idnits 2.17.1 draft-kamei-p2p-experiments-japan-06.txt: Checking boilerplate required by RFC 5378 and the IETF Trust (see https://trustee.ietf.org/license-info): ---------------------------------------------------------------------------- ** You're using the IETF Trust Provisions' Section 6.b License Notice from 12 Sep 2009 rather than the newer Notice from 28 Dec 2009. (See https://trustee.ietf.org/license-info/) 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 (September 5, 2011) is 4589 days in the past. Is this intentional? Checking references for intended status: Informational ---------------------------------------------------------------------------- No issues found here. Summary: 1 error (**), 0 flaws (~~), 1 warning (==), 1 comment (--). Run idnits with the --verbose option for more detailed information about the items above. -------------------------------------------------------------------------------- 2 P2PRG S. Kamei 3 Internet-Draft NTT Corporation 4 Intended status: Informational T. Momose 5 Expires: March 8, 2012 Cisco Systems 6 T. Inoue 7 T. Nishitani 8 NTT Communications 9 September 5, 2011 11 ALTO-Like Activities and Experiments in P2P Network Experiment Council 12 draft-kamei-p2p-experiments-japan-06 14 Abstract 16 This document introduces experiments to clarify how ALTO-like 17 approach was effective to reduce network traffic made by a Council in 18 Japan to harmonize P2P technology with the infrastructure. And this 19 also provides some suggestions that might be useful for ALTO 20 architecture learned through our experiments. 22 Status of this Memo 24 This Internet-Draft is submitted to IETF in full conformance with the 25 provisions of BCP 78 and BCP 79. 27 Internet-Drafts are working documents of the Internet Engineering 28 Task Force (IETF), its areas, and its working groups. Note that 29 other groups may also distribute working documents as Internet- 30 Drafts. 32 Internet-Drafts are draft documents valid for a maximum of six months 33 and may be updated, replaced, or obsoleted by other documents at any 34 time. It is inappropriate to use Internet-Drafts as reference 35 material or to cite them other than as "work in progress." 37 The list of current Internet-Drafts can be accessed at 38 http://www.ietf.org/ietf/1id-abstracts.txt. 40 The list of Internet-Draft Shadow Directories can be accessed at 41 http://www.ietf.org/shadow.html. 43 This Internet-Draft will expire on March 8, 2012. 45 Copyright Notice 47 Copyright (c) 2011 IETF Trust and the persons identified as the 48 document authors. All rights reserved. 50 This document is subject to BCP 78 and the IETF Trust's Legal 51 Provisions Relating to IETF Documents 52 (http://trustee.ietf.org/license-info) in effect on the date of 53 publication of this document. Please review these documents 54 carefully, as they describe your rights and restrictions with respect 55 to this document. Code Components extracted from this document must 56 include Simplified BSD License text as described in Section 4.e of 57 the Trust Legal Provisions and are provided without warranty as 58 described in the BSD License. 60 Table of Contents 62 1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . 3 63 2. Background in Japan . . . . . . . . . . . . . . . . . . . . . 3 64 2.1. P2P traffic . . . . . . . . . . . . . . . . . . . . . . . 3 65 2.2. Impact on network infrastructure . . . . . . . . . . . . . 3 66 2.3. The object of P2P Network Experiment Council . . . . . . . 4 67 3. The details of the experiments . . . . . . . . . . . . . . . . 5 68 3.1. Dummy Node . . . . . . . . . . . . . . . . . . . . . . . . 5 69 4. Hint Server ('08) . . . . . . . . . . . . . . . . . . . . . . 6 70 5. High-Level Trial Results . . . . . . . . . . . . . . . . . . . 11 71 5.1. Peer Selection with P2P . . . . . . . . . . . . . . . . . 11 72 5.2. Peer Selection with the Hint Server . . . . . . . . . . . 11 73 6. Considerations . . . . . . . . . . . . . . . . . . . . . . . . 11 74 6.1. Next steps . . . . . . . . . . . . . . . . . . . . . . . . 12 75 6.2. Feedback to ALTO WG . . . . . . . . . . . . . . . . . . . 12 76 6.2.1. Hierarchical architecture for ALTO servers . . . . . . 12 77 6.2.2. Measurement mechanism . . . . . . . . . . . . . . . . 13 78 7. Security Considerations . . . . . . . . . . . . . . . . . . . 13 79 8. IANA Considerations . . . . . . . . . . . . . . . . . . . . . 13 80 9. Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . 14 81 10. Informative References . . . . . . . . . . . . . . . . . . . . 14 82 Authors' Addresses . . . . . . . . . . . . . . . . . . . . . . . . 14 84 1. Introduction 86 An overlay network, which is used by P2P and other applications, 87 offers the advantage of allowing flexible provision of services while 88 hiding the lower layer network. The downside is that inefficient 89 routes are often taken in the lower IP network, thereby increasing 90 the network load. Several proposals have been made to build an 91 overlay network that takes account of the information about the lower 92 layer network. Since the management of the Internet is highly 93 distributed, it is difficult to implement such proposals and thus 94 optimize a network without the cooperation of network providers. 96 Recently, the controversy between the overlay network and the network 97 providers have been rekindled. Under these circumstances, some 98 researchers have studied overlay network control technology that 99 takes account of the network topology information obtained from 100 network providers. 102 One of the activities concerning this issue has been made by the P2P 103 Network Experiment Council in Japan. This document reports on the 104 issues addressed and experiments being made by the council, focusing 105 on the experiments made from 2007 to 2008. 107 2. Background in Japan 109 2.1. P2P traffic 111 As of 2008, the world most popular P2P file sharing application 112 Bittorrent isn't widely deployed in Japan. Instead, other Japan 113 specific file sharing P2P applications such as Winny [1], Share [2], 114 PerfectDark, and so on, still occupy 40% of the Internet traffic in 115 Japan even though many those P2P users were arrested for sharing 116 illegal files with these P2P apps. 118 Each P2P file sharing application has their original protocol. 119 Therefore, it is more difficult to control one by one unlike 120 Bittorrent. 122 2.2. Impact on network infrastructure 124 One of the advantage of using P2P technology for content delivery is 125 that peers exchange content directly among themselves. This reduces 126 the load on servers. Also, P2P applications can reduce upstream 127 traffic from an original content server. This is significant that 128 the charge for upstream traffic is usually traffic-sensitive for 129 content delivery services, and it is not negligible. 131 It is also known that server cost could be reduced with P2P 132 technology. However, the story is quite different for network 133 providers. From the viewpoint of network providers, the traffic that 134 content servers generate has shifted to the edge network and the 135 amount of traffic has not necessarily been reduced. Another problem 136 for network providers that an extremely inefficient routing may be 137 selected has been raised. It is because overlay network systems are 138 configured without any regard to the structure of the lower layer 139 network or network geometry. 141 In some cases, traffic on the Internet used to be limited by the 142 capacity of servers. For those cases, the improvement in the 143 scalability of servers has made it likely that network resources will 144 be used up before server resources are. Using P2P applications 145 increases the volume of traffic per user remarkably. 147 Faced with increase in the load on network infrastructure, network 148 providers are compelled to take actions to overcome the sudden 149 increase in facilities' cost. Representative actions include placing 150 content in IXs or data centers, introducing bandwidth control, and 151 raising the access fees[3]. 153 In the future, video posting sites, which has been delivered using 154 client-server applications, may adopt P2P system. The increase in 155 traffic arising from such a shift will be a great threat to the 156 network. 158 2.3. The object of P2P Network Experiment Council 160 In order to reduce Internet traffic and encourage legitimate use of 161 P2P technologies, the Japanese government led to establish a new 162 council called P2P Network Experiment Council conjunction with 163 commercial P2P application vendors and ISPs in 2006. 165 Then the council had started to develop regulations that include 166 several guidelines such like an advance notice to restrict bandwidth 167 to heavy traffic users. In accordance with the regulations, some 168 ISPs introduced solutions that reduce traffic caused by P2P file 169 sharing applications . 171 Besides this activity, the council also looked for new ways to avoid 172 traffic conjunction by commercial P2P applications with ISPs, 173 carriers, contents providers and P2P system vendors. In this work, 174 the council had experiments that introduced ALTO-like system and 175 observed how the traffic was reduced by redirecting to proper peers 176 on the real Internet in Japan. 178 This memo describes the overview of the experiments. 180 3. The details of the experiments 182 The council has already learned that the server cost could be reduced 183 with using P2P technology for contents delivering by investigating 184 data offered by the members of the council. For example, the data 185 brought by the vendors shows as follows: 187 90% of traffic was reduced with UG Live by Utagoe Inc[4]. 189 The costs of delivering to tens of thousand subscribers was 190 reduced to 1/5 with BBbroadcast with TV Bank Corp.[5] 192 On the other hand, these reduced server costs may affect network 193 load. One of the goals of our experiments are to visualize the 194 impacts and propose an architecture to reduce network load caused by 195 these new technologies. 197 To satisfy the above goals, the framework to be proposed should be 198 well generalized as possible that doesn't rely specific P2P 199 application behaviors because multi P2P application vendors join 200 these experiments. In addition, the traffic should be captured 201 beyond multi ISPs. 203 3.1. Dummy Node 205 As mentioned before, while the effect of delivery using P2P 206 technology on reducing the traffic and the load on servers is well 207 known, traffic behavior in the inter-ISP is not known. In Japan, 208 there is a backbone traffic report cooperated with ISPs and IXes [6]. 209 However, this measurement requires to capture packets on subscribers 210 line to know end user's activity. It is not realistic to measure the 211 behavior of P2P applications at user terminals connected to the 212 Internet because that would require a large-scale arrangement for 213 measurement, such as using Deep Packet Inspection (DPI) on aggregated 214 lines. 216 To solve these problems, we put several nodes called 'dummy nodes' in 217 the ISP's networks. The dummy nodes emulate an end user's PC and P2P 218 applications are running on the nodes. 220 By introducing dummy nodes, we can observe and evaluate how much P2P 221 applications have affected networks by measuring the traffic on dummy 222 nodes. Since this method can't measure every subscriber's traffic, 223 the accuracy would be less than other methods. But this make it 224 possible to adapt to situations many different P2P applications 225 coexist on a network. We can say this is suitable for these 226 experiments. 228 A dummy node consists of Intel PC server, Linux(CentOS), VMWare and 229 Windows XP works on VMWare. With this configuration, all packets can 230 be captured without any impacts to the network, nodes and application 231 behaviors. And it enable us to use different P2P applications for 232 windows and evaluate them generally. 234 To see behaviors of the node, incoming and outgoing packets are 235 captured on Linux because every packets are transmitted through it. 236 In these experiments, we captured source/destination address, port 237 number, amount of traffic and start/end time to see flow information. 239 60 Dummy nodes are put on access networks that are closest subscriber 240 as possible in different 40 networks. 242 +----------------------+ 243 |+--------------------+| 244 ||+------------------+|| 245 ||| P2P Application ||| 246 ||| WindowsXP ||| 247 ||| +--+ ||| 248 ||+--------|N |------+|| 249 || VMware |e | || 250 |+---------|t |-------+| 251 | Linux |IF| capture| 252 +----------| |--------+ 253 +--+ 255 Dummy nodes 257 Figure 1 259 4. Hint Server ('08) 261 In Japan, bottleneck in IP networks have been shifting from access 262 networks to backbone networks and equipments, such as bandwidth 263 between ISPs and capacity in IXs, since FTTH has rapidly spread all 264 over Japan. Under these circumstances, the Council proposed a less 265 restrictive and more flexible cooperation between ISPs than existent 266 P4P experiments [7]. The proposed method consists of the following 267 elements: (1) P2P clients, (2) P2P control servers, and (3) a hint 268 server: a peer selection hint server. (1) and (2) are existing 269 systems but whether (2) exists depends on each application. (3) is a 270 server that provides a hint as to the selection of a peer, and plays 271 a role equivalent to that of ALTO Server. Note that this proposal 272 was based on results of experiments using dummy nodes. The results 273 showed that it was possible to reduce unnecessary traffic that flows 274 across the boundaries of geographical districts or ISPs through 275 providing information about the physical network to P2P applications. 277 When a peer joins the network, it registers its location information 278 (IP address) and supplementary information (line speed, etc.) with 279 the hint server. The hint server makes a mapping of the new peer 280 (P2P client) based on network topology information obtained from the 281 ISP, generates a routing table in which peers are listed in the order 282 of priority for selection, and returns the table to the peer. 284 If all information can be made publicly, the above procedure can 285 produce a result which is close to overall optimization. However, 286 some information held by ISPs can often be confidential. Besides, in 287 some cases, the volume of calculation required to process all 288 information can be excessive. To avoid these problems, it is planned 289 to conduct experiments with a limited set of functions, analyze 290 experiments results, and gradually expand the scope of optimization. 292 A control mechanism that makes use of all possible information is 293 difficult not only technically but also difficulties to achieve 294 coordination among providers. In consideration of these 295 difficulties, the council has been limiting the implementation and 296 experiments to the following scope since 2006. 298 Figure 2 shows an outline of the hint server. 300 +---------+ GetLocation +-------------GeoIP DB Server---------+ 301 | | +-----------+ | +----------+ +-----------+ | 302 | |--|IP Address |-->| | GeoIP DB | |Quagga etc | | 303 | | +-----------+ | +----------+ +-----------+ | 304 | | | +-------------+ +----------------+ | 305 | | +-----------+ | | District | | Routing | | 306 | |--|AS Code: |---| | information | |information(DGP)| | 307 | | |Regional | | | | | | | 308 |P2P Peers| |Information| | | Range of | |AS Code(origin) | | 309 | or | +-----------+ | | IP address | | | | 310 | Contro| | | +-------------+ +----------------+ | 311 | Server | +-------------------------------------+ 312 | | | ^ 313 | | PeerSelection v | 314 | | +-----------+ +--------------------------------------+ 315 | |--|IP Address |-->| +--Priority Node Selection System--+ | 316 | | | List | | | | | 317 | | +-----------+ | | Peer candidate ranking | | 318 | | +-----------+ | | | | 319 | |--| Ranking |-->| +----------------------------------+ | 320 | | +-----------+ +--------------------------------------+ 321 +---------+ 323 Peer selection hint server 325 Figure 2 327 The network information used by the hint server is not information 328 solicited from individual ISPs but the AS number and district 329 information, which are more or less already public. Routing tables 330 are not generated. Instead, peers within the same ISP or the same 331 district are selected with higher priority in order to confine 332 traffic to within the same ISP or the same district. 334 When the hint server receives an IP address, it returns its attribute 335 information, to achieve the above. A peer can select a peer based on 336 the returned information. This operation is called GetLocation. 337 However, in preparation for the time when it becomes necessary to 338 hide topology information, an interface is provided through which a 339 priority order is returned in response to an input of a list of 340 candidate peers. This operation is called PeerSelection. 342 Although the priority node is selected based on the criterion that it 343 is within the same ISP or the same district, this type of selection 344 is not very effective if the number of participating peers is small. 345 Table 1 shows ratio of peers within the same AS or the same 346 prefecture calculated from the distribution of ASs and prefectures in 347 the IP address space from one-day data on a Winny network. 349 +--------------------+--------+ 350 | Conditions | ratio | 351 +--------------------+--------+ 352 | AS matches | 6.70% | 353 | Prefecture matches | 12.76% | 354 | Both match | 2.09% | 355 | Neither match | 78.45% | 356 +--------------------+--------+ 358 Table 1: AS and prefecture distributions 360 Since, in addition to the above, the presence/absence of content 361 affects the result, the control of selecting a peer within the same 362 district may be inadequate. Therefore, it is necessary to introduce 363 the weight of a continuous quantity that reflects the physical 364 distance or the AS path length as an indicator of the proximity of 365 the areas involved. 367 In consideration of the above, the following two measures are used 368 for the evaluation of proximity between peers in a hint server. 370 o AS path length (distance between ISPs) 372 AS path length calculated from BGP full routes. Since a full 373 routing table retrieved at an ISP can show only a best path, it 374 may not get an accurate length if the AS hop of both ISPs is too 375 large. To avoid this, we use multiple BGP information gotten at 376 different ISPs and combine them. Based on this concept, we used 377 BGP routing information's offered by three ISPs operated by big 378 telecommunication couriers and made a topology tree. Then it 379 enables to calculate the shortest path between given two ASes. 381 o Geographical distance 383 Distances between peers are measured using physical distance of 384 prefectural capitals that target peers belong to. The distance 385 between prefectural capitals is used to calculate physical 386 distance. Distances between prefectural capitals are sorted into 387 ascending order, and then into bands, with weights 1 to 15 388 assigned to them so that there are a more or less equal number of 389 "capital pairs" in each band. If either of their location is 390 indefinite, distance is equal to 15 and, if they are in the same 391 prefecture, distance is equal to 0. 393 Evaluation of distances between peers showed that the distribution 394 of distances was almost uniform when distances between peers are 395 normalized. This result suggests that using normalized distances 396 expands the area where the control by a Hint Server is effective. 398 An example of the request and the response 400 o Request 402 POST /PeerSelection HTTP/1.1 403 Host: ServerName 404 User-Agent: ClientName 405 Content-Type: text/plain; charset=utf-8 407 v=Version number 408 [application=Application identifier] 409 ip=IP address of physical interface 410 port=Port number of physical interface 411 [nat={no|upnp|unknown}] 412 [nat_ip=Global IP address using UPnP] 413 [nat_port= Global port number using UPnP] 414 [trans_id=transcation ID] 415 [pt=Flag of port type] 416 [ub=upload bandwidth] 417 [db=download bandwidth] 419 o Response 421 HTTP/1.1 200 OK 422 Date: Timestamp 423 Content-Type: text/plain; charset=utf-8 424 Cache-control: max-age=max age 425 Connection: close 427 v=Version number 428 ttl=ttl 429 server=hint server name 430 ... 431 trans_id=transaction ID 432 pt=Flag of port type 433 client_ip=Peer IP address observed from server 434 client_port=Peer port number observed from server 435 numpeers=number of respond peer 436 n=[src address] dst address / cost / option 438 5. High-Level Trial Results 440 5.1. Peer Selection with P2P 442 Table 2 shows the result of the analysis of communication in a node 443 of an ISP installed in Tokyo, as an example of measurement results. 445 +-----------------------------------------+------------+------------+ 446 | Conditions | Experiment | Experiment | 447 | | 1 | 2 | 448 +-----------------------------------------+------------+------------+ 449 | *Peers selected within the same ISP | 22% | 29% | 450 | *Peers selected within the same | 19% | 23% | 451 | district | | | 452 | *Peers selected within the same | 5% | 7% | 453 | district and the same ISP | | | 454 +-----------------------------------------+------------+------------+ 456 Table 2: Percentage of communication within the same ISP 458 The table shows that the probability of communication with peers in 459 the same ISP is proportional to the number of population and the 460 share of the ISP in each district. The data show that peers were 461 selected at random. Note that the vendor of a P2P application used 462 in these experiments explained that the mechanism of selection a peer 463 using network information can be implemented. However, peer 464 selection is normally based on past information because users often 465 cannot actually perceive the effect of using network information. 467 5.2. Peer Selection with the Hint Server 469 Since the main objective of these experiments was to verify the 470 operations of the hint server and P2P applications, the degree to 471 which traffic in the network was actually reduced was not evaluated. 472 However, the distances between a dummy node and a peer were obtained 473 from data on the dummy nodes. An examination of the distances 474 between a dummy node and a peer revealed that mean value of distance 475 after the hint server was introduced was reduced by 10% and that 95% 476 value of that was reduced by 5%. 478 6. Considerations 480 We clarified followings throughout our experiments. 482 1. Dispersed dummy nodes can figure out the behavior of peers and 483 traffic between inter-ISP networks, which peers are selected by 484 each peer. Therefore it proves that the importance of peer 485 selection control mechanism proposed in ALTO. 487 2. Using our peer selection control mechanism, called hint server, 488 could achieve significant differences. Our hint server can lead 489 each peer to select nearer peer. 491 In the experimental result of peer selection control, it is smaller 492 in intra-ISP traffic than other experiments[8] We think that it is 493 because there are smaller peers in each area of traffic control. 494 When there are many peers in one ISP, it is easy to select peers in 495 the same ISP. However, when there are small peers in one ISP, it is 496 difficult to select peers in the same ISP. In the situation of our 497 experiments, there are many ISPs of peers belonging, and there are 498 relatively smaller peers exist in same ISP. 500 Moreover, we didn't force P2P vendors to limit their implementation 501 policy, therefore we can observe differences how each implementations 502 weigh the information from the hint servers. Especially, in tree 503 overlay topology P2P applications, such mechanism is very effective, 504 on the other hand, in mesh overlay system, less effective. 506 6.1. Next steps 508 The experiments are on going as of 2011. Current experiments in 509 2011, we've changed the communication protocol to hint servers to 510 ALTO based because it is nearly standardized. In our implementation, 511 PIDs and the value of cost are mapped to ISP subnets, and ISP 512 distance respectively. We also implement services for compatibility 513 required by ALTO such as Service Capability and Map Services. But 514 the Endpoint Cost Service is mainly used because of backward 515 compatibility of our experiments. 517 We also study hierarchical hint server structure, in order to control 518 in coarse inter-ISPs and in detail intra-ISP. It is also effective 519 for limiting the area of information disclose. 521 6.2. Feedback to ALTO WG 523 This section describes what the authors learned with these 524 experiments would be useful for the ALTO WG. 526 6.2.1. Hierarchical architecture for ALTO servers 528 In our experiments, we present the possibility of traffic control 529 among multi-ISPs and multi-P2P applications using ALTO mechanism. On 530 the other hand, we found several problems in ISP operations to adapt 531 the mechanism. One is the granularity of network information. Among 532 inter-ISP area, it is relatively easy to treat information for public 533 purpose using BGP full route. On the other hand, among intra-ISP 534 area, it may be difficult to disclose private information of each 535 ISP. [9] propose some modification for ALTO protocol in order to hide 536 ISP information. We propose hierarchical structures. From the 537 viewpoint of cooperation between ISPs, fine-grained information is 538 not necessarily required and moreover it is difficult to exchange the 539 fine-grained information between ISPs. Considering this situation, 540 the authors use only coarse-grained information to control backbone 541 traffic in the experiments this year, though demand of controlling 542 traffic within an ISP using fine-grained information will arise in 543 the near future. Therefore it led us that introducing hierarchical 544 structure into ALTO is necessary to cope with both situations. 545 Actually, the authors plan to adapt a hierarchical control mechanism 546 in the next steps, which include the following two steps. 548 o In the first step, coarse-grained information about whole the 549 network is used to select ISPs. 551 o Next, fine-grained information within the ISP is used to select a 552 peer. 554 6.2.2. Measurement mechanism 556 In the experiments, there were two difficulties as follows: 558 o Evaluating effect of introducing a hint server was difficult, 559 since P2P applications had their own measurement mechanisms. 561 o How to treat priority orders of peers suggested by a hint server 562 could not be predetermined for P2P applications. 564 From these experiences, the authors consider that clarifying 565 requirements about measurement mechanisms for P2P applications are 566 necessary also in ALTO. 568 7. Security Considerations 570 This document does not propose any kind of protocol, practice or 571 standard. 573 8. IANA Considerations 575 No need to describe any request regarding number assignment. 577 9. Acknowledgments 579 Thanks to strong support by MIC (Ministry of Internal Affairs and 580 Communications of Japanese government), the council was established. 581 These experiments were performed under cooperation among P2P Network 582 Experiment Council members, and DREAMBOAT co.,ltd., Bitmedia Inc., 583 Utagoe. Inc. and Toyama IX have especially supported analyses of the 584 experiments. The authors appreciate Tohru Asami, Hiroshi Esaki and 585 Tatsuya Yamshita for their constructive comments. 587 10. Informative References 589 [1] "Winny on Wikipedia", . 591 [2] "Share on Wikipedia", 592 . 594 [3] Taniwaki, "Broadband Competition Policy in Japan", 2008, 595 . 597 [4] Utagoe Inc., "UGLive technology introduction", 598 http://www.utagoe.com/en/technology/grid/live/index.html, March, 599 2011. 601 [5] TVBank, "Live Delivery using `BB Broadcast'Achieving 96% Saving 602 in Traffic!", http:.wwww.tv-bank.com/jp/20081031.html, 2008 (in 603 Japanese). 605 [6] Cho, Fukuda, Esaki, and Kato, "The Impact and Implications of 606 the Growth in Residential User-to-User Traffic", SIGCOMM2006, 607 pp207-218, Pisa, Italy, September 2006. 609 [7] Open P4P, "P4P Field Tests: Yale-Pando-Verizon", 610 http://www.openp4p.net/front/, 2009. 612 [8] "RFC5632: Comcast's ISP Experiences in a Proactive Network 613 Provider Participation for P2P (P4P) Technical Trial", September 614 2009. 616 [9] "ALTO H12,draft-kiesel-alto-h12-02 (work in progress)", March 617 2010. 619 Authors' Addresses 621 Satoshi Kamei 622 NTT Service Integration Laboratories 623 3-9-11, Midori-cho 624 Musashino-shi, Tokyo 180-8585 625 JP 627 Phone: +81-422-59-6942 628 Email: kamei.satoshi@lab.ntt.co.jp 630 Tsuyoshi Momose 631 Cisco Systems G.K. 632 2-1-1 Nishi-Shinjuku 633 Shinjuku-ku, Tokyo 163-0409 634 JP 636 Phone: +81-3-5324-4154 637 Email: tmomose@cisco.com 639 Takeshi Inoue 640 NTT Communications 641 3-4-1, Shibaura 642 Minato-ku, Tokyo 108-8118 643 JP 645 Phone: +81-3-6733-7177 646 Email: inoue@jp.ntt.net 648 Tomohiro Nishitani 649 NTT Communications 650 1-1-6, Uchisaiwaicho 651 Chiyodaku, Tokyo 100-8019 652 JP 654 Phone: +81-50-3812-4742 655 Email: tomohiro.nishitani@ntt.com