English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Interference among concurrent transmissions in a wireless network is a keyfactor limiting the system performance. One way to alleviate this problem is tomanage the radio resources in order to maximize either the average or theworst-case performance. However, joint consideration of both metrics is oftenneglected as they are competing in nature. In this article, a mechanism forradio resource management using multi-agent deep reinforcement learning (RL) isproposed, which strikes the right trade-off between maximizing the average andthe $5^{th}$ percentile user throughput. Each transmitter in the network isequipped with a deep RL agent, receiving partial observations from the network(e.g., channel quality, interference level, etc.) and deciding whether to beactive or inactive at each scheduling interval for given radio resources, aprocess referred to as link scheduling. Based on the actions of all agents, thenetwork emits a reward to the agents, indicating how good their joint decisionswere. The proposed framework enables the agents to make decisions in adistributed manner, and the reward is designed in such a way that the agentsstrive to guarantee a minimum performance, leading to a fair resourceallocation among all users across the network. Simulation results demonstratethe superiority of our approach compared to decentralized baselines in terms ofaverage and $5^{th}$ percentile user throughput, while achieving performanceclose to that of a centralized exhaustive search approach. Moreover, theproposed framework is robust to mismatches between training and testingscenarios. In particular, it is shown that an agent trained on a network withlow transmitter density maintains its performance and outperforms the baselineswhen deployed in a network with a higher transmitter density.

When Multiple Agents Learn to Schedule: A Distributed Radio Resource Management Framework