Adaptive In-Network Collaborative Caching for Enhanced Ensemble Deep Learning at Edge

To enhance the quality and speed of data processing protect privacy security data, edge computing has been extensively applied to support data-intensive intelligent services at edge. Among these services, ensemble learning-based can, in natural, leverage distributed computation storage resources devices achieve efficient collection, processing, analysis. Collaborative caching close source, order take limited high-performance learning solutions. this goal, we propose an adaptive in-network collaborative scheme for First, representation structure is proposed record cached among different nodes. In addition, design a collaboration facilitate nodes cache valuable local learning, by scheduling according summarization representations from Our extensive simulations demonstrate high performance scheme, which significantly reduces latency transmission overhead.