TDMA Scheduling In Wireless Sensor Networks Based On Multi-Objective Soft Computing Optimization Algorithms

Due to the rapid convergence of sensor, micro-electro-mechanism system and wireless communication technologies, wireless sensor network (WSN) is a novel technology about acquiring and processing information. With functions of real-time data acquisition and wireless data transmission, WSN provides an entirely new computing platform that realizes the interaction between the mankind and the physical world. WSN has a huge potential of applications.Although the promising applications enabled by wireless sensor networks are very attractive, there are many system challenges to resolve. First of all, energy is an essential problem since sensors are usually battery-powered. Second, in some emergency applications, a short time of data collection is also required. Towards such a data gathering sensor networks, TDMA is a good choice to satisfy the above requirements: 1) TDMA can save energy by eliminating collisions, avoiding idle listening, or entering inactive states until their allocated time slots. Second, as a collision-free access method, TDMA can bound the delays of packets, guarantees reliable communication.Consequently, in this literature, a new optimization framework and a new encoding method are proposed to solve the TDMA slot allocation problem in WSN with many-to-one communication mode. Based on a new soft computing method, i.e., Particle Swarm Optimization (PSO), a hybrid algorithm called HPSO and a multiple objective algorithm called PAPSO are proposed to solve single objective optimization problem and multiple-objective optimization problem respectively. To the former one, two population-based algorithm, PSO and simulated annealing (SA), are hybridized to enhance the searching ability. Simulation results with different network scales shows that the proposed hybrid algorithm is superior over other algorithms on a specified objective, which can be the total time or the total energy for data collection. To the latter one, PAPSO is proposed, where two solution evaluation mechanisms are used, i.e., Coello Coello method for local evaluation and selection from Pareto Archive for global evaluation. PAPSO maintains a global Pareto solution set during the whole process. Simulation results show that PAPSO can get a Pareto set with a good covering performance, which can provide more choice between the delay performance and energy performance.