.png){kind=icon}
.png){kind=icon}
Return
Privacy-preserving distributed aggregative game algorithm under random communication compression
.png){kind=icon}
PRE
AI
DOI：10.1016/j.jfranklin.2026.109029.png)
.png){kind=icon}
Original request for help.png){kind=icon}
Share.png){kind=icon}
Save
Abstract
En 中文
To address the privacy leakage risks and excessive communication resource consumption in distributed aggregative games, this paper proposes a randomized quantization-based distributed projection-free Nash equilibrium seeking algorithm. The algorithm integrates the Frank-Wolfe method to avoid complex projection operations, utilizes a randomized quantization compression mechanism to reduce communication overhead, and achieves differential privacy protection through random errors introduced by quantization. Under switching undirected communication topologies, this paper analyzes the convergence of the aggregative term estimation error, theoretically proves that the algorithm strategies converge to the exact Nash equilibrium in the mean-square sense, and further proves that the proposed randomized quantization scheme achieves differential privacy. Finally, the effectiveness of the algorithm is verified through an energy consumption game simulation, demonstrating the convergence performance under different compression parameters. The results show that the proposed algorithm significantly reduces communication overhead while guaranteeing convergence accuracy.
Keywords:
Aggregative game
Differential privacy
Random compression
Distributed network
Nash equilibrium
Journal
J
IF:
3.7
Papers:
6.4K
Citations:
1.5W