An improved hybrid genetic algorithm for multi-user scheduling in 5G wireless networks

摘要

Motivated by the importance of allowing simultaneous user transmissions, especially in fifth-generation (5G) systems, this papers addresses the problem of maximising the number of links that can be activated simultaneously in a wireless network. Solving this problem under the physical signal-to-noise-plus-interference (SINR) model has been demonstrated to be NP-hard. Most previous studies focused on approximation algorithms with guaranteed performance ratios. Although such algorithms have tremendous theoretical value, their surprisingly low approximation ratios limit their practicality. Therefore, some recent studies introduced alternative solutions based on meta-heuristics, such as the genetic algorithm. This paper improves upon a previously proposed genetic algorithm by incorporating problem-specific knowledge into the algorithm. This results in a novel hybrid genetic algorithm that activates almost the same number of links as compared to the original one, while reducing the running time by more than 97%.