idnits 2.17.1 draft-ietf-intarea-broadcast-consider-04.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/1id-guidelines.txt: ---------------------------------------------------------------------------- No issues found here. Checking nits according to https://www.ietf.org/id-info/checklist : ---------------------------------------------------------------------------- No issues found here. Miscellaneous warnings: ---------------------------------------------------------------------------- == The copyright year in the IETF Trust and authors Copyright Line does not match the current year -- The document date (August 29, 2017) is 2432 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- == Outdated reference: A later version (-05) exists of draft-ietf-dnssd-privacy-00 == Outdated reference: A later version (-05) exists of draft-ietf-intarea-hostname-practice-00 -- Obsolete informational reference (is this intentional?): RFC 3315 (Obsoleted by RFC 8415) -- Obsolete informational reference (is this intentional?): RFC 4941 (Obsoleted by RFC 8981) Summary: 0 errors (**), 0 flaws (~~), 3 warnings (==), 3 comments (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 Internet Engineering Task Force R. Winter 3 Internet-Draft University of Applied Sciences Augsburg 4 Intended status: Informational M. Faath 5 Expires: March 2, 2018 Conntac GmbH 6 F. Weisshaar 7 University of Applied Sciences Augsburg 8 August 29, 2017 10 Privacy considerations for IP broadcast and multicast protocol designers 11 draft-ietf-intarea-broadcast-consider-04 13 Abstract 15 A number of application-layer protocols make use of IP broadcasts or 16 multicast messages for functions like local service discovery or name 17 resolution. Some of these functions can only be implemented 18 efficiently using such mechanisms. When using broadcasts or 19 multicast messages, a passive observer in the same broadcast/ 20 multicast domain can trivially record these messages and analyze 21 their content. Therefore, broadcast/multicast protocol designers 22 need to take special care when designing their protocols. 24 Status of This Memo 26 This Internet-Draft is submitted in full conformance with the 27 provisions of BCP 78 and BCP 79. 29 Internet-Drafts are working documents of the Internet Engineering 30 Task Force (IETF). Note that other groups may also distribute 31 working documents as Internet-Drafts. The list of current Internet- 32 Drafts is at http://datatracker.ietf.org/drafts/current/. 34 Internet-Drafts are draft documents valid for a maximum of six months 35 and may be updated, replaced, or obsoleted by other documents at any 36 time. It is inappropriate to use Internet-Drafts as reference 37 material or to cite them other than as "work in progress." 39 This Internet-Draft will expire on March 2, 2018. 41 Copyright Notice 43 Copyright (c) 2017 IETF Trust and the persons identified as the 44 document authors. All rights reserved. 46 This document is subject to BCP 78 and the IETF Trust's Legal 47 Provisions Relating to IETF Documents 48 (http://trustee.ietf.org/license-info) in effect on the date of 49 publication of this document. Please review these documents 50 carefully, as they describe your rights and restrictions with respect 51 to this document. Code Components extracted from this document must 52 include Simplified BSD License text as described in Section 4.e of 53 the Trust Legal Provisions and are provided without warranty as 54 described in the Simplified BSD License. 56 Table of Contents 58 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . 2 59 1.1. Types and usage of broadcast and multicast . . . . . . . 4 60 1.2. Requirements Language . . . . . . . . . . . . . . . . . . 4 61 2. Privacy considerations . . . . . . . . . . . . . . . . . . . 4 62 2.1. Message frequency . . . . . . . . . . . . . . . . . . . . 5 63 2.2. Persistent identifiers . . . . . . . . . . . . . . . . . 5 64 2.3. Anticipate user behavior . . . . . . . . . . . . . . . . 6 65 2.4. Consider potential correlation . . . . . . . . . . . . . 7 66 2.5. Configurability . . . . . . . . . . . . . . . . . . . . . 7 67 3. Operational considerations . . . . . . . . . . . . . . . . . 8 68 4. Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . 8 69 5. Other considerations . . . . . . . . . . . . . . . . . . . . 9 70 6. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 9 71 7. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 9 72 8. Security Considerations . . . . . . . . . . . . . . . . . . . 9 73 9. References . . . . . . . . . . . . . . . . . . . . . . . . . 10 74 9.1. Normative References . . . . . . . . . . . . . . . . . . 10 75 9.2. Informative References . . . . . . . . . . . . . . . . . 10 76 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . 12 78 1. Introduction 80 Broadcast and multicast messages have a large (and to the sender 81 unknown) receiver group by design. Because of that, these two 82 mechanisms are vital for a number of basic network functions such as 83 auto-configuration or link-layer address lookup. Also application 84 developers use broadcast/multicast messages to implement things like 85 local service or peer discovery and it appears that an increasing 86 number of applications make use of it as suggested by experimental 87 results obtained on campus networks including the IETF meeting 88 network [TRAC2016]. That is not entirely surprising. As RFC 919 89 [RFC0919] puts it, "The use of broadcasts [...] is a good base for 90 many applications". Broadcast and multicast functionality in a 91 subnetwork are therefore important as a lack thereof renders the 92 protocols underlying these mechanisms inoperable [RFC3819]. 94 Using broadcast/multicast can become problematic if the information 95 that is being distributed can be regarded as sensitive or when the 96 information that is distributed by multiple of these protocols can be 97 correlated in a way that sensitive data can be derived. This is 98 clearly true for any protocol, but broadcast/multicast is special in 99 at least two respects: 101 (a) The aforementioned large receiver group, consisting of receivers 102 unknown to the sender. This makes eavesdropping without special 103 privileges or a special location in the network trivial for 104 anybody in the broadcast/multicast domain. 106 (b) Encryption is more difficult when broadcast/multicast messages, 107 leaving content of these messages in the clear and making it 108 easier to spoof and replay them. 110 Given the above, privacy protection for protocols based on broadcast 111 or multicast communication is significantly more difficult compared 112 to unicast communication and at the same time invading the privacy is 113 much easier. 115 Privacy considerations of IETF-specified protocols have received some 116 attention in the recent past (e.g. RFC 7721 [RFC7721] or RFC 7919 117 [RFC7819]). There is also general guidance available for document 118 authors on when and how to include a privacy considerations section 119 in their documents and on how to evaluate the privacy implications of 120 Internet protocols [RFC6973]. RFC6973 also describes potential 121 threats to privacy in great detail and lists terminology that is also 122 used in this document. 124 In contrast to RFC6973, this document contains a number of privacy 125 considerations especially for broadcast/multicast protocol designers 126 that are intended to reduce the likelihood that a broadcast/multicast 127 protocol can be misused to collect sensitive data about devices, 128 users and groups of users on a broadcast/multicast domain. These 129 considerations particularly apply to protocols designed outside the 130 IETF for two reasons. For one, non-standard protocols will likely 131 not receive operational attention and support in making them more 132 secure such as e.g. DHCP snooping does for DHCP because they 133 typically are not documented. The other reason is that these 134 protocols have been designed in isolation, where a set of 135 considerations to follow is useful in the absence of a larger 136 community providing feedback. In particular, carelessly designed 137 broadcast/multicast protocols can break privacy efforts at different 138 layers of the protocol stack such as MAC address or IP address 139 randomization [RFC4941]. 141 1.1. Types and usage of broadcast and multicast 143 In IPv4, two major types of broadcast addresses exist, the limited 144 broadcast which is defined as all-ones (255.255.255.255, defined in 145 section 5.3.5.1 of [RFC1812]) and the directed broadcast with the 146 given network prefix of an IP address and the host part of all-ones 147 (defined in section 5.3.5.2. of [RFC1812]). Broadcast packets are 148 received by all nodes in a subnetwork. Limited broadcasts never 149 transit a router. The same is true for directed broadcasts by 150 default, but routers MAY provide an option to do this [RFC2644]. 151 IPv6 on the other hand does not provide broadcast addresses but 152 solely relies on multicast [RFC4291]. 154 In contrast to broadcast addresses, multicast addresses represent an 155 identifier for a set of interfaces that can be a set different from 156 all nodes in the subnetwork. All interfaces that are identified by a 157 given multicast address receive packets destined towards that address 158 and are called a multicast group. In both IPv4 and IPv6, multiple 159 pre-defined multicast addresses exist. The ones most relevant for 160 this document are the ones with subnet scope. For IPv4, an IP prefix 161 is reserved for this purpose called the Local Network Control Block 162 (224.0.0.0/24, defined in section 4 of [RFC5771]). For IPv6, the 163 relevant multicast addresses are the two All Nodes Addresses, which 164 every IPv6-capable host is required to recognize as identifying 165 itself (see section 2.7.1 of [RFC4291]). 167 Typical usage of these addresses include local service discovery 168 (e.g. mDNS [RFC6762] and LLMNR [RFC4795] make use of multicast), 169 autoconfiguration (e.g. DHCPv4 [RFC2131] uses broadcasts and DHCPv6 170 [RFC3315] uses multicast addresses) and other vital network services 171 such as address resolution or duplicate address detection. But 172 besides these core network functions, also applications make use of 173 broadcast and multicast functionality, often implementing proprietary 174 protocols. In sum, these protocols distribute a diverse set of 175 potentially privacy sensitive information to a large receiver group. 177 1.2. Requirements Language 179 The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", 180 "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this 181 document are to be interpreted as described in RFC 2119 [RFC2119]. 183 2. Privacy considerations 185 There are a few obvious and a few not necessarily obvious things 186 designers of broadcast/multicast protocols should consider in respect 187 to the privacy implications of their protocol. Most of these items 188 are based on protocol behavior observed as part of experiments on 189 operational networks [TRAC2016]. 191 2.1. Message frequency 193 Frequent broadcast/multicast traffic caused by an application can 194 give away user behavior and online connection times. This allows a 195 passive observer to potentially deduce a user's current activity 196 (e.g. a game) and it allows to create an online profile (i.e. times 197 the user is on the network). The higher the frequency of these 198 messages, the more accurate this profile will be. Given that 199 broadcasts/multicasts are only visible in the same broadcast/ 200 multicast domain, these messages also give the rough location of the 201 user away (e.g. a campus or building). 203 This behavior has e.g. been observed by a synchronization mechanism 204 of a popular application, where multiple messages have been sent per 205 minute via broadcast. Given this behavior, it is possible to record 206 a device's time on the network with a sub-minute accuracy given only 207 the traffic of this single application installed on the device. But 208 also services used for local name resolution in modern operating 209 systems utilize broadcast/multicast protocols (e.g. mDNS, LLMNR or 210 NetBIOS) to announce for example their shares regularly and allow a 211 tracking of the online time of a device. 213 If a protocol relies on frequent or periodic broadcast/multicast 214 messages, the frequency SHOULD be chosen conservatively, in 215 particular if the messages contain persistent identifiers (see next 216 subsection). Also, intelligent message suppression mechanisms such 217 as the ones employed in mDNS [RFC6762] SHOULD be implemented. The 218 lower the frequency of broadcast messages, the harder traffic 219 analysis and surveillance becomes. 221 2.2. Persistent identifiers 223 A few broadcast/multicast protocols observed in the wild make use of 224 persistent identifiers. This includes the use of host names or more 225 abstract persistent identifiers such as a universally unique 226 identifiers (UUID) or similar. These IDs, which e.g. identify the 227 installation of a certain application might not change across updates 228 of the software and are therefore extremely long lived. This allows 229 a passive observer to track a user precisely if broadcast/multicast 230 messages are frequent. This is even true in case the IP and/or MAC 231 address changes. Such identifiers also allow two different 232 interfaces (e.g. WiFi and Ethernet) to be correlated to the same 233 device. If the application makes use of persistent identifiers for 234 multiple installations of the same application for the same user, 235 this even allows to infer that different devices belong to the same 236 user. 238 The aforementioned broadcast messages from a synchronization 239 mechanism of a popular application also included a persistent 240 identifier in every broadcast. This identifier did never change 241 after the application was installed and allowed to track a device 242 even when it changed its network interface or when it connected to a 243 different network. 245 Persistent IDs are considered bad practice in general for broadcast 246 and multicast communication, as persistent application layer IDs will 247 make efforts on lower layers to randomize identifiers (e.g. 248 [I-D.huitema-6man-random-addresses]) useless or at least much more 249 difficult. When broadcast/multicast protocols need to make use of 250 IDs, frequent rotations of these IDs SHOULD be considered to make 251 user tracking more difficult. 253 2.3. Anticipate user behavior 255 A large number of users name their device after themselves, either 256 using their first name, last name or both. Often a host name 257 includes the type, model or maker of a device, its function or 258 includes language specific information. Based on data gathered 259 during experiments performed at IETF meetings and at a large campus 260 network, this appears currently to be prevalent user behavior 261 [TRAC2016]. For protocols using the host name as part of the 262 messages, this clearly will reveal personally identifiable 263 information to everyone on the local network. This information can 264 also be used to mount more sophisticated attacks, when e.g. the owner 265 of a device is identified (as an interesting target) or properties of 266 the device are known (e.g. known vulnerabilities). 268 Some of the most commonly used operating systems include the name the 269 user chooses for the user account during the installation process as 270 part of the host name of the device. The name of the operating 271 system can also be included, revealing therefore two pieces of 272 information, which can be regarded as private information if the host 273 name is used in broadcast/multicast messages. 275 Where possible, the use of host names and other user provided 276 information in broadcast/multicast protocols SHOULD be avoided. If 277 only a persistent ID is needed, this can be generated randomly. An 278 application might want to display the information it will broadcast 279 on the LAN at install/config time, so the user is at least aware of 280 the application's behavior. More host name considerations can be 281 found in [I-D.ietf-intarea-hostname-practice]. More information on 282 user participation can be found in RFC 6973 [RFC6973]. 284 2.4. Consider potential correlation 286 A large number of services and applications make use of the 287 broadcast/multicast mechanism. That means there are various sources 288 of information that are easily accessible by a passive observer. In 289 isolation, the information these protocols reveal might seem 290 harmless, but given multiple such protocols, it might be possible to 291 correlate this information. E.g. a protocol that uses frequent 292 messages including a UUID to identify the particular installation 293 does not give the identity of the user away. But a single message 294 including the user's host name might just do that and it can be 295 correlated using e.g. the MAC address of the device's interface. 297 In the experiments described in [TRAC2016], it was possible to 298 correlate frequently sent broadcast messages that included a unique 299 identifier with other broadcast/multicast messages containing 300 usernames (e.g. mDNS, LLMNR or NetBIOS), but also relationships to 301 other users. This allowed to reveal the real identity of the users 302 of many devices but it also gave some information about their social 303 environment away. 305 A broadcast protocol designer should be aware of the fact that even 306 if - in isolation - the information a protocol leaks seems harmless, 307 there might be ways to correlate that information with other 308 broadcast protocol information to reveal sensitive information about 309 a user. 311 2.5. Configurability 313 A lot of applications and services using broadcast/multicast 314 protocols do not include the means to declare "safe" environments 315 (e.g. based on the SSID of a WiFi network and the MAC addresses of 316 the access points). E.g. a device connected to a public WiFi will 317 likely broadcast the same information as when connected to the home 318 network. It would be beneficial if certain behavior could be 319 restricted to "safe" environments. 321 A popular operating system e.g. allows the user to specify the trust 322 level of the network the device connects to, which for example 323 restricts specific system services (using broadcast/multicast 324 messages for their normal operation) to be used in untrusted 325 networks. Such functionality could implemented as part of an 326 application. 328 An application developer making use of broadcasts/multicasts as part 329 of the application SHOULD make the broadcast feature, if possible, 330 configurable, so that potentially sensitive information does not leak 331 on public networks, where the threat to privacy is much larger. 333 3. Operational considerations 335 Besides changing end-user behavior, choosing sensible defaults as an 336 operating system vendor (e.g. for suggesting host names) and the 337 considerations for protocol designers mentioned in this document, 338 there are things that the network administrators/operators can do to 339 limit the above mentioned problems. 341 A feature not uncommonly found on access points e.g. is to filter 342 broadcast and multicast traffic. This will potentially break certain 343 applications or some of their functionality but will also protect the 344 users from potentially leaking sensitive information. 346 4. Summary 348 Increasingly, applications rely on broadcast and multicast messages. 349 For some, broadcasts/multicasts are the basis of their application 350 logic, others use broadcasts/multicasts to improve certain aspects of 351 the application but are fully functional in case broadcasts/ 352 multicasts fail. Irrespective of the role of broadcast and multicast 353 messages for the application, the designers of protocols that make 354 use of them should be very careful in their protocol design because 355 of the special nature of broad- and multicast. 357 It is not always possible to implement certain functionality via 358 unicast, but in case a protocol designer chooses to rely on 359 broadcast/multicast, the following should be carefully considered: 361 o IETF-specified protocols, such as mDNS [RFC6762], SHOULD be used 362 if possible as operational support might exist to protect against 363 the leakage of private information. Also, for some protocols 364 privacy extensions are being specified, which can be used if 365 implemented. E.g. for DNS-SD privacy extensions are documented in 366 [I-D.ietf-dnssd-privacy] 368 o Using user-specified information inside broadcast/multicast 369 messages SHOULD be avoided, as users will often use personal 370 information or other information aiding attackers, in particular 371 if the user is unaware about how that information is being used 373 o The use of persistent IDs in messages SHOULD be avoided, as this 374 allows user tracking, correlation and potentially has a 375 devastating effect on other privacy protection mechanisms 377 o If you really must use a broadcast/multicast protocol and cannot 378 use an IETF-specified protocol, then: 380 * You SHOULD be very conservative in how frequently you send 381 messages as an effort in data minimization 383 * You SHOULD seek advice from IETF-specified protocols such as 384 message suppression in mDNS 386 * You SHOULD try to design the protocol in a way that the 387 information cannot be correlated with other information in 388 broadcast/multicast messages 390 * You SHOULD let the user configure safe environments if possible 391 (e.g. based on the SSID) 393 5. Other considerations 395 Besides privacy implications, frequent broadcasting also represents a 396 performance problem. In particular in certain wireless technologies 397 such as 802.11, broadcast and multicast are transmitted at a much 398 lower rate (the lowest common denominator rate) compared to unicast 399 and therefore have a much bigger impact on the overall available 400 airtime [I-D.perkins-intarea-multicast-ieee802]. Further, it will 401 limit the ability for devices to go to sleep if frequent broadcasts 402 are being sent. A similar problem in respect to Router 403 Advertisements is addressed in 404 [I-D.ietf-v6ops-reducing-ra-energy-consumption]. In that respect 405 broadcasts can be used for another class of attacks that not related 406 to privacy. The potential impact on network performance should 407 nevertheless be considered by broadcast protocol designers. 409 6. Acknowledgments 411 We would like to thank Eliot Lear, Joe Touch and Stephane Bortzmeyer 412 for their valuable input to this document. 414 This work was partly supported by the European Commission under grant 415 agreement FP7-318627 mPlane. Support does not imply endorsement. 417 7. IANA Considerations 419 This memo includes no request to IANA. 421 8. Security Considerations 423 This document deals with privacy-related considerations of broadcast- 424 and multicast-based protocols. It contains advice for designers of 425 such protocols to minimize the leakage of privacy-sensitive 426 information. The intent of the advice is to make sure that 427 identities will remain anonymous and user tracking will be made 428 difficult. 430 It should be noted that certain applications could make use of 431 existing mechanisms to protect multicast traffic such as the ones 432 defined in [RFC5374]. Examples of such applications can be found in 433 Appendix A. of [RFC5374]. Given the required infrastructure and 434 assumptions about these applications and the security infrastructure, 435 many applications will not be able to make use of such mechanisms. 437 9. References 439 9.1. Normative References 441 [RFC2119] Bradner, S., "Key words for use in RFCs to Indicate 442 Requirement Levels", BCP 14, RFC 2119, March 1997. 444 9.2. Informative References 446 [I-D.huitema-6man-random-addresses] 447 Huitema, C., "Implications of Randomized Link Layers 448 Addresses for IPv6 Address Assignment", draft-huitema- 449 6man-random-addresses-03 (work in progress), March 2016. 451 [I-D.ietf-dnssd-privacy] 452 Huitema, C. and D. Kaiser, "Privacy Extensions for DNS- 453 SD", draft-ietf-dnssd-privacy-00 (work in progress), 454 October 2016. 456 [I-D.ietf-intarea-hostname-practice] 457 Huitema, C. and D. Thaler, "Current Hostname Practice 458 Considered Harmful", draft-ietf-intarea-hostname- 459 practice-00 (work in progress), October 2015. 461 [I-D.ietf-v6ops-reducing-ra-energy-consumption] 462 Yourtchenko, A. and L. Colitti, "Reducing energy 463 consumption of Router Advertisements", draft-ietf-v6ops- 464 reducing-ra-energy-consumption-03 (work in progress), 465 November 2015. 467 [I-D.perkins-intarea-multicast-ieee802] 468 Perkins, C., Stanley, D., Kumari, W., and J. Zuniga, 469 "Multicast Considerations over IEEE 802 Wireless Media", 470 draft-perkins-intarea-multicast-ieee802-03 (work in 471 progress), July 2017. 473 [RFC0919] Mogul, J., "Broadcasting Internet Datagrams", STD 5, RFC 474 919, DOI 10.17487/RFC0919, October 1984, 475 . 477 [RFC1812] Baker, F., Ed., "Requirements for IP Version 4 Routers", 478 RFC 1812, DOI 10.17487/RFC1812, June 1995, 479 . 481 [RFC2131] Droms, R., "Dynamic Host Configuration Protocol", RFC 482 2131, DOI 10.17487/RFC2131, March 1997, 483 . 485 [RFC2644] Senie, D., "Changing the Default for Directed Broadcasts 486 in Routers", BCP 34, RFC 2644, DOI 10.17487/RFC2644, 487 August 1999, . 489 [RFC3315] Droms, R., Ed., Bound, J., Volz, B., Lemon, T., Perkins, 490 C., and M. Carney, "Dynamic Host Configuration Protocol 491 for IPv6 (DHCPv6)", RFC 3315, DOI 10.17487/RFC3315, July 492 2003, . 494 [RFC3819] Karn, P., Ed., Bormann, C., Fairhurst, G., Grossman, D., 495 Ludwig, R., Mahdavi, J., Montenegro, G., Touch, J., and L. 496 Wood, "Advice for Internet Subnetwork Designers", BCP 89, 497 RFC 3819, DOI 10.17487/RFC3819, July 2004, 498 . 500 [RFC4291] Hinden, R. and S. Deering, "IP Version 6 Addressing 501 Architecture", RFC 4291, DOI 10.17487/RFC4291, February 502 2006, . 504 [RFC4795] Aboba, B., Thaler, D., and L. Esibov, "Link-local 505 Multicast Name Resolution (LLMNR)", RFC 4795, DOI 506 10.17487/RFC4795, January 2007, 507 . 509 [RFC4941] Narten, T., Draves, R., and S. Krishnan, "Privacy 510 Extensions for Stateless Address Autoconfiguration in 511 IPv6", RFC 4941, DOI 10.17487/RFC4941, September 2007, 512 . 514 [RFC5374] Weis, B., Gross, G., and D. Ignjatic, "Multicast 515 Extensions to the Security Architecture for the Internet 516 Protocol", RFC 5374, DOI 10.17487/RFC5374, November 2008, 517 . 519 [RFC5771] Cotton, M., Vegoda, L., and D. Meyer, "IANA Guidelines for 520 IPv4 Multicast Address Assignments", BCP 51, RFC 5771, DOI 521 10.17487/RFC5771, March 2010, 522 . 524 [RFC6762] Cheshire, S. and M. Krochmal, "Multicast DNS", RFC 6762, 525 DOI 10.17487/RFC6762, February 2013, 526 . 528 [RFC6973] Cooper, A., Tschofenig, H., Aboba, B., Peterson, J., 529 Morris, J., Hansen, M., and R. Smith, "Privacy 530 Considerations for Internet Protocols", RFC 6973, DOI 531 10.17487/RFC6973, July 2013, 532 . 534 [RFC7721] Cooper, A., Gont, F., and D. Thaler, "Security and Privacy 535 Considerations for IPv6 Address Generation Mechanisms", 536 RFC 7721, DOI 10.17487/RFC7721, March 2016, 537 . 539 [RFC7819] Jiang, S., Krishnan, S., and T. Mrugalski, "Privacy 540 Considerations for DHCP", RFC 7819, DOI 10.17487/RFC7819, 541 April 2016, . 543 [TRAC2016] 544 Faath, M., Weisshaar, F., and R. Winter, "How Broadcast 545 Data Reveals Your Identity and Social Graph", 7th 546 International Workshop on TRaffic Analysis and 547 Characterization IEEE TRAC 2016, September 2016. 549 Authors' Addresses 551 Rolf Winter 552 University of Applied Sciences Augsburg 553 Augsburg 554 DE 556 Email: rolf.winter@hs-augsburg.de 558 Michael Faath 559 Conntac GmbH 560 Augsburg 561 DE 563 Email: faath@conntac.net 564 Fabian Weisshaar 565 University of Applied Sciences Augsburg 566 Augsburg 567 DE 569 Email: fabian.weisshaar@hs-augsburg.de