idnits 2.17.1 draft-ietf-hip-rfc4423-bis-10.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year == The document seems to contain a disclaimer for pre-RFC5378 work, but was first submitted on or after 10 November 2008. The disclaimer is usually necessary only for documents that revise or obsolete older RFCs, and that take significant amounts of text from those RFCs. If you can contact all authors of the source material and they are willing to grant the BCP78 rights to the IETF Trust, you can and should remove the disclaimer. Otherwise, the disclaimer is needed and you can ignore this comment. (See the Legal Provisions document at https://trustee.ietf.org/license-info for more information.) -- The document date (April 12, 2015) is 3301 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for intended status: Informational ---------------------------------------------------------------------------- == Outdated reference: A later version (-12) exists of draft-ietf-hip-multihoming-05 == Outdated reference: A later version (-33) exists of draft-ietf-hip-native-nat-traversal-08 == Outdated reference: A later version (-11) exists of draft-ietf-hip-rfc5203-bis-06 == Outdated reference: A later version (-08) exists of draft-ietf-hip-rfc5204-bis-04 == Outdated reference: A later version (-10) exists of draft-ietf-hip-rfc5205-bis-05 == Outdated reference: A later version (-14) exists of draft-ietf-hip-rfc5206-bis-06 == Outdated reference: A later version (-09) exists of draft-ietf-hip-rfc6253-bis-01 -- Obsolete informational reference (is this intentional?): RFC 2535 (Obsoleted by RFC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 (~~), 9 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 Ericsson 6 Expires: October 14, 2015 April 12, 2015 8 Host Identity Protocol Architecture 9 draft-ietf-hip-rfc4423-bis-10 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 October 14, 2015. 42 Copyright Notice 44 Copyright (c) 2015 IETF Trust and the persons identified as the 45 document authors. All rights reserved. 47 This document is subject to BCP 78 and the IETF Trust's Legal 48 Provisions Relating to IETF Documents 49 (http://trustee.ietf.org/license-info) in effect on the date of 50 publication of this document. Please review these documents 51 carefully, as they describe your rights and restrictions with respect 52 to this document. Code Components extracted from this document must 53 include Simplified BSD License text as described in Section 4.e of 54 the Trust Legal Provisions and are provided without warranty as 55 described in the Simplified BSD License. 57 This document may contain material from IETF Documents or IETF 58 Contributions published or made publicly available before November 59 10, 2008. The person(s) controlling the copyright in some of this 60 material may not have granted the IETF Trust the right to allow 61 modifications of such material outside the IETF Standards Process. 62 Without obtaining an adequate license from the person(s) controlling 63 the copyright in such materials, this document may not be modified 64 outside the IETF Standards Process, and derivative works of it may 65 not be created outside the IETF Standards Process, except to format 66 it for publication as an RFC or to translate it into languages other 67 than English. 69 Table of Contents 71 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 3 72 2. Terminology . . . . . . . . . . . . . . . . . . . . . . . . . 4 73 2.1. Terms common to other documents . . . . . . . . . . . . . . 4 74 2.2. Terms specific to this and other HIP documents . . . . . . 5 75 3. Background . . . . . . . . . . . . . . . . . . . . . . . . . 6 76 3.1. A desire for a namespace for computing platforms . . . . . 6 77 4. Host Identity namespace . . . . . . . . . . . . . . . . . . . 8 78 4.1. Host Identifiers . . . . . . . . . . . . . . . . . . . . . 9 79 4.2. Host Identity Hash (HIH) . . . . . . . . . . . . . . . . . 10 80 4.3. Host Identity Tag (HIT) . . . . . . . . . . . . . . . . . . 10 81 4.4. Local Scope Identifier (LSI) . . . . . . . . . . . . . . . 11 82 4.5. Storing Host Identifiers in directories . . . . . . . . . . 12 83 5. New stack architecture . . . . . . . . . . . . . . . . . . . 12 84 5.1. On the multiplicity of identities . . . . . . . . . . . . . 13 85 6. Control plane . . . . . . . . . . . . . . . . . . . . . . . . 14 86 6.1. Base exchange . . . . . . . . . . . . . . . . . . . . . . . 14 87 6.2. End-host mobility and multi-homing . . . . . . . . . . . . 15 88 6.3. Rendezvous mechanism . . . . . . . . . . . . . . . . . . . 16 89 6.4. Relay mechanism . . . . . . . . . . . . . . . . . . . . . . 16 90 6.5. Termination of the control plane . . . . . . . . . . . . . 16 91 7. Data plane . . . . . . . . . . . . . . . . . . . . . . . . . 16 92 8. HIP and NATs . . . . . . . . . . . . . . . . . . . . . . . . 17 93 8.1. HIP and Upper-layer checksums . . . . . . . . . . . . . . . 18 94 9. Multicast . . . . . . . . . . . . . . . . . . . . . . . . . . 18 95 10. HIP policies . . . . . . . . . . . . . . . . . . . . . . . . 19 96 11. Design considerations . . . . . . . . . . . . . . . . . . . . 20 97 11.1. Benefits of HIP . . . . . . . . . . . . . . . . . . . . . 20 98 11.2. Drawbacks of HIP . . . . . . . . . . . . . . . . . . . . . 22 99 11.3. Deployment and adoption considerations . . . . . . . . . . 23 100 11.3.1. Deployment analysis . . . . . . . . . . . . . . . . . . 24 101 11.3.2. HIP in 802.15.4 networks . . . . . . . . . . . . . . . . 25 102 11.4. Answers to NSRG questions . . . . . . . . . . . . . . . . 25 103 12. Security considerations . . . . . . . . . . . . . . . . . . . 27 104 12.1. MiTM Attacks . . . . . . . . . . . . . . . . . . . . . . . 27 105 12.2. Protection against flooding attacks . . . . . . . . . . . 28 106 12.3. HITs used in ACLs . . . . . . . . . . . . . . . . . . . . 29 107 12.4. Alternative HI considerations . . . . . . . . . . . . . . 31 108 13. IANA considerations . . . . . . . . . . . . . . . . . . . . . 31 109 14. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 31 110 15. Changes from RFC 4423 . . . . . . . . . . . . . . . . . . . . 32 111 16. References . . . . . . . . . . . . . . . . . . . . . . . . . 32 112 16.1. Normative References . . . . . . . . . . . . . . . . . . . 32 113 16.2. Informative references . . . . . . . . . . . . . . . . . . 33 114 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 39 116 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 (although there 133 may be transient periods of time such as key replacement when more 134 than one identifier may be active). While this text later talks 135 about non-cryptographic Host Identifiers, the architecture focuses on 136 the case in which Host Identifiers are cryptographic in nature. 137 Specifically, the Host Identifier is the public key of an asymmetric 138 key-pair. Each Host Identity uniquely identifies a single host, 139 i.e., no two hosts have the same Host Identity. If two or more 140 computing platforms have the same Host Identifier, then they are 141 instantiating a distributed host. The Host Identifier can either be 142 public (e.g. published in the DNS), or unpublished. Client systems 143 will tend to have both public and unpublished Host Identifiers. 145 There is a subtle but important difference between Host Identities 146 and Host Identifiers. An Identity refers to the abstract entity that 147 is identified. An Identifier, on the other hand, refers to the 148 concrete bit pattern that is used in the identification process. 150 Although the Host Identifiers could be used in many authentication 151 systems, such as IKEv2 [RFC4306], the presented architecture 152 introduces a new protocol, called the Host Identity Protocol (HIP), 153 and a cryptographic exchange, called the HIP base exchange; see also 154 Section 6. HIP provides for limited forms of trust between systems, 155 enhance mobility, multi-homing and dynamic IP renumbering, aid in 156 protocol translation / transition, and reduce certain types of 157 denial-of-service (DoS) attacks. 159 When HIP is used, the actual payload traffic between two HIP hosts is 160 typically, but not necessarily, protected with ESP [RFC7402]. The 161 Host Identities are used to create the needed ESP Security 162 Associations (SAs) and to authenticate the hosts. When ESP is used, 163 the actual payload IP packets do not differ in any way from standard 164 ESP protected IP packets. 166 Much has been learned about HIP [RFC6538] since [RFC4423] was 167 published. This document expands Host Identities beyond use to 168 enable IP connectivity and security to general interhost secure 169 signalling at any protocol layer. The signal may establish a 170 security association between the hosts, or simply pass information 171 within the channel. 173 2. Terminology 175 2.1. Terms common to other documents 177 +-------------+-----------------------------------------------------+ 178 | Term | Explanation | 179 +-------------+-----------------------------------------------------+ 180 | Public key | The public key of an asymmetric cryptographic key | 181 | | pair. Used as a publicly known identifier for | 182 | | cryptographic identity authentication. Public is a | 183 | | a relative term here, ranging from "known to peers | 184 | | only" to "known to the world." | 185 | Private key | The private or secret key of an asymmetric | 186 | | cryptographic key pair. Assumed to be known only | 187 | | to the party identified by the corresponding public | 188 | | key. Used by the identified party to authenticate | 189 | | its identity to other parties. | 190 | Public key | An asymmetric cryptographic key pair consisting of | 191 | pair | public and private keys. For example, Rivest- | 192 | | Shamir-Adleman (RSA), Digital Signature Algorithm | 193 | | (DSA) and Elliptic Curve DSA (ECDSA) key pairs are | 194 | | such key pairs. | 195 | End-point | A communicating entity. For historical reasons, | 196 | | the term 'computing platform' is used in this | 197 | | document as a (rough) synonym for end-point. | 198 +-------------+-----------------------------------------------------+ 200 2.2. Terms specific to this and other HIP documents 202 It should be noted that many of the terms defined herein are 203 tautologous, self-referential or defined through circular reference 204 to other terms. This is due to the succinct nature of the 205 definitions. See the text elsewhere in this document and the base 206 specification [RFC7401] for more elaborate explanations. 208 +-------------------+-----------------------------------------------+ 209 | Term | Explanation | 210 +-------------------+-----------------------------------------------+ 211 | Computing | An entity capable of communicating and | 212 | platform | computing, for example, a computer. See the | 213 | | definition of 'End-point', above. | 214 | HIP base exchange | A cryptographic protocol; see also Section 6. | 215 | HIP packet | An IP packet that carries a 'Host Identity | 216 | | Protocol' message. | 217 | Host Identity | An abstract concept assigned to a 'computing | 218 | | platform'. See 'Host Identifier', below. | 219 | Host Identity | A name space formed by all possible Host | 220 | namespace | Identifiers. | 221 | Host Identity | A protocol used to carry and authenticate | 222 | Protocol | Host Identifiers and other information. | 223 | Host Identity | The cryptographic hash used in creating the | 224 | Hash | Host Identity Tag from the Host Identity. | 225 | Host Identity Tag | A 128-bit datum created by taking a | 226 | | cryptographic hash over a Host Identifier | 227 | | plus bits to identify which hash used. | 228 | Host Identifier | A public key used as a name for a Host | 229 | | Identity. | 230 | Local Scope | A 32-bit datum denoting a Host Identity. | 231 | Identifier | | 232 | Public Host | A published or publicly known Host Identifier | 233 | Identifier and | used as a public name for a Host Identity, | 234 | Identity | and the corresponding Identity. | 235 | Unpublished Host | A Host Identifier that is not placed in any | 236 | Identifier and | public directory, and the corresponding Host | 237 | Identity | Identity. Unpublished Host Identities are | 238 | | typically short lived in nature, being often | 239 | | replaced and possibly used just once. | 240 | Rendezvous | A mechanism used to locate mobile hosts based | 241 | Mechanism | on their HIT. | 242 +-------------------+-----------------------------------------------+ 244 3. Background 246 The Internet is built from three principal components: computing 247 platforms (end-points), packet transport (i.e., internetworking) 248 infrastructure, and services (applications). The Internet exists to 249 service two principal components: people and robotic services 250 (silicon-based people, if you will). All these components need to be 251 named in order to interact in a scalable manner. Here we concentrate 252 on naming computing platforms and packet transport elements. 254 There are two principal namespaces in use in the Internet for these 255 components: IP addresses, and Domain Names. Domain Names provide 256 hierarchically assigned names for some computing platforms and some 257 services. Each hierarchy is delegated from the level above; there is 258 no anonymity in Domain Names. Email, HTTP, and SIP addresses all 259 reference Domain Names. 261 The IP addressing namespace has been overloaded to name both 262 interfaces (at layer-3) and endpoints (for the endpoint-specific part 263 of layer-3, and for layer-4). In their role as interface names, IP 264 addresses are sometimes called "locators" and serve as an endpoint 265 within a routing topology. 267 IP addresses are numbers that name networking interfaces, and 268 typically only when the interface is connected to the network. 269 Originally, IP addresses had long-term significance. Today, the vast 270 number of interfaces use ephemeral and/or non-unique IP addresses. 271 That is, every time an interface is connected to the network, it is 272 assigned an IP address. 274 In the current Internet, the transport layers are coupled to the IP 275 addresses. Neither can evolve separately from the other. IPng 276 deliberations were strongly shaped by the decision that a 277 corresponding TCPng would not be created. 279 There are three critical deficiencies with the current namespaces. 280 Firstly, dynamic readdressing cannot be directly managed. Secondly, 281 confidentiality is not provided in a consistent, trustable manner. 282 Finally, authentication for systems and datagrams is not provided. 283 All of these deficiencies arise because computing platforms are not 284 well named with the current namespaces. 286 3.1. A desire for a namespace for computing platforms 287 An independent namespace for computing platforms could be used in 288 end-to-end operations independent of the evolution of the 289 internetworking layer and across the many internetworking layers. 290 This could support rapid readdressing of the internetworking layer 291 because of mobility, rehoming, or renumbering. 293 If the namespace for computing platforms is based on public-key 294 cryptography, it can also provide authentication services. If this 295 namespace is locally created without requiring registration, it can 296 provide anonymity. 298 Such a namespace (for computing platforms) and the names in it should 299 have the following characteristics: 301 o The namespace should be applied to the IP 'kernel' or stack. The 302 IP stack is the 'component' between applications and the packet 303 transport infrastructure. 305 o The namespace should fully decouple the internetworking layer from 306 the higher layers. The names should replace all occurrences of IP 307 addresses within applications (like in the Transport Control 308 Block, TCB). This replacement can be handled transparently for 309 legacy applications as the LSIs and HITs are compatible with IPv4 310 and IPv6 addresses [RFC5338]. However, HIP-aware applications 311 require some modifications from the developers, who may employ 312 networking API extensions for HIP [RFC6317]. 314 o The introduction of the namespace should not mandate any 315 administrative infrastructure. Deployment must come from the 316 bottom up, in a pairwise deployment. 318 o The names should have a fixed length representation, for easy 319 inclusion in datagram headers and existing programming interfaces 320 (e.g the TCB). 322 o Using the namespace should be affordable when used in protocols. 323 This is primarily a packet size issue. There is also a 324 computational concern in affordability. 326 o Name collisions should be avoided as much as possible. The 327 mathematics of the birthday paradox can be used to estimate the 328 chance of a collision in a given population and hash space. In 329 general, for a random hash space of size n bits, we would expect 330 to obtain a collision after approximately 1.2*sqrt(2**n) hashes 331 were obtained. For 64 bits, this number is roughly 4 billion. A 332 hash size of 64 bits may be too small to avoid collisions in a 333 large population; for example, there is a 1% chance of collision 334 in a population of 640M. For 100 bits (or more), we would not 335 expect a collision until approximately 2**50 (1 quadrillion) 336 hashes were generated. 338 o The names should have a localized abstraction so that they can be 339 used in existing protocols and APIs. 341 o It must be possible to create names locally. When such names are 342 not published, this can provide anonymity at the cost of making 343 resolvability very difficult. 345 o The namespace should provide authentication services. 347 o The names should be long lived, but replaceable at any time. This 348 impacts access control lists; short lifetimes will tend to result 349 in tedious list maintenance or require a namespace infrastructure 350 for central control of access lists. 352 In this document, a new namespace approaching these ideas is called 353 the Host Identity namespace. Using Host Identities requires its own 354 protocol layer, the Host Identity Protocol, between the 355 internetworking and transport layers. The names are based on public- 356 key cryptography to supply authentication services. Properly 357 designed, it can deliver all of the above stated requirements. 359 4. Host Identity namespace 361 A name in the Host Identity namespace, a Host Identifier (HI), 362 represents a statistically globally unique name for naming any system 363 with an IP stack. This identity is normally associated with, but not 364 limited to, an IP stack. A system can have multiple identities, some 365 'well known', some unpublished or 'anonymous'. A system may self- 366 assert its own identity, or may use a third-party authenticator like 367 DNSSEC [RFC2535], PGP, or X.509 to 'notarize' the identity assertion 368 to another namespace. It is expected that the Host Identifiers will 369 initially be authenticated with DNSSEC and that all implementations 370 will support DNSSEC as a minimal baseline. 372 In theory, any name that can claim to be 'statistically globally 373 unique' may serve as a Host Identifier. In the HIP architecture, the 374 public key of a private-public key pair has been chosen as the Host 375 Identifier because it can be self managed and it is computationally 376 difficult to forge. As specified in the Host Identity Protocol 377 [RFC7401] specification, a public-key-based HI can authenticate the 378 HIP packets and protect them from man-in-the-middle attacks. Since 379 authenticated datagrams are mandatory to provide much of HIP's 380 denial-of-service protection, the Diffie-Hellman exchange in HIP base 381 exchange has to be authenticated. Thus, only public-key HI and 382 authenticated HIP messages are supported in practice. 384 In this document, the non-cryptographic forms of HI and HIP are 385 presented to complete the theory of HI, but they should not be 386 implemented as they could produce worse denial-of-service attacks 387 than the Internet has without Host Identity. There has been past 388 research in challenge puzzles to use non-cryptographic HI, for Radio 389 Frequency IDentification (RFID), in an HIP exchange tailored to the 390 workings of such challenges (as described further in [urien-rfid] and 391 [urien-rfid-draft]). 393 4.1. Host Identifiers 395 Host Identity adds two main features to Internet protocols. The 396 first is a decoupling of the internetworking and transport layers; 397 see Section 5. This decoupling will allow for independent evolution 398 of the two layers. Additionally, it can provide end-to-end services 399 over multiple internetworking realms. The second feature is host 400 authentication. Because the Host Identifier is a public key, this 401 key can be used for authentication in security protocols like ESP. 403 The only completely defined structure of the Host Identity is that of 404 a public/private key pair. In this case, the Host Identity is 405 referred to by its public component, the public key. Thus, the name 406 representing a Host Identity in the Host Identity namespace, i.e., 407 the Host Identifier, is the public key. In a way, the possession of 408 the private key defines the Identity itself. If the private key is 409 possessed by more than one node, the Identity can be considered to be 410 a distributed one. 412 Architecturally, any other Internet naming convention might form a 413 usable base for Host Identifiers. However, non-cryptographic names 414 should only be used in situations of high trust - low risk. That is 415 any place where host authentication is not needed (no risk of host 416 spoofing) and no use of ESP. However, at least for interconnected 417 networks spanning several operational domains, the set of 418 environments where the risk of host spoofing allowed by non- 419 cryptographic Host Identifiers is acceptable is the null set. Hence, 420 the current HIP documents do not specify how to use any other types 421 of Host Identifiers but public keys. For instance, Back-to-My-Mac 422 [RFC6281] from Apple comes pretty close to the functionality of HIP, 423 but unlike HIP, it is based on non-cryptographic identifiers. 425 The actual Host Identifiers are never directly used at the transport 426 or network layers. The corresponding Host Identifiers (public keys) 427 may be stored in various DNS or other directories as identified 428 elsewhere in this document, and they are passed in the HIP base 429 exchange. A Host Identity Tag (HIT) is used in other protocols to 430 represent the Host Identity. Another representation of the Host 431 Identities, the Local Scope Identifier (LSI), can also be used in 432 protocols and APIs. 434 4.2. Host Identity Hash (HIH) 436 The Host Identity Hash is the cryptographic hash algorithm used in 437 producing the HIT from the HI. It is also the hash used throughout 438 the HIP protocol for consistency and simplicity. It is possible to 439 for the two hosts in the HIP exchange to use different hash 440 algorithms. 442 Multiple HIHs within HIP are needed to address the moving target of 443 creation and eventual compromise of cryptographic hashes. This 444 significantly complicates HIP and offers an attacker an additional 445 downgrade attack that is mitigated in the HIP protocol [RFC7401]. 447 4.3. Host Identity Tag (HIT) 449 A Host Identity Tag is a 128-bit representation for a Host Identity. 450 It is created from an HIH, an IPv6 prefix [RFC7343] and a hash 451 identifier. There are two advantages of using the HIT over using the 452 Host Identifier in protocols. Firstly, its fixed length makes for 453 easier protocol coding and also better manages the packet size cost 454 of this technology. Secondly, it presents the identity in a 455 consistent format to the protocol independent of the cryptographic 456 algorithms used. 458 In essence, the HIT is a hash over the public key. As such, two 459 algorithms affect the generation of a HIT: the public-key algorithm 460 of the HI and the used HIH. The two algorithms are encoded in the 461 bit presentation of the HIT. As the two communicating parties may 462 support different algorithms, [RFC7401] defines the minimum set for 463 interoperability. For further interoperability, the responder may 464 store its keys in DNS records, and thus the initiator may have to 465 couple destination HITs with appropriate source HITs according to 466 matching HIH. 468 In the HIP packets, the HITs identify the sender and recipient of a 469 packet. Consequently, a HIT should be unique in the whole IP 470 universe as long as it is being used. In the extremely rare case of 471 a single HIT mapping to more than one Host Identity, the Host 472 Identifiers (public keys) will make the final difference. If there 473 is more than one public key for a given node, the HIT acts as a hint 474 for the correct public key to use. 476 4.4. Local Scope Identifier (LSI) 478 An LSI is a 32-bit localized representation for a Host Identity. The 479 purpose of an LSI is to facilitate using Host Identities in existing 480 APIs for IPv4-based applications. Besides facilitating HIP-based 481 connectivity for legacy IPv4 applications, the LSIs are beneficial in 482 two other scenarios [RFC6538]. 484 In the first scenario, two IPv4-only applications are residing on two 485 separate hosts connected by IPv6-only network. With HIP-based 486 connectivity, the two applications are able to communicate despite of 487 the mismatch in the protocol families of the applications and the 488 underlying network. The reason is that the HIP layer translates the 489 LSIs originating from the upper layers into routable IPv6 locators 490 before delivering the packets on the wire. 492 The second scenario is the same as the first one, but with the 493 difference that one of the applications supports only IPv6. Now two 494 obstacles hinder the communication between the application: the 495 addressing families of the two applications differ, and the 496 application residing at the IPv4-only side is again unable to 497 communicate because of the mismatch between addressing families of 498 the application (IPv4) and network (IPv6). With HIP-based 499 connectivity for applications, this scenario works; the HIP layer can 500 choose whether to translate the locator of an incoming packet into an 501 LSI or HIT. 503 Effectively, LSIs improve IPv6 interoperability at the network layer 504 as described in the first scenario and at the application layer as 505 depicted in the second example. The interoperability mechanism 506 should not be used to avoid transition to IPv6; the authors firmly 507 believe in IPv6 adoption and encourage developers to port existing 508 IPv4-only applications to use IPv6. However, some proprietary, 509 closed-source, IPv4-only applications may never see the daylight of 510 IPv6, and the LSI mechanism is suitable for extending the lifetime of 511 such applications even in IPv6-only networks. 513 The main disadvantage of an LSI is its local scope. Applications may 514 violate layering principles and pass LSIs to each other in 515 application-layer protocols. As the LSIs are valid only in the 516 context of the local host, they may represent an entirely different 517 host when passed to another host. However, it should be emphasized 518 here that the LSI concept is effectively a host-based NAT and does 519 not introduce any more issues than the prevalent middlebox based NATs 520 for IPv4. In other words, the applications violating layering 521 principles are already broken by the NAT boxes that are ubiquitously 522 deployed. 524 4.5. Storing Host Identifiers in directories 526 The public Host Identifiers should be stored in DNS; the unpublished 527 Host Identifiers should not be stored anywhere (besides the 528 communicating hosts themselves). The (public) HI along with the 529 supported HIHs are stored in a new RR type. This RR type is defined 530 in HIP DNS Extension [I-D.ietf-hip-rfc5205-bis]. 532 Alternatively, or in addition to storing Host Identifiers in the DNS, 533 they may be stored in various other directories. For instance, a 534 directory based on the Lightweight Directory Access Protocol (LDAP) 535 or a Public Key Infrastructure (PKI) [I-D.ietf-hip-rfc6253-bis] may 536 be used. Alternatively, Distributed Hash Tables (DHTs) [RFC6537] 537 have successfully been utilized [RFC6538]. Such a practice may allow 538 them to be used for purposes other than pure host identification. 540 Some types of applications may cache and use Host Identifiers 541 directly, while others may indirectly discover them through symbolic 542 host name (such as FQDN) look up from a directory. Even though Host 543 Identities can have a substantially longer lifetime associated with 544 them than routable IP addresses, directories may be a better approach 545 to manage the lifespan of Host Identities. For example, an LDAP- 546 based directory or DHT can be used for locally published identities 547 whereas DNS can be more suitable for public advertisement. 549 5. New stack architecture 551 One way to characterize Host Identity is to compare the proposed new 552 architecture with the current one. Using the terminology from the 553 IRTF Name Space Research Group Report [nsrg-report] and, e.g., the 554 unpublished Internet-Draft Endpoints and Endpoint Names 555 [chiappa-endpoints], the IP addresses currently embody the dual role 556 of locators and end-point identifiers. That is, each IP address 557 names a topological location in the Internet, thereby acting as a 558 routing direction vector, or locator. At the same time, the IP 559 address names the physical network interface currently located at the 560 point-of-attachment, thereby acting as a end-point name. 562 In the HIP architecture, the end-point names and locators are 563 separated from each other. IP addresses continue to act as locators. 564 The Host Identifiers take the role of end-point identifiers. It is 565 important to understand that the end-point names based on Host 566 Identities are slightly different from interface names; a Host 567 Identity can be simultaneously reachable through several interfaces. 569 The difference between the bindings of the logical entities are 570 illustrated in Figure 1. Left side illustrates the current TCP/IP 571 architecture and right side the HIP-based architecture. 573 Transport ---- Socket Transport ------ Socket 574 association | association | 575 | | 576 | | 577 | | 578 End-point | End-point --- Host Identity 579 \ | | 580 \ | | 581 \ | | 582 \ | | 583 Location --- IP address Location --- IP address 585 Figure 1 587 Architecturally, HIP provides for a different binding of transport- 588 layer protocols. That is, the transport-layer associations, i.e., 589 TCP connections and UDP associations, are no longer bound to IP 590 addresses but rather to Host Identities. In practice, the Host 591 Identities are exposed as LSIs and HITs for legacy applications and 592 the transport layer to facilitate backward compatibility with 593 existing networking APIs and stacks. 595 5.1. On the multiplicity of identities 597 A host may have multiple identities both at the client and server 598 side. This raises some additional concerns that are addressed in 599 this section. 601 For security reasons, it may be a bad idea to duplicate the same Host 602 Identity on multiple hosts because the compromise of a single host 603 taints the identities of the other hosts. Management of machines 604 with identical Host Identities may also present other challenges and, 605 therefore, it is advisable to have a unique identity for each host. 607 Instead of duplicating identities, HIP opportunistic mode can be 608 employed, where the initiator leaves out the identifier of the 609 responder when initiating the key exchange and learns it upon the 610 completion of the exchange. The tradeoffs are related to lowered 611 security guarantees, but a benefit of the approach is to avoid 612 publishing of Host Identifiers in any directories [komu-leap]. The 613 approach could also be used for load balancing purposes at the HIP 614 layer because the identity of the responder can be decided 615 dynamically during the key exchange. Thus, the approach has the 616 potential to be used as a HIP-layer "anycast", either directly 617 between two hosts or indirectly through the rendezvous service 618 [komu-diss]. 620 At the client side, a host may have multiple Host Identities, for 621 instance, for privacy purposes. Another reason can be that the 622 person utilizing the host employs different identities for different 623 administrative domains as an extra security measure. If a HIP-aware 624 middlebox, such as a HIP-based firewall, is on the path between the 625 client and server, the user or the underlying system should carefully 626 choose the correct identity to avoid the firewall to unnecessarily 627 drop HIP-base connectivity [komu-diss]. 629 Similarly, a server may have multiple Host Identities. For instance, 630 a single web server may serve multiple different administrative 631 domains. Typically, the distinction is accomplished based on the DNS 632 name, but also the Host Identity could be used for this purpose. 633 However, a more compelling reason to employ multiple identities are 634 HIP-aware firewalls that are unable see the HTTP traffic inside the 635 encrypted IPsec tunnel. In such a case, each service could be 636 configured with a separate identity, thus allowing the firewall to 637 segregate the different services of the single web server from each 638 other [lindqvist-enterprise]. 640 6. Control plane 642 HIP decouples control and data plane from each other. Two end-hosts 643 initialize the control plane using a key exchange procedure called 644 the base exchange. The procedure can be assisted by new 645 infrastructural intermediaries called rendezvous or relay servers. 646 In the event of IP address changes, the end-hosts sustain control 647 plane connectivity with mobility and multihoming extensions. 648 Eventually, the end-hosts terminate the control plane and remove the 649 associated state. 651 6.1. Base exchange 653 The base exchange is a key exchange procedure that authenticates the 654 initiator and responder to each other using their public keys. 655 Typically, the initiator is the client-side host and the responder is 656 the server-side host. The roles are used by the state machine of a 657 HIP implementation, but discarded upon successful completion. 659 The exchange consists of four messages during which the hosts also 660 create symmetric keys to protect the control plane with Hash-based 661 message authentication codes (HMACs). The keys can be also used to 662 protect the data plane, and IPsec ESP [RFC7402] is typically used as 663 the data-plane protocol, albeit HIP can also accommodate others. 664 Both the control and data plane are terminated using a closing 665 procedure consisting of two messages. 667 In addition, the base exchange also includes a computational puzzle 668 [RFC7401] that the initiator must solve. The responder chooses the 669 difficulty of the puzzle which permits the responder to delay new 670 incoming initiators according to local policies, for instance, when 671 the responder is under heavy load. The puzzle can offer some 672 resiliency against DoS attacks because the design of the puzzle 673 mechanism allows the responder to remain stateless until the very end 674 of the base exchange [aura-dos]. HIP puzzles have also been studied 675 under steady-state DDoS attacks [beal-dos], on multiple adversary 676 models with varying puzzle difficulties [tritilanunt-dos] and with 677 ephemeral Host Identities [komu-mitigation]. 679 6.2. End-host mobility and multi-homing 681 HIP decouples the transport from the internetworking layer, and binds 682 the transport associations to the Host Identities (actually through 683 either the HIT or LSI). After the initial key exchange, the HIP 684 layer maintains transport-layer connectivity and data flows using its 685 mobility [I-D.ietf-hip-rfc5206-bis] and multihoming 686 [I-D.ietf-hip-multihoming] extensions. Consequently, HIP can provide 687 for a degree of internetworking mobility and multi-homing at a low 688 infrastructure cost. HIP mobility includes IP address changes (via 689 any method) to either party. Thus, a system is considered mobile if 690 its IP address can change dynamically for any reason like PPP, DHCP, 691 IPv6 prefix reassignments, or a NAT device remapping its translation. 692 Likewise, a system is considered multi-homed if it has more than one 693 globally routable IP address at the same time. HIP links IP 694 addresses together, when multiple IP addresses correspond to the same 695 Host Identity. If one address becomes unusable, or a more preferred 696 address becomes available, existing transport associations can easily 697 be moved to another address. 699 When a node moves while communication is already on-going, address 700 changes are rather straightforward. The peer of the mobile node can 701 just accept a HIP or an integrity protected ESP packet from any 702 address and ignore the source address. However, as discussed in 703 Section 12.2 below, a mobile node must send a HIP UPDATE packet to 704 inform the peer of the new address(es), and the peer must verify that 705 the mobile node is reachable through these addresses. This is 706 especially helpful for those situations where the peer node is 707 sending data periodically to the mobile node (that is, re-starting a 708 connection after the initial connection). 710 6.3. Rendezvous mechanism 712 Establishing a contact to a mobile, moving node is slightly more 713 involved. In order to start the HIP exchange, the initiator node has 714 to know how to reach the mobile node. For instance, the mobile node 715 can employ Dynamic DNS [RFC2136] to update its reachability 716 information in the DNS. To avoid the dependency to DNS, HIP provides 717 its own HIP-specific alternative: the HIP rendezvous mechanism as 718 defined in HIP Rendezvous specifications [I-D.ietf-hip-rfc5204-bis]. 720 Using the HIP rendezvous extensions, the mobile node keeps the 721 rendezvous infrastructure continuously updated with its current IP 722 address(es). The mobile nodes trusts the rendezvous mechanism in 723 order to properly maintain their HIT and IP address mappings. 725 The rendezvous mechanism is especially useful in scenarios where both 726 of the nodes are expected to change their address at the same time. 727 In such a case, the HIP UPDATE packets will cross each other in the 728 network and never reach the peer node. 730 6.4. Relay mechanism 732 The HIP relay mechanism [I-D.ietf-hip-native-nat-traversal] is an 733 alternative to the HIP rendezvous mechanism. The HIP relay mechanism 734 is more suitable for IPv4 networks with NATs because a HIP relay can 735 forward all control and data plane communications in order to 736 guarantee successful NAT traversal. 738 6.5. Termination of the control plane 740 The control plane between two hosts is terminated using a secure two 741 message exchange as specified in base exchange specification 742 [RFC7401]. The related state (i.e. host associations) should be 743 removed upon successful termination. 745 7. Data plane 746 The encapsulation format for the data plane used for carrying the 747 application-layer traffic can be dynamically negotiated during the 748 key exchange. For instance, HICCUPS extensions [RFC6078] define one 749 way to transport application-layer datagrams directly over the HIP 750 control plane, protected by asymmetric key cryptography. Also, S-RTP 751 has been considered as the data encapsulation protocol [hip-srtp]. 752 However, the most widely implemented method is the Encapsulated 753 Security Payload (ESP) [RFC7402] that is protected by symmetric keys 754 derived during the key exchange. ESP Security Associations (SAs) 755 offer both confidentiality and integrity protection, of which the 756 former can be disabled during the key exchange. In the future, other 757 ways of transporting application-layer data may be defined. 759 The ESP SAs are established and terminated between the initiator and 760 the responder hosts. Usually, the hosts create at least two SAs, one 761 in each direction (initiator-to-responder SA and responder-to- 762 initiator SA). If the IP addresses of either host changes, the HIP 763 mobility extensions can be used to re-negotiate the corresponding 764 SAs. 766 On the wire, the difference in the use of identifiers between the HIP 767 control and data plane is that the HITs are included in all control 768 packets, but not in the data plane when ESP is employed. Instead, 769 the ESP employs SPI numbers that act as compressed HITs. Any HIP- 770 aware middlebox (for instance, a HIP-aware firewall) interested in 771 the ESP based data plane should keep track between the control and 772 data plane identifiers in order to associate them with each other. 774 Since HIP does not negotiate any SA lifetimes, all lifetimes are 775 subject to local policy. The only lifetimes a HIP implementation 776 must support are sequence number rollover (for replay protection), 777 and SA timeout. An SA times out if no packets are received using 778 that SA. Implementations may support lifetimes for the various ESP 779 transforms and other data-plane protocols. 781 8. HIP and NATs 783 Passing packets between different IP addressing realms requires 784 changing IP addresses in the packet header. This may occur, for 785 example, when a packet is passed between the public Internet and a 786 private address space, or between IPv4 and IPv6 networks. The 787 address translation is usually implemented as Network Address 788 Translation (NAT) [RFC3022] or NAT Protocol translation (NAT-PT) 789 [RFC2766]. 791 In a network environment where identification is based on the IP 792 addresses, identifying the communicating nodes is difficult when NATs 793 are employed because private address spaces are overlapping. In 794 other words, two hosts cannot be distinguished from each other solely 795 based on their IP address. With HIP, the transport-layer end-points 796 (i.e. applications) are bound to unique Host Identities rather than 797 overlapping private addresses. This allows two end-points to 798 distinguish one other even when they are located in different private 799 address realms. Thus, the IP addresses are used only for routing 800 purposes and can be changed freely by NATs when a packet between two 801 HIP capable hosts traverses through multiple private address realms. 803 NAT traversal extensions for HIP [I-D.ietf-hip-native-nat-traversal] 804 can be used to realize the actual end-to-end connectivity through NAT 805 devices. To support basic backward compatibility with legacy NATs, 806 the extensions encapsulate both HIP control and data plane in UDP. 807 The extensions define mechanisms for forwarding the two planes 808 through an intermediary host called HIP relay and procedures to 809 establish direct end-to-end connectivity by penetrating NATs. 810 Besides this "native" NAT traversal mode for HIP, other NAT traversal 811 mechanisms have been successfully utilized, such as Teredo 812 [varjonen-split]. 814 Besides legacy NATs, a HIP-aware NAT has been designed and 815 implemented [ylitalo-spinat]. For a HIP-based flow, a HIP-aware NAT 816 or NAT-PT system tracks the mapping of HITs, and the corresponding 817 ESP SPIs, to an IP address. The NAT system has to learn mappings 818 both from HITs and from SPIs to IP addresses. Many HITs (and SPIs) 819 can map to a single IP address on a NAT, simplifying connections on 820 address poor NAT interfaces. The NAT can gain much of its knowledge 821 from the HIP packets themselves; however, some NAT configuration may 822 be necessary. 824 8.1. HIP and Upper-layer checksums 826 There is no way for a host to know if any of the IP addresses in an 827 IP header are the addresses used to calculate the TCP checksum. That 828 is, it is not feasible to calculate the TCP checksum using the actual 829 IP addresses in the pseudo header; the addresses received in the 830 incoming packet are not necessarily the same as they were on the 831 sending host. Furthermore, it is not possible to recompute the 832 upper-layer checksums in the NAT/NAT-PT system, since the traffic is 833 ESP protected. Consequently, the TCP and UDP checksums are 834 calculated using the HITs in the place of the IP addresses in the 835 pseudo header. Furthermore, only the IPv6 pseudo header format is 836 used. This provides for IPv4 / IPv6 protocol translation. 838 9. Multicast 840 A number of studies investigating HIP-based multicast have been 841 published (including [shields-hip], [xueyong-hip], [xueyong-hip], 843 [amir-hip], [kovacshazi-host] and [xueyong-secure]). In particular, 844 so-called Bloom filters, that allow compressing of multiple labels 845 into small data structures, may be a promising way forward 846 [sarela-bloom]. However, the different schemes have not been adopted 847 by the HIP working group (nor the HIP research group in IRTF), so the 848 details are not further elaborated here. 850 10. HIP policies 852 There are a number of variables that influence the HIP exchange that 853 each host must support. All HIP implementations should support at 854 least 2 HIs, one to publish in DNS or similar directory service and 855 an unpublished one for anonymous usage. Although unpublished HIs 856 will be rarely used as responder HIs, they are likely to be common 857 for initiators. Support for multiple HIs is recommended. This 858 provides new challenges for systems or users to decide which type of 859 HI to expose when they start a new session. 861 Opportunistic mode (where the initiator starts a HIP exchange without 862 prior knowledge of the responder's HI) presents a security tradeoff. 863 At the expense of being subject to MITM attacks, the opportunistic 864 mode allows the initiator to learn the identity of the responder 865 during communication rather than from an external directory. 866 Opportunistic mode can be used for registration to HIP-based services 867 [I-D.ietf-hip-rfc5203-bis] (i.e. utilized by HIP for its own internal 868 purposes) or by the application layer [komu-leap]. For security 869 reasons, especially the latter requires some involvement from the 870 user to accept the identity of the responder similar to how SSH 871 prompts the user when connecting to a server for the first time 872 [pham-leap]. In practice, this can be realized in end-host based 873 firewalls in the case of legacy applications [karvonen-usable] or 874 with native APIs for HIP APIs [RFC6317] in the case of HIP-aware 875 applications. 877 Many initiators would want to use a different HI for different 878 responders. The implementations should provide for a policy mapping 879 of initiator HITs to responder HITs. This policy should also include 880 preferred transforms and local lifetimes. 882 Responders would need a similar policy, describing the hosts allowed 883 to participate in HIP exchanges, and the preferred transforms and 884 local lifetimes. 886 11. Design considerations 888 11.1. Benefits of HIP 890 In the beginning, the network layer protocol (i.e., IP) had the 891 following four "classic" invariants: 893 1. Non-mutable: The address sent is the address received. 895 2. Non-mobile: The address doesn't change during the course of an 896 "association". 898 3. Reversible: A return header can always be formed by reversing the 899 source and destination addresses. 901 4. Omniscient: Each host knows what address a partner host can use 902 to send packets to it. 904 Actually, the fourth can be inferred from 1 and 3, but it is worth 905 mentioning explicitly for reasons that will be obvious soon if not 906 already. 908 In the current "post-classic" world, we are intentionally trying to 909 get rid of the second invariant (both for mobility and for multi- 910 homing), and we have been forced to give up the first and the fourth. 911 Realm Specific IP [RFC3102] is an attempt to reinstate the fourth 912 invariant without the first invariant. IPv6 is an attempt to 913 reinstate the first invariant. 915 Few client-side systems on the Internet have DNS names that are 916 meaningful. That is, if they have a Fully Qualified Domain Name 917 (FQDN), that name typically belongs to a NAT device or a dial-up 918 server, and does not really identify the system itself but its 919 current connectivity. FQDNs (and their extensions as email names) 920 are application-layer names; more frequently naming services than 921 particular systems. This is why many systems on the Internet are not 922 registered in the DNS; they do not have services of interest to other 923 Internet hosts. 925 DNS names are references to IP addresses. This only demonstrates the 926 interrelationship of the networking and application layers. DNS, as 927 the Internet's only deployed and distributed database, is also the 928 repository of other namespaces, due in part to DNSSEC and application 929 specific key records. Although each namespace can be stretched (IP 930 with v6, DNS with KEY records), neither can adequately provide for 931 host authentication or act as a separation between internetworking 932 and transport layers. 934 The Host Identity (HI) namespace fills an important gap between the 935 IP and DNS namespaces. An interesting thing about the HI is that it 936 actually allows a host to give up all but the 3rd network-layer 937 invariant. That is to say, as long as the source and destination 938 addresses in the network-layer protocol are reversible, HIP takes 939 care of host identification, and reversibility allows a local host to 940 receive a packet back from a remote host. The address changes 941 occurring during NAT transit (non-mutable) or host movement (non- 942 omniscient or non-mobile) can be managed by the HIP layer. 944 With the exception of High-Performance Computing applications, the 945 Sockets API is the most common way to develop network applications. 946 Applications use the Sockets API either directly or indirectly 947 through some libraries or frameworks. However, the Sockets API is 948 based on the assumption of static IP addresses, and DNS with its 949 lifetime values was invented at later stages during the evolution of 950 the Internet. Hence, the Sockets API does not deal with the lifetime 951 of addresses [RFC6250]. As the majority of the end-user equipment is 952 mobile today, their addresses are effectively ephemeral, but the 953 Sockets API still gives a fallacious illusion of persistent IP 954 addresses to the unwary developer. HIP can be used to solidify this 955 illusion because HIP provides persistent surrogate addresses to the 956 application layer in the form of LSIs and HITs. 958 The persistent identifiers as provided by HIP are useful in multiple 959 scenarios (see, e.g., [ylitalo-diss] or [komu-diss], for a more 960 elaborate discussion): 962 o When a mobile host moves physically between two different WLAN 963 networks and obtains a new address, an application using the 964 identifiers remains isolated regardless of the topology changes 965 while the underlying HIP layer re-establishes connectivity (i.e. a 966 horizontal handoff). 968 o Similarly, the application utilizing the identifiers remains again 969 unaware of the topological changes when the underlying host 970 equipped with WLAN and cellular network interfaces switches 971 between the two different access technologies (i.e. a vertical 972 handoff). 974 o Even when hosts are located in private address realms, 975 applications can uniquely distinguish different hosts from each 976 other based on their identifiers. In other words, it can be 977 stated that HIP improves Internet transparency for the application 978 layer [komu-diss]. 980 o Site renumbering events for services can occur due to corporate 981 mergers or acquisitions, or by changes in Internet Service 982 Provider. They can involve changing the entire network prefix of 983 an organization, which is problematic due to hard-coded addresses 984 in service configuration files or cached IP addresses at the 985 client side [RFC5887]. Considering such human errors, a site 986 employing location-independent identifiers as promoted by HIP may 987 experience less problems while renumbering their network. 989 o More agile IPv6 interoperability can be achieved, as discussed in 990 Section 4.4. IPv6-based applications can communicate using HITs 991 with IPv4-based applications that are using LSIs. Additionally, 992 the underlying network type (IPv4 or IPv6) becomes independent of 993 the addressing family of the application. 995 o HITs (or LSIs) can be used in IP-based access control lists as a 996 more secure replacement for IPv6 addresses. Besides security, HIT 997 based access control has two other benefits. First, the use of 998 HITs can potentially halve the size of access control lists 999 because separate rules for IPv4 are not needed [komu-diss]. 1000 Second, HIT-based configuration rules in HIP-aware middleboxes 1001 remain static and independent of topology changes, thus 1002 simplifying administrative efforts particularly for mobile 1003 environments. For instance, the benefits of HIT based access 1004 control have been harnessed in the case of HIP-aware firewalls, 1005 but can be utilized directly at the end-hosts as well [RFC6538]. 1007 While some of these benefits could be and have been redundantly 1008 implemented by individual applications, providing such generic 1009 functionality at the lower layers is useful because it reduces 1010 software development effort and networking software bugs (as the 1011 layer is tested with multiple applications). It also allows the 1012 developer to focus on building the application itself rather than 1013 delving into the intricacies of mobile networking, thus facilitating 1014 separation of concerns. 1016 HIP could also be realized by combining a number of different 1017 protocols, but the complexity of the resulting software may become 1018 substantially larger, and the interaction between multiple possibly 1019 layered protocols may have adverse effects on latency and throughput. 1020 It is also worth noting that virtually nothing prevents realizing the 1021 HIP architecture, for instance, as an application-layer library, 1022 which has been actually implemented in the past [xin-hip-lib]. 1023 However, the tradeoff in moving the HIP layer to the application 1024 layer is that legacy applications may not be supported. 1026 11.2. Drawbacks of HIP 1028 In computer science, many problems can be solved with an extra layer 1029 of indirection. However, the indirection always involves some costs 1030 as there is no such a thing as "free lunch". In the case of HIP, the 1031 main costs could be stated as follows: 1033 o In general, a new layer and a new namespace always involve some 1034 initial effort in terms of implementation, deployment and 1035 maintenance. Some education of developers and administrators may 1036 also be needed. However, the HIP community at the IETF has spent 1037 years in experimenting, exploring, testing, documenting and 1038 implementing HIP to ease the adoption costs. 1040 o HIP decouples identifier and locator roles of IP addresses. 1041 Consequently, a mapping mechanism is needed to associate them 1042 together. A failure to map a HIT to its corresponding locator may 1043 result in failed connectivity because a HIT is "flat" by its 1044 nature and cannot be looked up from the hierarchically organized 1045 DNS. HITs are flat by design due to a security tradeoff. The 1046 more bits are allocated for the hash in the HIT, the less likely 1047 there will be (malicious) collisions. 1049 o From performance viewpoint, HIP control and data plane processing 1050 introduces some overhead in terms of throughput and latency as 1051 elaborated below. 1053 The key exchange introduces some extra latency (two round trips) in 1054 the initial transport layer connection establishment between two 1055 hosts. With TCP, additional delay occurs if the underlying network 1056 stack implementation drops the triggering SYN packet during the key 1057 exchange. The same cost may also occur during HIP handoff 1058 procedures. However, subsequent TCP sessions using the same HIP 1059 association will not bear this cost (within the key lifetime). Both 1060 the key exchange and handoff penalties can be minimized by caching 1061 TCP packets. The latter case can further be optimized with TCP user 1062 timeout extensions [RFC5482] as described in further detail by 1063 Schuetz et al [schuetz-intermittent]. 1065 The most CPU-intensive operations involve the use of the asymmetric 1066 keys and Diffie-Hellman key derivation at the control plane, but this 1067 occurs only during the key exchange, its maintenance (handoffs, 1068 refreshing of key material) and tear down procedures of HIP 1069 associations. The data plane is typically implemented with ESP 1070 because it has a smaller overhead due to symmetric key encryption. 1071 Naturally, even ESP involves some overhead in terms of latency 1072 (processing costs) and throughput (tunneling) (see e.g. 1073 [ylitalo-diss] for a performance evaluation). 1075 11.3. Deployment and adoption considerations 1076 This section describes some deployment and adoption considerations 1077 related to HIP from a technical perspective. 1079 11.3.1. Deployment analysis 1081 HIP has commercially been utilized at Boeing airplane factory for 1082 their internal purposes [paine-hip]. It has been included in a 1083 security product called Tofino to support layer-two Virtual Private 1084 Networks [henderson-vpls] to facilitate, e.g, supervisory control and 1085 data acquisition (SCADA) security. However, HIP has not been a "wild 1086 success" [RFC5218] in the Internet as argued by Levae et al 1087 [leva-barriers]. Here, we briefly highlight some of their findings 1088 based on interviews with 19 experts from the industry and academia. 1090 From a marketing perspective, the demand for HIP has been low and 1091 substitute technologies have been favored. Another identified reason 1092 has been that some technical misconceptions related to the early 1093 stages of HIP specifications still persist. Two identified 1094 misconceptions are that HIP does not support NAT traversal, and that 1095 HIP must be implemented in the OS kernel. Both of these claims are 1096 untrue; HIP does have NAT traversal extensions 1097 [I-D.ietf-hip-native-nat-traversal], and kernel modifications can be 1098 avoided with modern operating systems by diverting packets for 1099 userspace processing. 1101 The analysis by Levae et al clarifies infrastructural requirements 1102 for HIP. In a minimal set up, a client and server machine have to 1103 run HIP software. However, to avoid manual configurations, usually 1104 DNS records for HIP are set up. For instance, the popular DNS server 1105 software Bind9 does not require any changes to accommodate DNS 1106 records for HIP because they can be supported in binary format in its 1107 configuration files [RFC6538]. HIP rendezvous servers and firewalls 1108 are optional. No changes are required to network address points, 1109 NATs, edge routers or core networks. HIP may require holes in legacy 1110 firewalls. 1112 The analysis also clarifies the requirements for the host components 1113 that consist of three parts. First, a HIP control plane component is 1114 required, typically implemented as a userspace daemon. Second, a 1115 data plane component is needed. Most HIP implementations utilize the 1116 so called BEET mode of ESP that has been available since Linux kernel 1117 2.6.27, but is included also as a userspace component in a few of the 1118 implementations. Third, HIP systems usually provide a DNS proxy for 1119 the local host that translates HIP DNS records to LSIs and HITs, and 1120 communicates the corresponding locators to HIP userspace daemon. 1121 While the third component is not mandatory, it is very useful for 1122 avoiding manual configurations. The three components are further 1123 described in the HIP experiment report [RFC6538]. 1125 Based on the interviews, Levae et al suggest further directions to 1126 facilitate HIP deployment. Transitioning the HIP specifications to 1127 the standards track may help, but other measures could be taken. As 1128 a more radical measure, the authors suggest to implement HIP as a 1129 purely application-layer library [xin-hip-lib] or other kind of 1130 middleware. On the other hand, more conservative measures include 1131 focusing on private deployments controlled by a single stakeholder. 1132 As a more concrete example of such a scenario, HIP could be used by a 1133 single service provider to facilitate secure connectivity between its 1134 servers [komu-cloud]. 1136 11.3.2. HIP in 802.15.4 networks 1138 The IEEE 802 standards have been defining MAC layered security. Many 1139 of these standards use EAP [RFC3748] as a Key Management System (KMS) 1140 transport, but some like IEEE 802.15.4 [IEEE.802-15-4.2011] leave the 1141 KMS and its transport as "Out of Scope". 1143 HIP is well suited as a KMS in these environments: 1145 o HIP is independent of IP addressing and can be directly 1146 transported over any network protocol. 1148 o Master Keys in 802 protocols are commonly pair-based with group 1149 keys transported from the group controller using pair-wise keys. 1151 o AdHoc 802 networks can be better served by a peer-to-peer KMS than 1152 the EAP client/server model. 1154 o Some devices are very memory constrained and a common KMS for both 1155 MAC and IP security represents a considerable code savings. 1157 11.4. Answers to NSRG questions 1159 The IRTF Name Space Research Group has posed a number of evaluating 1160 questions in their report [nsrg-report]. In this section, we provide 1161 answers to these questions. 1163 1. How would a stack name improve the overall functionality of the 1164 Internet? 1165 HIP decouples the internetworking layer from the transport 1166 layer, allowing each to evolve separately. The decoupling 1167 makes end-host mobility and multi-homing easier, also across 1168 IPv4 and IPv6 networks. HIs make network renumbering easier, 1169 and they also make process migration and clustered servers 1170 easier to implement. Furthermore, being cryptographic in 1171 nature, they provide the basis for solving the security 1172 problems related to end-host mobility and multi-homing. 1174 2. What does a stack name look like? 1176 A HI is a cryptographic public key. However, instead of using 1177 the keys directly, most protocols use a fixed size hash of the 1178 public key. 1180 3. What is its lifetime? 1182 HIP provides both stable and temporary Host Identifiers. 1183 Stable HIs are typically long lived, with a lifetime of years 1184 or more. The lifetime of temporary HIs depends on how long 1185 the upper-layer connections and applications need them, and 1186 can range from a few seconds to years. 1188 4. Where does it live in the stack? 1190 The HIs live between the transport and internetworking layers. 1192 5. How is it used on the end points? 1194 The Host Identifiers may be used directly or indirectly (in 1195 the form of HITs or LSIs) by applications when they access 1196 network services. Additionally, the Host Identifiers, as 1197 public keys, are used in the built in key agreement protocol, 1198 called the HIP base exchange, to authenticate the hosts to 1199 each other. 1201 6. What administrative infrastructure is needed to support it? 1203 In some environments, it is possible to use HIP 1204 opportunistically, without any infrastructure. However, to 1205 gain full benefit from HIP, the HIs must be stored in the DNS 1206 or a PKI, and a new rendezvous mechanism is needed 1207 [I-D.ietf-hip-rfc5205-bis]. 1209 7. If we add an additional layer would it make the address list in 1210 SCTP unnecessary? 1212 Yes 1214 8. What additional security benefits would a new naming scheme 1215 offer? 1217 HIP reduces dependency on IP addresses, making the so called 1218 address ownership [Nik2001] problems easier to solve. In 1219 practice, HIP provides security for end-host mobility and 1220 multi-homing. Furthermore, since HIP Host Identifiers are 1221 public keys, standard public key certificate infrastructures 1222 can be applied on the top of HIP. 1224 9. What would the resolution mechanisms be, or what characteristics 1225 of a resolution mechanisms would be required? 1227 For most purposes, an approach where DNS names are resolved 1228 simultaneously to HIs and IP addresses is sufficient. 1229 However, if it becomes necessary to resolve HIs into IP 1230 addresses or back to DNS names, a flat resolution 1231 infrastructure is needed. Such an infrastructure could be 1232 based on the ideas of Distributed Hash Tables, but would 1233 require significant new development and deployment. 1235 12. Security considerations 1237 This section includes discussion on some issues and solutions related 1238 to security in the HIP architecture. 1240 12.1. MiTM Attacks 1242 HIP takes advantage of the new Host Identity paradigm to provide 1243 secure authentication of hosts and to provide a fast key exchange for 1244 ESP. HIP also attempts to limit the exposure of the host to various 1245 denial-of-service (DoS) and man-in-the-middle (MitM) attacks. In so 1246 doing, HIP itself is subject to its own DoS and MitM attacks that 1247 potentially could be more damaging to a host's ability to conduct 1248 business as usual. 1250 Resource exhausting denial-of-service attacks take advantage of the 1251 cost of setting up a state for a protocol on the responder compared 1252 to the 'cheapness' on the initiator. HIP allows a responder to 1253 increase the cost of the start of state on the initiator and makes an 1254 effort to reduce the cost to the responder. This is done by having 1255 the responder start the authenticated Diffie-Hellman exchange instead 1256 of the initiator, making the HIP base exchange 4 packets long. The 1257 first packet sent by the responder can be prebuilt to further 1258 mitigate the costs. This packet also includes a computational puzzle 1259 that can optionally be used to further delay the initiator, for 1260 instance, when the responder is overloaded. The details are 1261 explained in the base exchange specification [RFC7401]. 1263 Man-in-the-middle (MitM) attacks are difficult to defend against, 1264 without third-party authentication. A skillful MitM could easily 1265 handle all parts of the HIP base exchange, but HIP indirectly 1266 provides the following protection from a MitM attack. If the 1267 responder's HI is retrieved from a signed DNS zone or securely 1268 obtained by some other means, the initiator can use this to 1269 authenticate the signed HIP packets. Likewise, if the initiator's HI 1270 is in a secure DNS zone, the responder can retrieve it and validate 1271 the signed HIP packets. However, since an initiator may choose to 1272 use an unpublished HI, it knowingly risks a MitM attack. The 1273 responder may choose not to accept a HIP exchange with an initiator 1274 using an unknown HI. 1276 Other types of MitM attacks against HIP can be mounted using ICMP 1277 messages that can be used to signal about problems. As a overall 1278 guideline, the ICMP messages should be considered as unreliable 1279 "hints" and should be acted upon only after timeouts. The exact 1280 attack scenarios and countermeasures are described in full detail the 1281 base exchange specification [RFC7401]. 1283 The need to support multiple hashes for generating the HIT from the 1284 HI affords the MitM to mount a potentially powerful downgrade attack 1285 due to the a-priori need of the HIT in the HIP base exchange. The 1286 base exchange has been augmented to deal with such an attack by 1287 restarting on detecting the attack. At worst this would only lead to 1288 a situation in which the base exchange would never finish (or would 1289 be aborted after some retries). As a drawback, this leads to an 1290 6-way base exchange which may seem bad at first. However, since this 1291 only occurs in an attack scenario and since the attack can be handled 1292 (so it is not interesting to mount anymore), we assume the subsequent 1293 messages do not represent a security threat. Since the MitM cannot 1294 be successful with a downgrade attack, these sorts of attacks will 1295 only occur as 'nuisance' attacks. So, the base exchange would still 1296 be usually just four packets even though implementations must be 1297 prepared to protect themselves against the downgrade attack. 1299 In HIP, the Security Association for ESP is indexed by the SPI; the 1300 source address is always ignored, and the destination address may be 1301 ignored as well. Therefore, HIP-enabled Encapsulated Security 1302 Payload (ESP) is IP address independent. This might seem to make 1303 attacking easier, but ESP with replay protection is already as well 1304 protected as possible, and the removal of the IP address as a check 1305 should not increase the exposure of ESP to DoS attacks. 1307 12.2. Protection against flooding attacks 1309 Although the idea of informing about address changes by simply 1310 sending packets with a new source address appears appealing, it is 1311 not secure enough. That is, even if HIP does not rely on the source 1312 address for anything (once the base exchange has been completed), it 1313 appears to be necessary to check a mobile node's reachability at the 1314 new address before actually sending any larger amounts of traffic to 1315 the new address. 1317 Blindly accepting new addresses would potentially lead to flooding 1318 Denial-of-Service attacks against third parties [RFC4225]. In a 1319 distributed flooding attack an attacker opens high volume HIP 1320 connections with a large number of hosts (using unpublished HIs), and 1321 then claims to all of these hosts that it has moved to a target 1322 node's IP address. If the peer hosts were to simply accept the move, 1323 the result would be a packet flood to the target node's address. To 1324 prevent this type of attack, HIP mobility extensions include a return 1325 routability check procedure where the reachability of a node is 1326 separately checked at each address before using the address for 1327 larger amounts of traffic. 1329 A credit-based authorization approach for host mobility with the Host 1330 Identity Protocol [I-D.ietf-hip-rfc5206-bis] can be used between 1331 hosts for sending data prior to completing the address tests. 1332 Otherwise, if HIP is used between two hosts that fully trust each 1333 other, the hosts may optionally decide to skip the address tests. 1334 However, such performance optimization must be restricted to peers 1335 that are known to be trustworthy and capable of protecting themselves 1336 from malicious software. 1338 12.3. HITs used in ACLs 1340 At end-hosts, HITs can be used in IP-based access control lists at 1341 the application and network layers. At middleboxes, HIP-aware 1342 firewalls [lindqvist-enterprise] can use HITs or public keys to 1343 control both ingress and egress access to networks or individual 1344 hosts, even in the presence of mobile devices because the HITs and 1345 public keys are topologically independent. As discussed earlier in 1346 Section 7, once a HIP session has been established, the SPI value in 1347 an ESP packet may be used as an index, indicating the HITs. In 1348 practice, firewalls can inspect HIP packets to learn of the bindings 1349 between HITs, SPI values, and IP addresses. They can even explicitly 1350 control ESP usage, dynamically opening ESP only for specific SPI 1351 values and IP addresses. The signatures in HIP packets allow a 1352 capable firewall to ensure that the HIP exchange is indeed occurring 1353 between two known hosts. This may increase firewall security. 1355 A potential drawback of HITs in ACLs is their 'flatness' means they 1356 cannot be aggregated, and this could potentially result in larger 1357 table searches in HIP-aware firewalls. A way to optimize this could 1358 be to utilize Bloom filters for grouping of HITs [sarela-bloom]. 1360 However, it should be noted that it is also easier to exclude 1361 individual, misbehaving hosts out when the firewall rules concern 1362 individual HITs rather than groups. 1364 There has been considerable bad experience with distributed ACLs that 1365 contain public key related material, for example, with SSH. If the 1366 owner of a key needs to revoke it for any reason, the task of finding 1367 all locations where the key is held in an ACL may be impossible. If 1368 the reason for the revocation is due to private key theft, this could 1369 be a serious issue. 1371 A host can keep track of all of its partners that might use its HIT 1372 in an ACL by logging all remote HITs. It should only be necessary to 1373 log responder hosts. With this information, the host can notify the 1374 various hosts about the change to the HIT. There have been attempts 1375 to develop a secure method to issue the HIT revocation notice 1376 [zhang-revocation]. 1378 Some of the HIP-aware middleboxes, such as firewalls 1379 [lindqvist-enterprise] or NATs [ylitalo-spinat], may observe the on- 1380 path traffic passively. Such middleboxes are transparent by their 1381 nature and may not get a notification when a host moves to a 1382 different network. Thus, such middleboxes should maintain soft state 1383 and timeout when the control and data plane between two HIP end-hosts 1384 has been idle too long. Correspondingly, the two end-hosts may send 1385 periodically keepalives, such as UPDATE packets or ICMP messages 1386 inside the ESP tunnel, to sustain state at the on-path middleboxes. 1388 One general limitation related to end-to-end encryption is that 1389 middleboxes may not be able to participate to the protection of data 1390 flows. While the issue may affect also other protocols, Heer at al 1391 [heer-end-host] have analyzed the problem in the context of HIP. 1392 More specifically, when ESP is used as the data-plane protocol for 1393 HIP, the association between the control and data plane is weak and 1394 can be exploited under certain assumptions. In the scenario, the 1395 attacker has already gained access to the target network protected by 1396 a HIP-aware firewall, but wants to circumvent the HIP-based firewall. 1397 To achieve this, the attacker passively observes a base exchange 1398 between two HIP hosts and later replays it. This way, the attacker 1399 manages to penetrate the firewall and can use a fake ESP tunnel to 1400 transport its own data. This is possible because the firewall cannot 1401 distinguish when the ESP tunnel is valid. As a solution, HIP-aware 1402 middleboxes may participate to the control plane interaction by 1403 adding random nonce parameters to the control traffic, which the end- 1404 hosts have to sign to guarantee the freshness of the control traffic 1405 [heer-midauth]. As an alternative, extensions for transporting data 1406 plane directly over the control plane can be used [RFC6078]. 1408 12.4. Alternative HI considerations 1410 The definition of the Host Identifier states that the HI need not be 1411 a public key. It implies that the HI could be any value; for example 1412 a FQDN. This document does not describe how to support such a non- 1413 cryptographic HI, but examples of such protocol variants do exist 1414 ([urien-rfid], [urien-rfid-draft]). A non-cryptographic HI would 1415 still offer the services of the HIT or LSI for NAT traversal. It 1416 would be possible to carry HITs in HIP packets that had neither 1417 privacy nor authentication. Such schemes may be employed for 1418 resource constrained devices, such as small sensors operating on 1419 battery power, but are not further analyzed here. 1421 If it is desirable to use HIP in a low security situation where 1422 public key computations are considered expensive, HIP can be used 1423 with very short Diffie-Hellman and Host Identity keys. Such use 1424 makes the participating hosts vulnerable to MitM and connection 1425 hijacking attacks. However, it does not cause flooding dangers, 1426 since the address check mechanism relies on the routing system and 1427 not on cryptographic strength. 1429 13. IANA considerations 1431 This document has no actions for IANA. 1433 14. Acknowledgments 1435 For the people historically involved in the early stages of HIP, see 1436 the Acknowledgments section in the Host Identity Protocol 1437 specification. 1439 During the later stages of this document, when the editing baton was 1440 transferred to Pekka Nikander, the comments from the early 1441 implementers and others, including Jari Arkko, Tom Henderson, Petri 1442 Jokela, Miika Komu, Mika Kousa, Andrew McGregor, Jan Melen, Tim 1443 Shepard, Jukka Ylitalo, Sasu Tarkoma, and Jorma Wall, were 1444 invaluable. Also, the comments from Lars Eggert, Spencer Dawkins and 1445 Dave Crocker were also useful. 1447 The authors want to express their special thanks to Tom Henderson, 1448 who took the burden of editing the document in response to IESG 1449 comments at the time when both of the authors were busy doing other 1450 things. Without his perseverance original document might have never 1451 made it as RFC4423. 1453 This main effort to update and move HIP forward within the IETF 1454 process owes its impetuous to a number of HIP development teams. The 1455 authors are grateful for Boeing, Helsinki Institute for Information 1456 Technology (HIIT), NomadicLab of Ericsson, and the three 1457 universities: RWTH Aachen, Aalto and University of Helsinki, for 1458 their efforts. Without their collective efforts HIP would have 1459 withered as on the IETF vine as a nice concept. 1461 Thanks also for Suvi Koskinen for her help with proofreading and with 1462 the reference jungle. 1464 15. Changes from RFC 4423 1466 In a nutshell, the changes from RFC 4423 [RFC4423] are mostly 1467 editorial, including clarifications on topics described in a 1468 difficult way and omitting some of the non-architectural 1469 (implementation) details that are already described in other 1470 documents. A number of missing references to the literature were 1471 also added. New topics include the drawbacks of HIP, discussion on 1472 802.15.4 and MAC security, deployment considerations and description 1473 of the base exchange. 1475 16. References 1477 16.1. Normative References 1479 [I-D.ietf-hip-multihoming] 1480 Henderson, T., Vogt, C., and J. Arkko, "Host Multihoming 1481 with the Host Identity Protocol", draft-ietf-hip- 1482 multihoming-05 (work in progress), January 2015. 1484 [I-D.ietf-hip-native-nat-traversal] 1485 Keranen, A. and J. Melen, "Native NAT Traversal Mode for 1486 the Host Identity Protocol", draft-ietf-hip-native-nat- 1487 traversal-08 (work in progress), January 2015. 1489 [I-D.ietf-hip-rfc5203-bis] 1490 Laganier, J. and L. Eggert, "Host Identity Protocol (HIP) 1491 Registration Extension", draft-ietf-hip-rfc5203-bis-06 1492 (work in progress), September 2014. 1494 [I-D.ietf-hip-rfc5204-bis] 1495 Laganier, J. and L. Eggert, "Host Identity Protocol (HIP) 1496 Rendezvous Extension", draft-ietf-hip-rfc5204-bis-04 (work 1497 in progress), June 2014. 1499 [I-D.ietf-hip-rfc5205-bis] 1500 Laganier, J., "Host Identity Protocol (HIP) Domain Name 1501 System (DNS) Extension", draft-ietf-hip-rfc5205-bis-05 1502 (work in progress), July 2014. 1504 [I-D.ietf-hip-rfc5206-bis] 1505 Henderson, T., Vogt, C., and J. Arkko, "Host Mobility with 1506 the Host Identity Protocol", draft-ietf-hip-rfc5206-bis-06 1507 (work in progress), July 2013. 1509 [I-D.ietf-hip-rfc6253-bis] 1510 Heer, T. and S. Varjonen, "Host Identity Protocol 1511 Certificates", draft-ietf-hip-rfc6253-bis-01 (work in 1512 progress), October 2013. 1514 [RFC5482] Eggert, L. and F. Gont, "TCP User Timeout Option", RFC 1515 5482, March 2009. 1517 [RFC7343] Laganier, J. and F. Dupont, "An IPv6 Prefix for Overlay 1518 Routable Cryptographic Hash Identifiers Version 2 1519 (ORCHIDv2)", RFC 7343, September 2014. 1521 [RFC7401] Moskowitz, R., Heer, T., Jokela, P., and T. Henderson, 1522 "Host Identity Protocol Version 2 (HIPv2)", RFC 7401, 1523 April 2015. 1525 [RFC7402] Jokela, P., Moskowitz, R., and J. Melen, "Using the 1526 Encapsulating Security Payload (ESP) Transport Format with 1527 the Host Identity Protocol (HIP)", RFC 7402, April 2015. 1529 16.2. Informative references 1531 [IEEE.802-15-4.2011] 1532 , "Information technology - Telecommunications and 1533 information exchange between systems - Local and 1534 metropolitan area networks - Specific requirements - Part 1535 15.4: Wireless Medium Access Control (MAC) and Physical 1536 Layer (PHY) Specifications for Low-Rate Wireless Personal 1537 Area Networks (WPANs)", IEEE Standard 802.15.4, September 1538 2011, . 1541 [Nik2001] Nikander, P., "Denial-of-Service, Address Ownership, and 1542 Early Authentication in the IPv6 World", in Proceesings of 1543 Security Protocols, 9th International Workshop, Cambridge, 1544 UK, April 25-27 2001, LNCS 2467, pp. 12-26, Springer, 1545 2002. 1547 [RFC2136] Vixie, P., Thomson, S., Rekhter, Y., and J. Bound, 1548 "Dynamic Updates in the Domain Name System (DNS UPDATE)", 1549 RFC 2136, April 1997. 1551 [RFC2535] Eastlake, D., "Domain Name System Security Extensions", 1552 RFC 2535, March 1999. 1554 [RFC2766] Tsirtsis, G. and P. Srisuresh, "Network Address 1555 Translation - Protocol Translation (NAT-PT)", RFC 2766, 1556 February 2000. 1558 [RFC3022] Srisuresh, P. and K. Egevang, "Traditional IP Network 1559 Address Translator (Traditional NAT)", RFC 3022, January 1560 2001. 1562 [RFC3102] Borella, M., Lo, J., Grabelsky, D., and G. Montenegro, 1563 "Realm Specific IP: Framework", RFC 3102, October 2001. 1565 [RFC3748] Aboba, B., Blunk, L., Vollbrecht, J., Carlson, J., and H. 1566 Levkowetz, "Extensible Authentication Protocol (EAP)", RFC 1567 3748, June 2004. 1569 [RFC4225] Nikander, P., Arkko, J., Aura, T., Montenegro, G., and E. 1570 Nordmark, "Mobile IP Version 6 Route Optimization Security 1571 Design Background", RFC 4225, December 2005. 1573 [RFC4306] Kaufman, C., "Internet Key Exchange (IKEv2) Protocol", RFC 1574 4306, December 2005. 1576 [RFC4423] Moskowitz, R. and P. Nikander, "Host Identity Protocol 1577 (HIP) Architecture", RFC 4423, May 2006. 1579 [RFC5218] Thaler, D. and B. Aboba, "What Makes For a Successful 1580 Protocol?", RFC 5218, July 2008. 1582 [RFC5338] Henderson, T., Nikander, P., and M. Komu, "Using the Host 1583 Identity Protocol with Legacy Applications", RFC 5338, 1584 September 2008. 1586 [RFC5887] Carpenter, B., Atkinson, R., and H. Flinck, "Renumbering 1587 Still Needs Work", RFC 5887, May 2010. 1589 [RFC6078] Camarillo, G. and J. Melen, "Host Identity Protocol (HIP) 1590 Immediate Carriage and Conveyance of Upper-Layer Protocol 1591 Signaling (HICCUPS)", RFC 6078, January 2011. 1593 [RFC6250] Thaler, D., "Evolution of the IP Model", RFC 6250, May 1594 2011. 1596 [RFC6281] Cheshire, S., Zhu, Z., Wakikawa, R., and L. Zhang, 1597 "Understanding Apple's Back to My Mac (BTMM) Service", RFC 1598 6281, June 2011. 1600 [RFC6317] Komu, M. and T. Henderson, "Basic Socket Interface 1601 Extensions for the Host Identity Protocol (HIP)", RFC 1602 6317, July 2011. 1604 [RFC6537] Ahrenholz, J., "Host Identity Protocol Distributed Hash 1605 Table Interface", RFC 6537, February 2012. 1607 [RFC6538] Henderson, T. and A. Gurtov, "The Host Identity Protocol 1608 (HIP) Experiment Report", RFC 6538, March 2012. 1610 [amir-hip] 1611 Amir, K., Forsgren, H., Grahn, K., Karvi, T., and G. 1612 Pulkkis, "Security and Trust of Public Key Cryptography 1613 for HIP and HIP Multicast", International Journal of 1614 Dependable and Trustworthy Information Systems (IJDTIS), 1615 2(3), 17-35, DOI: 10.4018/jdtis.2011070102, 2013. 1617 [aura-dos] 1618 Aura, T., Nikander, P., and J. Leiwo, "DOS-resistant 1619 Authentication with Client Puzzles", 8th International 1620 Workshop on Security Protocols, pages 170-177. Springer, , 1621 April 2001. 1623 [beal-dos] 1624 Beal, J. and T. Shephard, "Deamplification of DoS Attacks 1625 via Puzzles", , October 2004. 1627 [chiappa-endpoints] 1628 Chiappa, J., "Endpoints and Endpoint Names: A Proposed 1629 Enhancement to the Internet Architecture", URL 1630 http://www.chiappa.net/~jnc/tech/endpoints.txt, 1999. 1632 [heer-end-host] 1633 Heer, T., Hummen, R., Komu, M., Goetz, S., and K. Wehre, 1634 "End-host Authentication and Authorization for Middleboxes 1635 based on a Cryptographic Namespace", ICC2009 Communication 1636 and Information Systems Security Symposium, , 2009. 1638 [heer-midauth] 1639 Heer, T. and M. Komu, "End-Host Authentication for HIP 1640 Middleboxes", Working draft draft-heer-hip-middle-auth-02, 1641 September 2009. 1643 [henderson-vpls] 1644 Henderson, T. and D. Mattes, "HIP-based Virtual Private 1645 LAN Service (HIPLS)", Working draft draft-henderson-hip- 1646 vpls-07, Dec 2013. 1648 [hip-srtp] 1649 Tschofenig, H., Muenz, F., and M. Shanmugam, "Using SRTP 1650 transport format with HIP", Working draft draft- 1651 tschofenig-hiprg-hip-srtp-01, October 2005. 1653 [karvonen-usable] 1654 Karvonen, K., Komu, M., and A. Gurtov, "Usable Security 1655 Management with Host Identity Protocol", 7th ACS/IEEE 1656 International Conference on Computer Systems and 1657 Applications, (AICCSA-2009), 2009. 1659 [komu-cloud] 1660 Komu, M., Sethi, M., Mallavarapu, R., Oirola, H., Khan, 1661 R., and S. Tarkoma, "Secure Networking for Virtual 1662 Machines in the Cloud", International Workshop on Power 1663 and QoS Aware Computing (PQoSCom2012), IEEE, ISBN: 1664 978-1-4244-8567-3, September 2012. 1666 [komu-diss] 1667 Komu, M., "A Consolidated Namespace for Network 1668 Applications, Developers, Administrators and Users", 1669 Dissertation, Aalto University, Espoo, Finland ISBN: 1670 978-952-60-4904-5 (printed), ISBN: 978-952-60-4905-2 1671 (electronic). , December 2012. 1673 [komu-leap] 1674 Komu, M. and J. Lindqvist, "Leap-of-Faith Security is 1675 Enough for IP Mobility", 6th Annual IEEE Consumer 1676 Communications and Networking Conference IEEE CCNC 2009, 1677 Las Vegas, Nevada, , January 2009. 1679 [komu-mitigation] 1680 Komu, M., Tarkoma, S., and A. Lukyanenko, "Mitigation of 1681 Unsolicited Traffic Across Domains with Host Identities 1682 and Puzzles", 15th Nordic Conference on Secure IT Systems 1683 (NordSec 2010), Springer Lecture Notes in Computer 1684 Science, Volume 7127, pp. 33-48, ISBN: 978-3-642-27936-2, 1685 October 2010. 1687 [kovacshazi-host] 1688 Kovacshazi, Z. and R. Vida, "Host Identity Specific 1689 Multicast", International conference on Networking and 1690 Services (ICNS'06), IEEE Computer Society, Los Alamitos, 1691 CA, USA, 1692 http://doi.ieeecomputersociety.org/10.1109/ICNS.2007.66, 1693 2007. 1695 [leva-barriers] 1696 Levae, A., Komu, M., and S. Luukkainen, "Adoption Barriers 1697 of Network-layer Protocols: the Case of Host Identity 1698 Protocol", The International Journal of Computer and 1699 Telecommunications Networking, ISSN: 1389-1286, March 1700 2013. 1702 [lindqvist-enterprise] 1703 Lindqvist, J., Vehmersalo, E., Manner, J., and M. Komu, 1704 "Enterprise Network Packet Filtering for Mobile 1705 Cryptographic Identities", International Journal of 1706 Handheld Computing Research, 1 (1), 79-94, , January-March 1707 2010. 1709 [nsrg-report] 1710 Lear, E. and R. Droms, "What's In A Name:Thoughts from the 1711 NSRG", draft-irtf-nsrg-report-10 (work in progress), 1712 September 2003. 1714 [paine-hip] 1715 Paine, R., "Beyond HIP: The End to Hacking As We Know It", 1716 BookSurge Publishing, ISBN: 1439256047, 9781439256046, 1717 2009. 1719 [pham-leap] 1720 Pham, V. and T. Aura, "Security Analysis of Leap-of-Faith 1721 Protocols", Seventh ICST International Conference on 1722 Security and Privacy for Communication Networks, , 1723 September 2011. 1725 [sarela-bloom] 1726 Saerelae, M., Esteve Rothenberg, C., Zahemszky, A., 1727 Nikander, P., and J. Ott, "BloomCasting: Security in Bloom 1728 filter based multicast", , Lecture Notes in Computer 1729 Science 2012, , pages 1-16, Springer Berlin Heidelberg, 1730 2012. 1732 [schuetz-intermittent] 1733 Schuetz, S., Eggert, L., Schmid, S., and M. Brunner, 1734 "Protocol enhancements for intermittently connected 1735 hosts", SIGCOMM Comput. Commun. Rev., 35(3):5-18, , July 1736 2005. 1738 [shields-hip] 1739 Shields, C. and J. Garcia-Luna-Aceves, "The HIP protocol 1740 for hierarchical multicast routing", Proceedings of the 1741 seventeenth annual ACM symposium on Principles of 1742 distributed computing, pages 257-266. ACM, New York, NY, 1743 USA, ISBN: 0-89791-977-7, DOI: 10.1145/277697.277744, 1744 1998. 1746 [tritilanunt-dos] 1747 Tritilanunt, S., Boyd, C., Foo, E., and J. Nieto, 1748 "Examining the DoS Resistance of HIP", OTM Workshops (1), 1749 volume 4277 of Lecture Notes in Computer Science, pages 1750 616-625,Springer , 2006. 1752 [urien-rfid-draft] 1753 Urien, P., Lee, G., and G. Pujolle, "HIP support for 1754 RFIDs", IRTF Working draft draft-irtf-hiprg-rfid-07, April 1755 2013. 1757 [urien-rfid] 1758 Urien, P., Chabanne, H., Bouet, M., de Cunha, D., Guyot, 1759 V., Pujolle, G., Paradinas, P., Gressier, E., and J. 1760 Susini, "HIP-based RFID Networking Architecture", IFIP 1761 International Conference on Wireless and Optical 1762 Communications Networks, DOI: 10.1109/WOCN.2007.4284140, 1763 July 2007. 1765 [varjonen-split] 1766 Varjonen, S., Komu, M., and A. Gurtov, "Secure and 1767 Efficient IPv4/IPv6 Handovers Using Host-Based Identifier- 1768 Location Split", Journal of Communications Software and 1769 Systems, 6(1), 2010, ISSN: 18456421, 2010. 1771 [xin-hip-lib] 1772 Xin, G., "Host Identity Protocol Version 2.5", Master's 1773 Thesis, Aalto University, Espoo, Finland, , June 2012. 1775 [xueyong-hip] 1776 Xueyong, Z., Zhiguo, D., and W. Xinling, "A Multicast 1777 Routing Algorithm Applied to HIP-Multicast Model", 1778 Proceedings of the 2011 International Conference on 1779 Network Computing and Information Security - Volume 01 1780 (NCIS '11), Vol. 1. IEEE Computer Society, Washington, DC, 1781 USA, pages 169-174, DOI: 10.1109/NCIS.2011.42, 2011. 1783 [xueyong-secure] 1784 Xueyong, Z. and J. Atwood, "A Secure Multicast Model for 1785 Peer-to-Peer and Access Networks Using the Host Identity 1786 Protocol", Consumer Communications and Networking 1787 Conference. CCNC 2007. 4th IEEE, pages 1098,1102, DOI: 1788 10.1109/CCNC.2007.221, January 2007. 1790 [ylitalo-diss] 1791 Ylitalo, J., "Secure Mobility at Multiple Granularity 1792 Levels over Heterogeneous Datacom Networks", Dissertation, 1793 Helsinki University of Technology, Espoo, Finland ISBN 1794 978-951-22-9531-9, 2008. 1796 [ylitalo-spinat] 1797 Ylitalo, J., Salmela, P., and H. Tschofenig, "SPINAT: 1798 Integrating IPsec into overlay routing", Proceedings of 1799 the First International Conference on Security and Privacy 1800 for Emerging Areas in Communication Networks (SecureComm 1801 2005). Athens, Greece. IEEE Computer Society, pages 1802 315-326, ISBN: 0-7695-2369-2, September 2005. 1804 [zhang-revocation] 1805 Zhang, D., Kuptsov, D., and S. Shen, "Host Identifier 1806 Revocation in HIP", IRTF Working draft draft-irtf-hiprg- 1807 revocation-05, Mar 2012. 1809 Authors' Addresses 1811 Robert Moskowitz (editor) 1812 Verizon 1813 1000 Bent Creek Blvd, Suite 200 1814 Mechanicsburg, PA 1815 USA 1817 Email: robert.moskowitz@verizon.com 1819 Miika Komu 1820 Ericsson 1821 Hirsalantie 11 1822 02420 Jorvas 1823 Finland 1825 Email: miika.komu@ericsson.com