Privacy-Aware Multi-Device Cooperative Edge Inference with Distributed Resource Bidding

TL;DR

This paper presents a privacy-aware multi-device cooperative edge inference system that uses distributed resource bidding and feature compression, optimizing classification accuracy while protecting data privacy in mobile edge computing environments.

Contribution

It introduces a decentralized bidding mechanism and a MADDPG-based algorithm for privacy-preserving cooperative AI inference at the edge.

Findings

Achieves 0.31-0.95% accuracy improvement with privacy protection.

Enhances performance by 1.54-1.67% considering inference data challenges.

Effectively balances privacy and inference accuracy in MEC settings.

Abstract

Mobile edge computing (MEC) has empowered mobile devices (MDs) in supporting artificial intelligence (AI) applications through collaborative efforts with proximal MEC servers. Unfortunately, despite the great promise of device-edge cooperative AI inference, data privacy becomes an increasing concern. In this paper, we develop a privacy-aware multi-device cooperative edge inference system for classification tasks, which integrates a distributed bidding mechanism for the MEC server's computational resources. Intermediate feature compression is adopted as a principled approach to minimize data privacy leakage. To determine the bidding values and feature compression ratios in a distributed fashion, we formulate a decentralized partially observable Markov decision process (DEC-POMDP) model, for which, a multi-agent deep deterministic policy gradient (MADDPG)-based algorithm is developed. Simulation results demonstrate the effectiveness of the proposed algorithm in privacy-preserving cooperative edge inference. Specifically, given a sufficient level of data privacy protection, the proposed algorithm achieves 0.31-0.95% improvements in classification accuracy compared to the approach being agnostic to the wireless channel conditions. The performance is further enhanced by 1.54-1.67% by considering the difficulties of inference data.