Target Coverage Research Based On The Model In Wireless Sensor Network

As a lowcost and self-organizing network technology, a brand new platform for gettinginformation is opend up by wireless sensor network. Target coverage control, as an importantbasis of all the wireless sensor network technologies, is the primary task. This technology isput forward later than the area coverage which is researched generally and based on binarysensing model. However this model is only a simply description for sensors, so in order toreflect the availability of sensors, the issue of sensors target coverage through directional andprobability model are studied.On condition that all the targets can be covered, the best sensors set is get. There are twokey points of research: target coverage based on priority in directional sensor networks andtarget coverage of heterogeneous sensor networks in three-dimensional space.(1) Directional sensors imitate the sensors such as rotatable monitor effectively.According to the directional perception model, this paper constructs the direction choosingmatrix so as to make directional sensors can choose directions from the matrix. Also thepriority constraint condition is introduced in order to improve the system model. According tothe deficiency of real-time and optimal results in the already existing algorithms for this issue,this paper improves its performance by using two kinds of intelligent algorithm. And thefollowed solutions are proposed:The first solution increases the diversity of the population by bringing vaccinationmechanism in PSO. In addition, introduces weighting factor to weigh the local and globalsearch abilities, so the better results can be found availably.The second solution uses the simulated annealing algorithm which can expand searchspace by accepting the poor results at a certain probability. Also this algorithm can beginoptimizing only from one population with pithy steps and get the best results more quickly.Through empirical evidence, those two intelligence algorithms which are proposed bythe paper are proved suitable for the optimization of the target coverage for directionalsensors. The overall performance of the network is improved: the number of sensors isreduced and the real-time characteristic is enhanced.(2) The perception of the sensors belongs to probability events, so it presses to the realitymore closely by using probability model. The three-dimensional space is closer to nature environment than plane area and the study in the three-dimensional space is more accurate.The super points are put into the network to remedy the energy loss of common pointsincreasing when the size of network is amplified. Target coverage of heterogeneous sensornetworks is studied in three-dimensional space based on probability model. According to thedefects of algorithm which has already applied, the problem is optimized further by usingimmunodominace clone algorithm. Through the contrast between immunodominance clonealgorithm and genetic algorithm, the dominance in application has been found from the newalgorithm: it has been found that only optimizing with cloned antibodies can help save theoperating time. At last the simulation results show that the immunodominace clone algorithmhas more desirable results when it is used for reducing the simulation time.