Safe Multi-Agent Deep Reinforcement Learning for Privacy-Aware Edge-Device Collaborative DNN Inference
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866910035747012608 |
|---|---|
| author | Wang, Hong Fan, Xuwei Cheng, Zhipeng Yuan, Yachao Min, Minghui Liwang, Minghui Xia, Xiaoyu |
| author_facet | Wang, Hong Fan, Xuwei Cheng, Zhipeng Yuan, Yachao Min, Minghui Liwang, Minghui Xia, Xiaoyu |
| contents | As Deep Neural Network (DNN) inference becomes increasingly prevalent on edge and mobile platforms, critical challenges emerge in privacy protection, resource constraints, and dynamic model deployment. This paper proposes a privacy-aware collaborative inference framework, in which adaptive model partitioning is performed across edge devices and servers. To jointly optimize inference delay, energy consumption, and privacy cost under dynamic service demands and resource constraints, we formulate the joint problem as a Constrained Markov Decision Process (CMDP) that integrates model deployment, user-server association, model partitioning, and resource allocation. We propose a Hierarchical Constrained Multi-Agent Proximal Policy Optimization with Lagrangian relaxation (HC-MAPPO-L) algorithm, a safe reinforcement learning-based framework that enhances Multi-Agent Proximal Policy Optimization (MAPPO) with adaptive Lagrangian dual updates to enforce long-term delay constraints. To ensure tractability while maintaining coordination, we decompose the CMDP into three hierarchically structured policy layers: an auto-regressive based model deployment policy, a Lagrangian-enhanced user association and model partitioning policy, and an attention-based resource allocation policy. Extensive experimental results demonstrate that HC-MAPPO-L consistently satisfies stringent delay constraints while achieving a superior balance among energy consumption and privacy cost, outperforming representative baseline algorithms across varying problem scales and resource configurations. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_00129 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Safe Multi-Agent Deep Reinforcement Learning for Privacy-Aware Edge-Device Collaborative DNN Inference Wang, Hong Fan, Xuwei Cheng, Zhipeng Yuan, Yachao Min, Minghui Liwang, Minghui Xia, Xiaoyu Multiagent Systems Machine Learning As Deep Neural Network (DNN) inference becomes increasingly prevalent on edge and mobile platforms, critical challenges emerge in privacy protection, resource constraints, and dynamic model deployment. This paper proposes a privacy-aware collaborative inference framework, in which adaptive model partitioning is performed across edge devices and servers. To jointly optimize inference delay, energy consumption, and privacy cost under dynamic service demands and resource constraints, we formulate the joint problem as a Constrained Markov Decision Process (CMDP) that integrates model deployment, user-server association, model partitioning, and resource allocation. We propose a Hierarchical Constrained Multi-Agent Proximal Policy Optimization with Lagrangian relaxation (HC-MAPPO-L) algorithm, a safe reinforcement learning-based framework that enhances Multi-Agent Proximal Policy Optimization (MAPPO) with adaptive Lagrangian dual updates to enforce long-term delay constraints. To ensure tractability while maintaining coordination, we decompose the CMDP into three hierarchically structured policy layers: an auto-regressive based model deployment policy, a Lagrangian-enhanced user association and model partitioning policy, and an attention-based resource allocation policy. Extensive experimental results demonstrate that HC-MAPPO-L consistently satisfies stringent delay constraints while achieving a superior balance among energy consumption and privacy cost, outperforming representative baseline algorithms across varying problem scales and resource configurations. |
| title | Safe Multi-Agent Deep Reinforcement Learning for Privacy-Aware Edge-Device Collaborative DNN Inference |
| topic | Multiagent Systems Machine Learning |
| url | https://arxiv.org/abs/2603.00129 |