Joint User Association and Beamforming Multi-Objective Optimization for Distributed mmWave Networks

In this letter, we investigate multi-objective optimization methods for joint user association and beamforming in downstream distributed millimeter-wave (mmWave) networks. Multiple goals are considered together, including maximizing system throughput, minimizing end-to-end latency between users and remote radio units (RRUs), and RRU load balancing. We use Fuzzy Inference System (FIS) to dynamically infer the weights of different objective functions according to user needs, assisting operators to provide customer-centric network access services. In addition, we innovatively adopt the extended gene coding method in the genetic algorithm. The simulation results show that extending the genetic code can significantly improve the performance of the genetic algorithm. The user-centric network access service can be realized by the fuzzy logic and the genetic algorithm. In this way, network throughput, end-to-end delay between users and RRUs, and load balancing can be simultaneously improved.