| With the appearance of miniature sensor，which has sensation ability, computingability and communication ability, WSN(wireless sensor network), a new network,which has the interdisciplinary characteristics, has been explored and studied by manyscholars. It is widely used and has a good application prospect as a networktechnology, and it also has become a hot research field at home and abroad.Because of its network structure, composition, and the development of othersubjects especially the rapid sensor, wireless sensor network can be used so widely.Generally speaking, the wireless sensor network is a wireless self-organized networkcomposed of cheap and a large number of sensor nodes, and its network is known togood robustness and expansibility. Therefore, the main technologies of wirelesssensor networks includes network routing and coverage and topology technology,network node location and tracking and time synchronization technology, networksecurity technology, network data management and middleware technology andhardware technology. First of all, the design of efficient routing protocols can improvethe communication network connectivity, reducing the energy consumption of nodes,prolong the network survival time; Secondly, the significance of the node location inwireless sensor networks is indubitable, so people need to choose an appropriate nodepositioning method according to different application environments, the cost and thedemand; Thirdly, as the most wireless sensor network is a kind of non-centralized orsemi-centralized self-organization network, so the nodes are decentralized, and theamount of them in the network is huge. Some specific application scenes are moresusceptible to others or the enemies, so to taking appropriate measures, selecting highsecurity network protocols can greatly increase the survival rate and the survival period of network.Nowadays In many scenarios of wireless sensor network applications, thefundamental meaning of its application is that people can receive the requiredinformation instantaneously through it at a specific time, and then people can analyzethese information so that they can know where or what happened in the network.Therefore, the sensor node provides the context information of the environment forthe users, and80%of them are related to the position information. Even in someapplications there are only location information.At present, the localization algorithm of wireless sensor networks can be dividedinto two algorithms, such as range-based algorithm and range-free algorithms,roughly according to many needing to measure the actual distance between nodes ornot in positioning process. Although the precision of range-based algorithm is high,but the hardware cost on the whole network of it is higher, so it is not appropriate forthe most wireless sensor network application scenes which require low cost and lowpower consumption; And in many of the range-free localization algorithms, there arestill a lot of problems yet to be solved, mainly in the following points:(1) Relying on the beacon node ratio. Although the beacon nodes in thenetwork uses range-free localization algorithm is indispensable, but toomuch the number of them will also increase the cost, even if it has only10%of beacon node, the whole network will also increase the price10times.But if the number of beacon node is too small, it will also cause thelocalization accuracy of many algorithms, such as the DV-HOP algorithm,falling rapidly.(2) Needing more network node density. although the accuracy of thepositioning algorithm based on Amorphous location method can reachabove90%, the variance reached10%. But this method is at the cost of thehigher node density, and scalability of network is bad.(3) Effect of network topological structure on algorithms. The wireless sensornetwork nodes are usually randomly, under conditions of practical application, even we don’t consider the effects of3D terrain to the location,but only in two-dimension condition, the network based on anisotropytopology can fully affect on many algorithms such as DV-HOP.This paper, Based on the DV-HOP algorithm which is belong to range-freelocalization algorithm in wireless sensor network, firstly, maintains the zero numberof the nodes with the ranging function, then deals with the highly influence of averagesingle jumping distance calculation error of the DV-HOP algorithm using in networkwhich is heterogeneous, and then modifies the calculation of it. Finally, in order tosolve the problem that is in the final positioning process of DV-HOP, it is verysensitive on average jumping distance error. So the paper choose quantum-behavedparticle swarm optimization algorithm to calculate the results of the DV-HOP.This paper presents a new algorithm named SQDV-HOP algorithm. In the lastpart of the paper, the author uses MATLAB software create a network environmentbased on WSN, and simulates SQDV-HOP algorithm and DV-HOP algorithm underdifferent conditions. The experimental results show that, in the network which hasdifferent number of anchor nodes, location accuracy of SQDV-HOP algorithm isbetter than the DV-HOP algorithm which is improved by about55%; In theenvironments which has different total number of nodes, the former’s locationaccuracy is55%higher than that of the latter; In the situations which has differentnode communication radius, the former’s location accuracy is55%higher than that ofthe latter. Experiments show SQDV-HOP algorithm has higher positioning accuracy.But the paper still has many problems that need further improvement. Forexample, so as to improve the optimization performance of the algorithm, the QPSOalgorithm is used in the algorithm can still be improved; Moreover, the correction ofthe anchor node average hop distance value has the very large promotion space. |
PDF Full Text Request