Скачать презентацию Self-Organized Aggregation of Wi-Fi Networks Elias C Efstathiou Скачать презентацию Self-Organized Aggregation of Wi-Fi Networks Elias C Efstathiou

8bc6baecf68f11cdf85dcb89d66359a0.ppt

  • Количество слайдов: 1

Self-Organized Aggregation of Wi-Fi Networks Elias C. Efstathiou, Advisor: George C. Polyzos Mobile Multimedia Self-Organized Aggregation of Wi-Fi Networks Elias C. Efstathiou, Advisor: George C. Polyzos Mobile Multimedia Laboratory, Department of Computer Science Athens University of Economics and Business Abstract: We present a peer-to-peer (P 2 P) scheme that could fuel the deployment of free public wireless networks in cities; we call it P 2 P Wireless Network Confederation (P 2 PWNC). Unlike existing approaches, P 2 PWNC does not rely on central planning. In P 2 PWNC, wireless service is provided to those who provide service to others based on an algorithm that runs in isolation on every peer. 1 5 Motivation Incomplete Information • Teams have partial view of receipt graph • Private Wi-Fi networks, connected to the Internet, cover major metropolitan areas • Teams do not cooperate in a distributed storage scheme and there is no central repository of receipts • We could use one another’s networks to access the Internet • Must rely on members to disseminate receipts: – members carry a subset of the receipts their team knows about – before requesting service, members transmit these receipts to Plus, Wi-Fi phones are now available the prospective provider Motorola CN 620 2 6 Evolutionary Learning Framework • Assume cost c when providing a unit of service and The Problem • How to encourage Wi-Fi connection sharing? benefit b when receiving one • For various b/c ratios (>1): – “Give to those who give” • – – – 3 – let each team follow one of various predefined sharing strategies – let teams interact at random – allow teams to learn, i. e. switch to better-performing strategies Design a reciprocity scheme that is fully self organized collusion proof easy to join When do cooperative strategies outperform non-cooperative ones? Example of a conditionally cooperative strategy from [2] (black) outperforming “never share” (blue) and “always share” (red) 7 Solution Architecture • Participants form small teams, typical teams = households • Team IDs are uncertified, no need for authorities • Teams operate one or more Wi-Fi access points (APs) 4 • Simple text-based protocol [3]: • – Blue team – – Green team – : Wi-Fi access point : team member Wi-Fi network view Team view member PK team PK RCPT: Client transmits a stored or a newly signed Y 5 consult receipt graph: UPDT: Client requests latest receipts from own team Summary, Contact, References Z G 6 timestamp – We design, simulate, and implement [2, 3, 4, 5] a self-managed solution for Wi-Fi aggregation, which can be deployed spontaneously X B signed by team SK consumer cert. Additional information: the Peer-to-Peer Wireless Network Confederation section at http: //mm. aueb. gr/research/ and via e-mail to [email protected] gr (Elias Efstathiou) • receipt graph contains the community’s history volume of traffic signed by consumer SK 4 Decision function: ? Keys and certificates based on Elliptic Curve Cryptography (Receipt size: 150 bytes) RREQ: AP requests an updated receipt for the current 8 W providing team PK • receipt 1 All IDs = unique public/private key (PK/SK) pairs 3 Receipt: CACK: Positive/negative response from AP AP code runs on the Linux-based Linksys WRT 54 GS AP session (thus splitting consumer/provider risk) Receipt Graph, When to Cooperate? 2 Member certificate: CONN: Client requests service from access point (AP) (request includes member certificate) White team • Team members sign receipts whenever they access another team’s APs Protocol, Implementation Green team AP: “Should I cooperate with the member from the Blue team who is requesting service? ” • graph arrows encode the owes to relationship • we study collusion-proof reciprocity strategies [1] that allow the evolution of cooperation White team PK Green member cert timestamp Timestamp volume of traffic Weight Signed by Green member SK [1] Feldman, Lai, Stoica, Chuang, Robust Incentive Techniques for Peer-to-Peer Networks, Proc. ACM EC’ 04. [2] Efstathiou, Polyzos, Self-Organized Peering of Wireless LAN Hotspots, European Transactions on Telecommunications (special issue on Self-Organization in Mobile Networking, in press). [3] Efstathiou, Frangoudis, Polyzos, in preparation. [4] Antoniadis, Courcoubetis, Efstathiou, Polyzos, Strulo, Peer-to-Peer Wireless LAN Consortia: Economic Modelling and Architecture, Proc. IEEE P 2 P’ 03. [5] Efstathiou, Polyzos, A Peer-to-Peer Approach to Wireless LAN Roaming, Proc. ACM Mobi. Com WMASH’ 03.