| Popis: |
For the past few years, mmWave based user-centric cell-free (UCCF) networks have gained phenomenal attentions in the wireless industry. In the UCCF network system, a number of smart devices in a geographical area are served simultaneously by multiple distributed access points (APs). In this paper, we study a resource allocation problem and design a novel resource control scheme for multi-connective mobile devices. By considering the main features of mmWave communications, APs are clustered and devices are grouped. To decide each AP’s service price, we adopt the ideas of cooperativemulti-agent learning (CMAL). For multi-connective devices, the generalized Gini bargaining solution (GGBS) is applied to solve the communication task distribution problem. Through the cooperative learning, individual APs learn the best price policy in an interactive manner. Based on the GGBS, each device adaptively distributes its service request into corresponding APs. Our learning and bargaining algorithms are tightly coupled and work together to reach a socially optimal outcome. The main merit possessed by our joint approach is its flexibility to reach a reciprocal consensus under dynamic UCCF network environments. Extensive simulations are conducted to study the performance of the presented scheme. Numerical results confirm that our proposed approach outperforms in the system throughput and device payoff while maintaining comparable fairness among APs. Finally, we conclude this article, and highlight open future research issues. |
