Optimizing Traffic and Latency in Peer-to-Peer Networks Through Advanced Replication and Polling Algorithms

Abstract

Peer-to-peer (P2P) networks have gained prominence as scalable platforms for large file distribution, but face challenges in managing network traffic and latency. This paper proposes enhancements to the Integrated File Replication and Consistency Maintenance (IRM) algorithm to reduce polling traffic, ensure consistency, and improve response handling in Chord-based systems. Our novel polling strategies—Greedy, Lazy, and Intercept Polling—enable replicas to serve requests independently. Simulations using Amazon EC2 demonstrate a 50–60% drop in polling traffic and a 30% reduction in latency, while maintaining 99% data consistency. These results validate the robustness of our method, particularly under churn-heavy scenarios.