Summary of A Federated Online Restless Bandit Framework For Cooperative Resource Allocation, by Jingwen Tong et al.
A Federated Online Restless Bandit Framework for Cooperative Resource Allocation
by Jingwen Tong, Xinran Li, Liqun Fu, Jun Zhang, Khaled B. Letaief
First submitted to arxiv on: 12 Jun 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Signal Processing (eess.SP)
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary The paper proposes a federated online restless multi-armed bandit (RMAB) framework to tackle the cooperative resource allocation problem in Markov reward processes (MRPs) with unknown system dynamics. The framework is designed to mitigate communication overhead and data privacy issues through federated learning, enabling multiple agents to collaboratively learn the system dynamics while maximizing their accumulated rewards. A Federated Thompson Sampling-enabled Whittle Index (FedTSWI) algorithm is developed to solve this multi-agent online RMAB problem efficiently and privately. The algorithm enjoys a high communication and computation efficiency, as well as a privacy guarantee. The paper also derives a regret upper bound for the FedTSWI algorithm and demonstrates its effectiveness in an online multi-user multi-channel access scenario. Numerical results show that the proposed algorithm achieves a fast convergence rate of O(sqrt(Tlog(T))) and better performance compared to baselines. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary The paper solves a big problem in resource allocation, where multiple agents need to work together while learning about a system they don’t fully understand. They develop a new method called Federated Thompson Sampling-enabled Whittle Index (FedTSWI) that lets these agents make good decisions quickly and privately. This is important because it could be used in many different areas, like network control or energy management. The paper shows that FedTSWI works well and can even learn faster as more agents join the system. |
Keywords
* Artificial intelligence * Federated learning
