Research On Energy-Efficient Link Schedule And Data Transmission Technology In Wireless Sensor Networks

Wireless sensor networks have many characteristics, such as rapid networking, low cost, invulnerability, and so on. The great application value and development prospects vigorously promote the launch of research work on wireless sensor networks. As wireless sensor networks have multi-hop communication pattern, many-to-one flow characteristic and resource-constrained hardware, in this thesis, from the perspective of energy efficiency, the link schedule and data transmission technologies for wireless sensor networks are studied deeply.Link schedule is the primary issue confronted in wireless sensor networks research. Through systematic analysis to the relationship among link schedule, route selecting, schedule period and energy consumption, a model of joint optimization on energy-efficient link schedule and routing is proposed to minimize schedule peroid and energy consumption. Focusing on this model, a heuristic link scheduling algorithm based on two-stage resolution mechanism is proposed. Firstly, the algorithm uses a weight approach and variable deletion to transform the model, and the topology of network routing and lower limit of schedule period are obtained by using integer programming. Based on topology of network routing and lower limit of schedule period obtained, a maximum interference degree based heuristic time slot allocation algorithm is presented. Further, specificto the defect of heuristic link scheduling algorithm, an energy-efficient link schedule algorithm based on NSGA-â…¡ is proposed. Utilizing reverse routing tree as the basis of chromosome encoding, the algorithm sloves the minimum shcedule period under every chromosome encoding using heuristic time slot allocation based on maximum interference priority, and sloves energy consumption based on chromosome encoding to complete population initiation, then uses nondominated sorting to slice the population and calculates the crowding distance, then gets the best individual by evolutionary. Finally, simulation experiments are analyzed to verify the proposed algorithms.Data transmission is another problem in urgent need to be solved in wireles sensor networks. In this thesis, compressive sensing theory is introduced to reduce amount of data tramismited by sensor nodes. Genetic algorithm is employed to find compressed sensing based data collection tree in wireless sensor networks, under which to achieve the maximum network lifespan. Further, viewing from atomic selection and matching direction, an improved gradient matching algorithm is presented. The algorithm selects more than one atomic into atomic set based on threshold. At the same time, backtracking mechanism for matching direction is introduced. Making use of last matching direction to revise the current one, the number of iterations in need is reduced and the matching speed near the objective value is accelerated. The speed of the reconstruction of the original data is accelerated and the accuracy is improved. A series of simulations verify the effectiveness of the proposed algorithms.