The Decrease of Freeriding Effect to the Resource Providers in Peer-to-Peer File Sharing System with Social Network Information

Abstract

Freeriders only use system resources without sharing any resources to others or sharing re-sources which are not used by the system. This behavior is one significant problem that needs to be considered in improving the quality of P2P file sharing system. Moreover, it can lead the system to lack of quality resources, and hence, its performance will gradually decrease and further possibly collapse. Many studies raised concern on how social network information can reduce this behavior and improve the system performance. This research applies social network information by using a SocialGnutella algorithm to P2P file sharing system. The algorithm is simulated to observe how the system used to numerous peers in the contact list. This simulation used 512 and 1024 peers with each peer having five transactions and 30 repetitions in merge. Available bandwidth and utility of resource provider used to measure the system performance during the simulation. From the simulation, average utility of resource providers in the system using social network information reached 0.959 and 0.979 for 512 and 1024, respectively. Whereas in the system which is not used social network information reached 0.527 and 0.531. Average utility is significantly higher based on the t-test result at 99.9% confidence level. Hence, system which is used social network information can decrease the freeriding behaviour and directly impact the resource provider.