Research on a Lightweight Full-Stack Edge Execution Optimization Framework Based on Serverless and WebAssembly

This paper proposes a lightweight full-stack execution framework integrating Serverless architecture with WebAssembly runtime optimization to enhance performance and energy efficiency in edge deployments. The system employs modular task decomposition and Light-Container Isolation (LCI) technology to achieve cross-node function reuse on AWS Lambda and Cloudflare Workers platforms. An Reinforcement Learning Scheduler (RL-Scheduler) predicts request distribution in real-time, dynamically allocating CPU cycles and memory limits. Targeted testing demonstrates a 52% reduction in cold start time, a 33% decrease in average execution latency, and a 21% reduction in energy consumption under 3,000 concurrent tasks. Results confirm the framework effectively enhances execution autonomy and cross-platform portability for edge Serverless systems in multi-tenant environments.