Resilient Edge Data Caching Using Uncertainty-Aware Optimization In Mobile Edge Computing

Abstract

Mobile Edge Computing (MEC) has emerged as a promising paradigm to reduce latency and improve data accessibility by caching frequently requested content closer to end users. However, existing edge data caching techniques primarily rely on static or popularity-based strategies, which fail to adapt effectively to dynamic user demands and do not account for uncertainties such as sudden demand fluctuations and edge server failures. These limitations can lead to increased latency, reduced data availability, and inefficient resource utilization.

To address these challenges, this project proposes an Uncertainty-aware Edge Data Caching (uEDC) framework designed for dynamic and unreliable edge environments. The system models the caching problem as a robust optimization task and introduces two algorithms: uEDC-B, which provides optimal caching decisions under worst-case conditions, and uEDC-L, a scalable approximation approach suitable for real-time applications. The framework continuously monitors user demand, network conditions, and server availability to make adaptive caching decisions.

In addition, the system incorporates security mechanisms such as data encryption and integrity verification to ensure safe and reliable data access. Experimental results show that the proposed approach significantly reduces data retrieval latency, improves caching efficiency, and enhances data availability even under uncertain conditions. The proposed framework provides a robust and scalable solution for intelligent edge data caching in modern distributed systems.