Efficiency Enhancement Architecture of SQL Queries Through Hierarchical Cache

Abstract

Secure Multi-Party Computation (MPC) enables privacy-preserving collaborative data analysis but suffers from high computational overhead, communication latency, and inefficient query execution. Traditional caching mechanisms, such as SMPCache, attempt to optimize performance but face limitations due to high storage overhead, low cache hit rates, and inefficient cache lookup strategies. This study suggests a Hierarchical SMPCache (H-SMPCache) The approach to address these issues, inspired by CPU caching, introduces a multi-level caching structure (L1, L2, L3) to efficiently manage SQL queries in MPC environments. The proposed hierarchical caching model aims to classify queries based on execution cost, frequency, and computational complexity.