Research on key technologies in wireless communications based on evolutionary algorithms

Wireless communications have been developing rapidly in both the research and industry. However, such developments have raised many challenges across various layers in wireless communications. In particular, many optimization processes have been proved to be NP-hard problems, which hinder further technological advances. Evolutionary Algorithm (EA), as part of the Artificial Intelligence, provides a generic heuristic optimization technique motivated by natural evolution. EAs usually work well in a category of combinatorial NP-hard problems by building better solutions through the recombination of the best part of past solutions, rather than attempting all possible combinations. As such EA is able to find near optimal solutions through solution generation, selection and rearrangement, reducing the complexity of solving NP-hard problems. Our research is motivated by the need to optimize difficult discrete optimization problems in wireless communication systems. In particular, we developed novel EA algorithms to address some of the NP-hard problems across various layers, ranging from Physical layer, Data Link layer, to Network layer, in wireless communication systems. We demonstrate that our EA designs achieve significant performance improvements for the systems under investigation with lower computational complexity and fast convergence. The main innovations and contributions of this thesis are as follows: In the Physical layer, we take on the challenge of the reduction of the peak-to-average power ratio (PAPR) in orthogonal frequency division multiplexing (OFDM) systems. We propose a modified chaos clonal shuffled frog leaping algorithm (MCCSFLA) for PAPR reduction. We also analyze MCCSFLA using Markov chain theory and prove that the proposed algorithm converges to the global optimum. Simulation results show that the proposed algorithm achieves better PAPR reduction than using others heuristics. Additionally, MCCSFLA has lower computational complexity and faster convergence than other heuristics. In the Data link layer, we investigate the target coverage problem in large-scale self-organizing wireless sensor networks (WSNs), we propose a method based on a quantum ant colony evolutionary algorithm. We build the WSNs target coverag esystem model, design and apply the proposed method to WSNs. Simulation results show that the proposed method can significantly increase the target coverage rate in WSNs. Furthermore, we expand our investigation to the communication coverage problem in WSNs. We propose a new quantum immune clonal evolutionary algorithm for the duty cycle sequence design with a full coverage constraint. Simulation results show that the proposed algorithm not only maintains full coverage of all the targets in the monitoring area but also extends the network lifetime of the WSN. Additionally, in order to reduce the communication energy consumption in large-scale WSNs, we propose a fuzzy simulated evolutionary computation clustering method. We design a fuzzy controller for the algorithm parameter adjustment. Simulation results show that the proposed method can significantly reduce the energy consumption of large-scale WSNs. In the Network layer, we take on the challenge of providing Quality of Service (QoS) routing for multimedia wireless sensor networks. A novel parallel elite clonal quantum evolutionary algorithm is proposed to solve the multi-constraints QoS routing problem. Simulation results demonstrate that the proposed algorithm achieves lower energy consumption at a faster convergence rate than the other heuristic algorithms. Wireless communication is a key technology in the modern society and we believe new evolutionary algorithms can contribute a growing number of solutions in this area.