Congestion aware overlay network

Autor: Md. Moinuddin Bhuiyan
Rok vydání: 2021
Předmět:
DOI: 10.32920/ryerson.14655078.v1
Popis: Advances in wireless technology and mobile computing have provided a major impetus towards development of P2P network over MANET. Both peer-to-peer overlays and mobile ad-hoc network share the key characteristics of self-organizing and decentralization; both need to solve same fundamental problem, that is, how to route queries or packets efficiently. Congestion affecting the performance of P2P overlay on wireless ad-hoc network occurs not only due to buffer overflow but also due to error prone nature of the wireless medium used and the mobility of users. When paths of multiple downloading sessions intersect with each other, they mutually cause congestion for each other that result in packet drops. We identified that P2P MANET suffers query performance primarily due to significant congestion during download session. The unawareness of congestion in P2P MANET system may lead to the longer user response time, high overhead and many packet losses. In order to alleviate the congestion problem, we propose a modification to RAON, CANON, which includes congestion awareness for improving the RAON performance. We evaluated our system through simulation, which shows that CAON improves the performance of query success rate, query delay, packet losses significantly. We also evaluated the impact of mobility model on our system by simulating under both Random way point mobility model and Manhattan mobility model.
Databáze: OpenAIRE