4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung" An Architecture Concept for Mobile P2P File Sharing Services Andersen F.-U., Kappler C. (Siemens AG) Ivan Dedinski, Oberender J., de Meer H. (Universität Passau) Tutschku K., Hossfeld T., Mäder A., Tran-Gia P. (Universität Würzburg) University of Würzburg Department of Distributed Systems, Prof. Dr.- Ing. P. Tran-Gia University of Passau Institute of Computer Networks and Computer, Prof. Dr. H. de Meer SIEMENS AG I&C Mobile Networks Research & Concepts, F.-U. Andersen Dr. C. Kappler
Motivation (Trends and Assumptions) Trends Billing models for GPRS/UMTS: Towards time based and flat rate Rising or evolving amount of user created mobile content - Photos / small Videos of any (rights-free) kind - (self-composed) Ring tones, free text files / Information ISP network traffic is dominated by -to- file sharing applications - Telefonica reports up to 80% fixed network p2p traffic Assumptions DRM: very complex, inter-disciplinary issue (but ouf of scope) - Alternatives: DRM light / Creative Commons Licence model P2P experience is highly attractive to many users - Curiosity factor (what do other users have to offer?) 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung", 28.7.2004 Seite 2
Objectives of MoPi ( Mobile P2P ) Problems for P2P in 2.5/3G mobile networks P2P traffic should be kept local for fixed/wireless operators Asymmetric speed / P2P-unattractive charging Variable terminal connectivity High signalling traffic in unmodified P2P systems Operators wish to participate in service creation Qualitative + quantitative analysis of P2P in 2.5G / 3G CDF 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 0.2 0.1 with cache without cache 0 0 5 10 15 20 25 download time [h] Approach / options: Operator-managed P2P services: Operator controls service and traffic dynamics Improved P2P performance for mobile environments Realize in 3G by hybrid P2P structure Introduce infrastructure elements to address the above problems 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung", 28.7.2004 Seite 3
State of the art / related work Smart cache for KaZaA (Joltid Inc.) and other P2P systems man-in-the-middle approach. Protocol messages are intercepted and then tried to satisfy from a cache database Project JXTA for mobile devices (J2ME, JXME) mobile P2P messaging / lightweight JXTA APIs for mobiles relays enable connection to JXTA networks rendezvous s with full capabilities (message routing / proxy services) no consideration of cellular mobile requirements A platform and applications for mobile -to- communications (Kato et al., WWW2003, DoCoMo / Ericsson) mobiles cannot run full P2P client software proxy mixed architecture: hybrid, super + pure P2P part. multicast overlay carries out multicast forwarding by multihop or multidestination unicast (ALM). 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung", 28.7.2004 Seite 4
Option Space Cartography (I) Classical operator domain resource mediation centralized decentralized MBMS client/ server Cache IMS Proxy VPN static overlay Pure P2P e-donkey Classical user centric domain Hybrid domain centralized decentralized resource access 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung", 28.7.2004 Seite 5
Option Space Cartography (II) Shifting a pure P2P system towards the client-server model Classical operator domain resource mediation centralized decentralized client/ server Solution with e-donkey + Cache + Proxy Pure P2P e-donkey Classical user centric domain Hybrid domain centralized decentralized resource control 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung", 28.7.2004 Seite 6
Scope of work Initial basic idea Mobile P2P (MoPi) Decisions taken during project definition phase Cellular mobile Environment (2.5/3G) User services for sharing content Bluetooth, IR, WLAN, satellite, Network + control services, resilience 4G Legend: selected option evaluated but not selected option Decisions taken during WP1 project phase Fix-Mobile P2P Interconnect Hybrid P2P Edonkey + derivatives Mobile-to- Mobile P2P Pure P2P JXTA Optimization of PC-based P2P access via 3G Gnutella II, Fasttrack, WinMX, BitTorrent DoCoMo- Ericsson proposal 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung", 28.7.2004 Seite 7
Design decisions Option space Infrastructure options Algorithm options Legend: selected option evaluated but not selected option current work focus MBMS Distribution at GGSN level IMS / SIP Presence information P2P Cache Edonkey Index server Mod Caching strategies Advanced signalling Super organization SIP presence Transfer session IMS P2P application server Proxy P2P Crawler Search session Mobile P2P Architecture / System proposal 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung", 28.7.2004 Seite 8
Caching Peer (pat. pend.) sub-system P2P file exchange Common P2P / overlay signaling Caching specific signaling Mob. Mob. GGSN - Selected for final architecture; strong improvement of performance; implementation without changes in client software Fixed P2P network user domain Caching Index server operator domain Air traffic generated trigger level for cache # of downloads 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung", 28.7.2004 Seite 9
Crawling Peer (pat. pend.) sub-system P2P file exchange Common P2P / overlay signaling Mob. Mob. Caching specific signaling GGSN - Selected for final architecture; allows for integration of mobile s into the global P2P community Fixed P2P network Index server Crawling user domain operator domain 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung", 28.7.2004 Seite 10
Smart Proxy (pat. pend.) sub-system Legacy mobile signaling P2P file exchange Enhanced MoPi signaling Common P2P / overlay signaling -Some of its core functionality can be taken over by the Mob. GGSN caching and crawling s Mob. - Deselected for final architecture; Fixed P2P network user domain P2P proxy Firewall Indexserver operator domain has lots of interesting options for further research (e.g. queue sharing) 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung", 28.7.2004 Seite 11
Proposed Architecture Mobile Control Domain (Presence Information) Internet Index Server 2.5/3G mobile network Crawling Peer Cache Peer Internet Peers 0 Mobile Peers user domain operator domain P2P file exchange mediation signaling enhanced P2P signalling 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung", 28.7.2004 Seite 12
Simulation model and first results Requested file 3000 2500 2000 1500 1000 Popular file(s) CDF 1 0.9 0.8 0.7 0.6 0.5 0.4 0.3 500 Transferred data volume [kb] 0 0 5 10 15 20 25 Time [h] 3.5 x 104 3 2.5 2 1.5 1 0.5 downlink uplink 0 0 100 200 300 400 500 time [min] 0.2 0.1 with cache without cache 0 0 5 10 15 20 25 download time [h] Mean value without cache: 11.2915 h Mean value with cache: 5.8751 h Simulation model type: Queue-based,eventoriented (virtual time base) and implemented in Java. Event inter-arrival time is randomized. 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung", 28.7.2004 Seite 13
First results and conclusions (project is still running) The project shows that It is possible to use P2P with GPRS with limitations The limitations can be addressed by using the described infrastructure elements / architecture If the proposed architecture is used, it has the following advantages: More performant P2P data exchange via the 2.5/3G cellular air IF Increasing reliability of mobile-to-mobile file transfers Caching ( only 1 air interface needed) Based upon a fixed network P2P standard (edonkey) Connecting to fixed network P2P world in a controlled way Operator-provided index-server manages content mediation Redirection of clients to the operator s index server reducing interdomain traffic For the most parts, also applicable to fixed networks Motivation for end users to subscribe to a mobile P2P service because of a noticeable performance improvement 4. Würzburger Workshop "IP Netzmanagement, IP Netzplanung und Optimierung", 28.7.2004 Seite 14