RIPPLE: A DECENTRALIZED EDGE-BASED DATA DEDUPLICATION FRAMEWORK

Abstract

With its advantages in ensuring low data retrieval latency and reducing backhaul network traffic, edge computing is becoming a backbone solution for many latency-sensitive appli cations. An increasingly large number of data is being generated at the edge, stretching the limited capacity of edge storage systems. Improving resource utilization for edge storage systems has become a significant challenge in recent years. Existing solutions attempt to achieve this goal through data placement optimization, data partitioning, data sharing, etc. These approaches overlook the data redundancy in edge storage systems, which produces substantial storage resource wastage. This motivates the need for an approach for data deduplication at the edge. However, existing data deduplication methods rely on centralized control, which is not always feasible in practical edge computing environments. Ripple is a novel framework that enables edge servers to perform data deduplication in a fully decentralized manner. Unlike traditional approaches relying on centralized coordination, Ripple establishes a local data index on each edge server, empowering them to independently identify and eliminate redundant data. This decentralized design allows Ripple to: 1. efficiently detect and remove duplicate data, 2. uphold low-latency data retrieval requirements, and (removing Unwanted Data) 3. ensure data availability after deduplication. Extensive trace-driven experiments on a realworld testbed validate the effectiveness of Ripple. Compared to existing state-of-the-art techniques, Ripple achieves a 60.42% reduction in data retrieval latency and enhances the deduplication ratio by up to 16.79%, demonstrating its practical advantages in edge computing environments.