Joint Optimization of Model Partitioning and Resource Allocation for Anti-Jamming Collaborative Inference Systems

2026-03-02

AI Summary

This work addresses the vulnerability of intermediate features to malicious interference during wireless transmission in edge-assisted deep inference on resource-constrained devices, which severely degrades both accuracy and latency performance. To mitigate this issue, the paper pioneers the integration of anti-jamming objectives into edge-device collaborative inference systems and proposes a joint optimization framework aimed at maximizing system utility—defined as a trade-off between inference accuracy and delay. The framework co-optimizes model partitioning, computational resource allocation, and transmit power. To solve the resulting mixed-integer nonlinear problem, an alternating optimization strategy is devised, leveraging KKT conditions, convex optimization, and a quantum-inspired genetic algorithm to efficiently handle subproblems. Extensive simulations demonstrate that the proposed scheme significantly outperforms existing baselines in terms of system utility (RDA).

Technology Category

• Machine Learning: Learning on the Edge & Model Compression
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Search and Optimization: Learning to Search

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

With the increasing computational demands of deep neural network (DNN) inference on resource-constrained devices, DNN partitioning-based device-edge collaborative inference has emerged as a promising paradigm. However, the transmission of intermediate feature data is vulnerable to malicious jamming, which significantly degrades the overall inference performance. To counter this threat, this letter focuses on an anti-jamming collaborative inference system in the presence of a malicious jammer. In this system, a DNN model is partitioned into two distinct segments, which are executed by wireless devices and edge servers, respectively. We first analyze the effects of jamming and DNN partitioning on inference accuracy via data regression. Based on this, our objective is to maximize the system's revenue of delay and accuracy (RDA) under inference accuracy and computing resource constraints by jointly optimizing computation resource allocation, devices' transmit power, and DNN partitioning. To address the mixed-integer nonlinear programming problem, we propose an efficient alternating optimization-based algorithm, which decomposes the problem into three subproblems that are solved via Karush-Kuhn-Tucker conditions, convex optimization methods, and a quantum genetic algorithm, respectively. Extensive simulations demonstrate that our proposed scheme outperforms baselines in terms of RDA.

Problem

• anti-jamming
• collaborative inference
• DNN partitioning
• resource allocation
• jamming attack

Innovation

• DNN partitioning
• anti-jamming
• collaborative inference
• resource allocation
• quantum genetic algorithm