile IP Version 6 Route Optimization Security 1605 Design Background", RFC 4225, December 2005. 1607 [RFC4306] Kaufman, C., "Internet Key Exchange (IKEv2) Protocol", 1608 RFC 4306, December 2005. 1610 [RFC4423] Moskowitz, R. and P. Nikander, "Host Identity Protocol 1611 (HIP) Architecture", RFC 4423, May 2006. 1613 [RFC5218] Thaler, D. and B. Aboba, "What Makes For a Successful 1614 Protocol?", RFC 5218, July 2008. 1616 [RFC5338] Henderson, T., Nikander, P., and M. Komu, "Using the Host 1617 Identity Protocol with Legacy Applications", RFC 5338, 1618 September 2008. 1620 [RFC5887] Carpenter, B., Atkinson, R., and H. Flinck, "Renumbering 1621 Still Needs Work", RFC 5887, May 2010. 1623 [RFC6078] Camarillo, G. and J. Melen, "Host Identity Protocol (HIP) 1624 Immediate Carriage and Conveyance of Upper-Layer Protocol 1625 Signaling (HICCUPS)", RFC 6078, January 2011. 1627 [RFC6250] Thaler, D., "Evolution of the IP Model", RFC 6250, 1628 May 2011. 1630 [RFC6281] Cheshire, S., Zhu, Z., Wakikawa, R., and L. Zhang, 1631 "Understanding Apple's Back to My Mac (BTMM) Service", 1632 RFC 6281, June 2011. 1634 [RFC6317] Komu, M. and T. Henderson, "Basic Socket Interface 1635 Extensions for the Host Identity Protocol (HIP)", 1636 RFC 6317, July 2011. 1638 [RFC6537] Ahrenholz, J., "Host Identity Protocol Distributed Hash 1639 Table Interface", RFC 6537, February 2012. 1641 [RFC6538] Henderson, T. and A. Gurtov, "The Host Identity Protocol 1642 (HIP) Experiment Report", RFC 6538, March 2012. 1644 [amir-hip] 1645 Amir, K., Forsgren, H., Grahn, K., Karvi, T., and G. 1646 Pulkkis, "Security and Trust of Public Key Cryptography 1647 for HIP and HIP Multicast", International Journal of 1648 Dependable and Trustworthy Information Systems (IJDTIS), 1649 2(3), 17-35, DOI: 10.4018/jdtis.2011070102, 2013. 1651 [aura-dos] 1652 Aura, T., Nikander, P., and J. Leiwo, "DOS-resistant 1653 Authentication with Client Puzzles", 8th International 1654 Workshop on Security Protocols, pages 170-177. Springer, , 1655 April 2001. 1657 [beal-dos] 1658 Beal, J. and T. Shephard, "Deamplification of DoS Attacks 1659 via Puzzles", , October 2004. 1661 [chiappa-endpoints] 1662 Chiappa, J., "Endpoints and Endpoint Names: A Proposed 1663 Enhancement to the Internet Architecture", 1664 URL http://www.chiappa.net/~jnc/tech/endpoints.txt, 1999. 1666 [heer-end-host] 1667 Heer, T., Hummen, R., Komu, M., Goetz, S., and K. Wehre, 1668 "End-host Authentication and Authorization for Middleboxes 1669 based on a Cryptographic Namespace", ICC2009 Communication 1670 and Information Systems Security Symposium, , 2009. 1672 [heer-midauth] 1673 Heer, T. and M. Komu, "End-Host Authentication for HIP 1674 Middleboxes", Working draft draft-heer-hip-middle-auth-02, 1675 September 2009. 1677 [henderson-vpls] 1678 Henderson, T. and D. Mattes, "", Working 1679 draft draft-henderson-hip-vpls-06, June 2013. 1681 [hip-srtp] 1682 Tschofenig, H., Muenz, F., and M. Shanmugam, "Using SRTP 1683 transport format with HIP", Working 1684 draft draft-tschofenig-hiprg-hip-srtp-01, October 2005. 1686 [karvonen-usable] 1687 Karvonen, K., Komu, M., and A. Gurtov, "Usable Security 1688 Management with Host Identity Protocol", 7th ACS/IEEE 1689 International Conference on Computer Systems and 1690 Applications, (AICCSA-2009), 2009. 1692 [komu-cloud] 1693 Komu, M., Sethi, M., Mallavarapu, R., Oirola, H., Khan, 1694 R., and S. Tarkoma, "Secure Networking for Virtual 1695 Machines in the Cloud", International Workshop on Power 1696 and QoS Aware Computing (PQoSCom2012), IEEE, ISBN: 978-1- 1697 4244-8567-3, September 2012. 1699 [komu-diss] 1700 Komu, M., "A Consolidated Namespace for Network 1701 Applications, Developers, Administrators and Users", 1702 Dissertation, Aalto University, Espoo, Finland ISBN: 978- 1703 952-60-4904-5 (printed), ISBN: 978-952-60-4905-2 1704 (electronic). , December 2012. 1706 [komu-leap] 1707 Komu, M. and J. Lindqvist, "Leap-of-Faith Security is 1708 Enough for IP Mobility", 6th Annual IEEE Consumer 1709 Communications and Networking Conference IEEE CCNC 2009, 1710 Las Vegas, Nevada, , January 2009. 1712 [komu-mitigation] 1713 Komu, M., Tarkoma, S., and A. Lukyanenko, "Mitigation of 1714 Unsolicited Traffic Across Domains with Host Identities 1715 and Puzzles", 15th Nordic Conference on Secure IT Systems 1716 (NordSec 2010), Springer Lecture Notes in Computer 1717 Science, Volume 7127, pp. 33-48, ISBN: 978-3-642-27936-2, 1718 October 2010. 1720 [kovacshazi-host] 1721 Kovacshazi, Z. and R. Vida, "Host Identity Specific 1722 Multicast", International conference on Networking and 1723 Services (ICNS'06), IEEE Computer Society, Los Alamitos, 1724 CA, USA, http://doi.ieeecomputersociety.org/10.1109/ 1725 ICNS.2007.66, 2007. 1727 [leva-barriers] 1728 Levae, A., Komu, M., and S. Luukkainen, "Adoption Barriers 1729 of Network-layer Protocols: the Case of Host Identity 1730 Protocol", The International Journal of Computer and 1731 Telecommunications Networking, ISSN: 1389-1286, 1732 March 2013. 1734 [lindqvist-enterprise] 1735 Lindqvist, J., Vehmersalo, E., Manner, J., and M. Komu, 1736 "Enterprise Network Packet Filtering for Mobile 1737 Cryptographic Identities", International Journal of 1738 Handheld Computing Research, 1 (1), 79-94, , January- 1739 March 2010. 1741 [nsrg-report] 1742 Lear, E. and R. Droms, "What's In A Name:Thoughts from the 1743 NSRG", draft-irtf-nsrg-report-10 (work in progress), 1744 September 2003. 1746 [paine-hip] 1747 Paine, R., "Beyond HIP: The End to Hacking As We Know It", 1748 BookSurge Publishing, ISBN: 1439256047, 9781439256046, 1749 2009. 1751 [pham-leap] 1752 Pham, V. and T. Aura, "Security Analysis of Leap-of-Faith 1753 Protocols", Seventh ICST International Conference on 1754 Security and Privacy for Communication Networks, , 1755 September 2011. 1757 [sarela-bloom] 1758 Saerelae, M., Esteve Rothenberg, C., Zahemszky, A., 1759 Nikander, P., and J. Ott, "BloomCasting: Security in Bloom 1760 filter based multicast", , Lecture Notes in Computer 1761 Science 2012, , pages 1-16, Springer Berlin Heidelberg, 1762 2012. 1764 [scultz-intermittent] 1765 Schuetz, S., Eggert, L., Schmid, S., and M. Brunner, 1766 "Protocol enhancements for intermittently connected 1767 hosts", SIGCOMM Comput. Commun. Rev., 35(3):5-18, , 1768 July 2005. 1770 [shields-hip] 1771 Shields, C. and J. Garcia-Luna-Aceves, "The HIP protocol 1772 for hierarchical multicast routing", Proceedings of the 1773 seventeenth annual ACM symposium on Principles of 1774 distributed computing, pages 257-266. ACM, New York, NY, 1775 USA, ISBN: 0-89791-977-7, DOI: 10.1145/277697.277744, 1776 1998. 1778 [tritilanunt-dos] 1779 Tritilanunt, S., Boyd, C., Foo, E., and J. Nieto, 1780 "Examining the DoS Resistance of HIP", OTM Workshops (1), 1781 volume 4277 of Lecture Notes in Computer Science, pages 1782 616-625,Springer , 2006. 1784 [urien-rfid] 1785 Urien, P., Chabanne, H., Bouet, M., de Cunha, D., Guyot, 1786 V., Pujolle, G., Paradinas, P., Gressier, E., and J. 1787 Susini, "HIP-based RFID Networking Architecture", IFIP 1788 International Conference on Wireless and Optical 1789 Communications Networks, DOI: 10.1109/WOCN.2007.4284140, 1790 July 2007. 1792 [urien-rfid-draft] 1793 Urien, P., Lee, G., and G. Pujolle, "HIP support for 1794 RFIDs", IRTF Working draft draft-irtf-hiprg-rfid-07, 1795 April 2013. 1797 [varjonen-split] 1798 Varjonen, S., Komu, M., and A. Gurtov, "Secure and 1799 Efficient IPv4/IPv6 Handovers Using Host-Based Identifier- 1800 Location Split", Journal of Communications Software and 1801 Systems, 6(1), 2010, ISSN: 18456421, 2010. 1803 [xin-hip-lib] 1804 Xin, G., "Host Identity Protocol Version 2.5", Master's 1805 Thesis, Aalto University, Espoo, Finland, , June 2012. 1807 [xueyong-hip] 1808 Xueyong, Z., Zhiguo, D., and W. Xinling, "A Multicast 1809 Routing Algorithm Applied to HIP-Multicast Model", 1810 Proceedings of the 2011 International Conference on 1811 Network Computing and Information Security - Volume 01 1812 (NCIS '11), Vol. 1. IEEE Computer Society, Washington, DC, 1813 USA, pages 169-174, DOI: 10.1109/NCIS.2011.42, 2011. 1815 [xueyong-secure] 1816 Xueyong, Z. and J. Atwood, "A Secure Multicast Model for 1817 Peer-to-Peer and Access Networks Using the Host Identity 1818 Protocol", Consumer Communications and Networking 1819 Conference. CCNC 2007. 4th IEEE, pages 1098,1102, DOI: 1820 10.1109/CCNC.2007.221, January 2007. 1822 [ylitalo-diss] 1823 Ylitalo, J., "Secure Mobility at Multiple Granularity 1824 Levels over Heterogeneous Datacom Networks", Dissertation, 1825 Helsinki University of Technology, Espoo, Finland ISBN 1826 978-951-22-9531-9, 2008. 1828 [ylitalo-spinat] 1829 Ylitalo, J., Salmela, P., and H. Tschofenig, "SPINAT: 1831 Integrating IPsec into overlay routing", Proceedings of 1832 the First International Conference on Security and Privacy 1833 for Emerging Areas in Communication Networks (SecureComm 1834 2005). Athens, Greece. IEEE Computer Society, pages 315- 1835 326, ISBN: 0-7695-2369-2, September 2005. 1837 [zhang-revocation] 1838 Zhang, D., Kuptsov, D., and S. Shen, "Host Identifier 1839 Revocation in HIP", IRTF Working 1840 draft draft-irtf-hiprg-revocation-05, Mar 2012. 1842 Authors' Addresses 1844 Robert Moskowitz (editor) 1845 Verizon 1846 1000 Bent Creek Blvd, Suite 200 1847 Mechanicsburg, PA 1848 USA 1850 Email: robert.moskowitz@verizon.com 1852 Miika Komu 1853 Aalto University 1854 Konemiehentie 2 1855 Espoo 1856 Finland 1858 Email: miika.komu@aalto.fi