Optimization of Wireless Sensor Networks based on Bio-inspired Evolutionary Computing with Collective-effect and Context-awareness

• Main Authors: Yi-Ting Chen, 陳怡婷
• Other Authors: Mong-Fong Horng
• Format: Others
• Language: en_US
• Published: 2015
• Online Access: http://ndltd.ncl.edu.tw/handle/76932095813984053600

博士 === 國立高雄應用科技大學 === 電子工程系碩士班 === 103 === In the past, due to battery power limitation, wireless sensor networks (WSNs) urgently request for efficient management approaches to prolong network lifetime and advance network performance such as coverage, reachability, shortest routes, and so on. An efficient management contributes to solve the critical problems in WSNs, including energy waste, interference and defective connectivity. In this work, energy-efficiency WSNs (EEWSNs) is presented to develop WSNs with long lifetime and premium network performance. Firstly, the issues of transmission power and network reachability are investigated in depth to propose beneficial approaches based on topology control and route management. The network structure, network topology and network routing are considered to design Modern WSNs. In the network structure and network topology, a three-tier hierarchical network with full-coverage and full-reachability is constructed. Then, the network routing including reactive routing strategies and proactive routing strategies is performed in this constructed hierarchical network to manage data routes. In addition, Next-Generation Evolutionary Computing (EC 2.0) is created to establish Modern WSNs with energy-efficiency WSNs. The proposed EC 2.0 is also significant part of Modern WSNs. In traditional ECs, cooperation and competition are the dominated principles to evolve population toward optimal solutions. In EC 2.0, conflict theory is introduced to help the efficiency of solution discovery. Conflict between individuals is healthful behavior for population evolution. Constructive conflict promotes the overall quality of population. Conflict, competition and cooperation are the three pillars of collective effects investigated in this work. Additionally, the context-awareness property is another feature of EC 2.0. The context-awareness indicates that the individuals are able to perceive the environmental information by physic laws. In experiments, the performance of the proposed EC 2.0 are evaluated by the statistical analysis and these indicators including accuracy (ACU), success rate (SR), problem-solving speed (PSS) and stability (STA). A series of numeric results shows that the performance of EC 2.0 is better than the traditional ECs particularly for high-dimension problems. In summary, the proposed EC 2.0 breaks the bottleneck of traditional ECs to create the new paradigm in ECs. As well as, EC 2.0 not only represents a better performance but also requires a less iteration in discovery global optimal solution. Furthermore, these proposed approaches of EC 2.0 are also applied to optimize the performance of WSNs. In conclusion, this dissertation creates EC 2.0 to present novel conception different from traditional ECs to design Modern WSNs. In this dissertation, the created EC 2.0 includes echo-aided bat algorithm (EABA), guidable bat algorithm (GBA) and Flip-Flop ant colony system (FFACS). Besides, many beneficial approaches mentioned above are proposed to achieve these two objectives of prolonging network lifetime and advancement network performance.