Research On Coverage Optimization Methods In WSNs

The target coverage method in WSNs can prolong the network lifetime under the premise of ensuring the continuous monitoring of all the target points.The type of sensor's sensing direction and sensing radius directly affect the complexity of target coverage problem in the network.This paper investigates the target coverage problem under the condition that the sensing radius of wireless sensor is fixed and adjustable,and the condition that sensing directions are both directional and omnidirectional.Three coverage algorithms are proposed to prolong the network lifetime respectively.In directional sensor networks which sensors have fixed sensing radius,the greedy algorithm is easy to fall into the local optimal solution when solving the target coverage problem.The genetic algorithm needs to predict the theoretical upper limit of network lifetime.Aiming at solving the problems above,a multi-objective coverage optimization memetic algorithm is proposed to schedule directional wireless sensors.The algorithm uses one-dimensional linear array structure to encode chromosomes,establishes the model of biological evolution process,and initializes a certain number of chromosomes to form population in a randomized manner.The enhancement operation is proposed to recombine chromosomes for increasing the diversity of population,improves the probability of finding a better solution;The fitness function evaluates the chromosome in a linear way from the three dimensions of the number of cover sets,the variance of residual time and the number of unused sensors.In order to maintain the quality of solutions in the population,the algorithm uses selecting and eliminating operations to preserve the high quality solutions in the population and to eliminate chromosomes with lower fitness values.After many generations of selection,elimination and enhancement,the chromosome with the highest fitness value is used as the scheduling scheme to prolong the network lifetime.As for the omnidirectional sensor network in which sensors have adjustable sensing radius,this paper proposes a heuristic target coverage algorithm to prolong the network lifetime.The algorithm is divided into two stages,namely,the cover set construction and the cover set optimization: the stage of cover set construction uses selection strategy based on three factors including sensor's coverage contribution,energy waste and residual energy to construct cover set.The strategy reduces the number of sensors in the cover set and overlapped coverage to improve network energy efficiency.The second stage optimizes the number of sensors and the sensing radius of the cover set constructed in the previous stage.The energy is saved further to prolong the network lifetime.Finally,an imperialist competition coverage optimization algorithm is proposed to solve the target coverage problem in an omnidirectional sensor network with a fixed sensing radius.The algorithm uses a linear array to encode the “country” and a random function to construct the population of “country”.The colonial countries make an approximate change to the empire through imperial assimilation operation which improves the quality of population on the whole.The revolution operation is used to reconstruct the values of a proportion of countries in the population for increasing the diversity of population and to improve the probability of finding the optimal solution;The imperial competition operation is proposed to eliminate countries with low fitness value in the population for ensuring the quality of the solutions contained in the countries.The initial country population undergoes multiple generations of the operations of assimilation,competition and revolution.The algorithm chooses the “country” with the highest fitness value as the solution to the target coverage problem.Through experimental verification and comparative analysis,all three algorithms proposed in this paper can effectively solve the target coverage problem in the corresponding network,and acquire a long network lifetime in the meantime.