TY - JOUR
T1 - Evolution of resource reciprocation strategies in P2P networks
AU - Park, Hyunggon
AU - Van Der Schaar, Mihaela
N1 - Funding Information:
Manuscript received April 05, 2009; accepted September 12, 2009. First published October 13, 2009; current version published February 10, 2010. This work was supported by the National Science Foundation (NSF) under CCF 0830556 and CCF 0541867. The material in this paper was presented in part at the 34th IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP), Taipei, Taiwan, April 2009. The associate editor coordinating the review of this manuscript and approving it for publication was Dr. Z. Jane Wang.
PY - 2010/3
Y1 - 2010/3
N2 - In this paper, we consider the resource reciprocation among self-interested peers in peer-to-peer (P2P) networks, which is modeled as a stochastic game. Peers play the game by determining their optimal strategies for resource distributions using a Markov decision process (MDP) framework. The optimal strategies enable the peers to maximize their long-term utility. Unlike in conventional MDP frameworks, we consider heterogeneous peers that have different and limited ability to characterize their resource reciprocation with other peers. This is due to the large complexity requirements associated with their decision making processes. We analytically investigate these tradeoffs and show how to determine the optimal number of state descriptions, which maximizes each peer's average cumulative download rates given a limited time for computing the optimal strategies. We also investigate how the resource reciprocation evolves over time as peers adapt their reciprocation strategies by changing the number of state descriptions. Then, we study how resulting download rates affect their performance as well as that of the other peers with which they interact. Our simulation results quantify the tradeoffs between the number of state descriptions and the resulting utility. We also show that evolving resource reciprocation can improve the performance of peers which are simultaneously refining their state descriptions.
AB - In this paper, we consider the resource reciprocation among self-interested peers in peer-to-peer (P2P) networks, which is modeled as a stochastic game. Peers play the game by determining their optimal strategies for resource distributions using a Markov decision process (MDP) framework. The optimal strategies enable the peers to maximize their long-term utility. Unlike in conventional MDP frameworks, we consider heterogeneous peers that have different and limited ability to characterize their resource reciprocation with other peers. This is due to the large complexity requirements associated with their decision making processes. We analytically investigate these tradeoffs and show how to determine the optimal number of state descriptions, which maximizes each peer's average cumulative download rates given a limited time for computing the optimal strategies. We also investigate how the resource reciprocation evolves over time as peers adapt their reciprocation strategies by changing the number of state descriptions. Then, we study how resulting download rates affect their performance as well as that of the other peers with which they interact. Our simulation results quantify the tradeoffs between the number of state descriptions and the resulting utility. We also show that evolving resource reciprocation can improve the performance of peers which are simultaneously refining their state descriptions.
KW - Evolution of resource reciprocation
KW - Markov decision process (MDP)
KW - Peer-to-peer (P2P) network
KW - Resource reciprocation
KW - Stochastic game
UR - http://www.scopus.com/inward/record.url?scp=79955625536&partnerID=8YFLogxK
U2 - 10.1109/TSP.2009.2033731
DO - 10.1109/TSP.2009.2033731
M3 - Article
AN - SCOPUS:79955625536
SN - 1053-587X
VL - 58
SP - 1205
EP - 1218
JO - IEEE Transactions on Signal Processing
JF - IEEE Transactions on Signal Processing
IS - 3 PART 1
ER -