LSec-REC：面向电商推荐服务的轻量化安全通信机制设计与性能分析

LSec-REC: Design and Performance Analysis of Lightweight Secure Communication Mechanism for E-Commerce Recommendation Services

摘要: 随着电子商务的快速发展，中小电商推荐服务平台的数据安全问题日益凸显。云端–边缘–端侧协同架构在提升推荐服务性能的同时，也扩大了数据攻击面，安全机制的引入必然带来额外的时间与算力开销，而中小电商算力有限，难以承受高安全开销。如何在保障数据安全与控制额外开销之间取得平衡，是该场景下的关键矛盾。针对电商推荐服务场景，本文设计了一种融合加密、传输安全与隐私保护的轻量化安全通信机制(简称：LSec-REC)。为量化该机制带来的额外开销，提出了包含时延、资源开销和安全-性能权衡的综合性能模型(简称：P3-Model)。仿真实验表明，LSec-REC机制在保障必要安全等级的前提下，能有效控制额外开销，加密与传输时延增加有限，CPU占用增幅较小，整体性能优于传统安全传输方案。本文提出的LSec-REC机制可为同类电商推荐服务平台的安全通信设计提供参考。

Abstract: With the rapid development of e-commerce, data security issues of recommendation service platforms for small and medium-sized e-commerce enterprises have become increasingly prominent. While the cloud-edge-end collaborative architecture improves the performance of recommendation services, it also expands the attack surface. The introduction of security mechanisms inevitably brings additional time and computing overhead, whereas small and medium-sized e-commerce enterprises have limited computing power and can hardly afford high security overhead. How to strike a balance between ensuring data security and controlling additional overhead is the key contradiction in this scenario. Targeting the e-commerce recommendation service scenario, this paper designs a lightweight secure communication mechanism integrating encryption, transmission security, and privacy protection (referred to as the LSec-REC). To quantify the additional overhead introduced by this mechanism, a comprehensive performance model incorporating latency, resource overhead, and security-performance trade-off (referred to as the P3-Model) is proposed. Simulation results show that the LSec-REC mechanism can effectively control additional overhead while ensuring the necessary security level, with limited increases in encryption and transmission latency and a small increase in CPU usage, demonstrating overall performance superior to traditional secure transmission schemes. The proposed LSec-REC mechanism can provide a reference for the secure communication design of similar e-commerce recommendation service platforms.