Research On Range-free Localization In Wireless Sensor Networks

Wireless Sensor Networks (WSN), which is the integration of sensor techniques, MEMS techniques and network techniques, is an innovative technology of information acquisition and processing and has changed the way humans interact with the natural world. It is a self-organized and multi-hop wireless communication network composed of a large number of small nodes with limited computing and communication capabilities. Self-organization, mini-size, low-cost, flexibility and other characters of WSN make it show great potential application value in the military national defense, environmental monitoring, anti-terrorism, disaster relief, intelligent home and many other areas.Node localization is the basic key issue in research of WSN, which is also the premise of a variety of applications in WSN and the support of many key technologies. It brings big challenges to node localization problem due to the large-scale, micro-nature of nodes, limited computing and communication capability and other characteristics of WSN. In response to these challenges, the academia has carried out research on Range-free localization technology. Compared to range-based localization technology, Range-free localization technology is in low level requirement to node hardware, low power consumption while the localization accuracy is able to meet most application requirements. Range-free localization algorithm generally includes two stages: The first is to estimate the distance between the unknown nodes and beacon nodes, the second is to calculate the coordinates for the unknown nodes. In this paper, based on summing up the existing technology and research, the rage-free localization technology is discussed, and the main contents include the following aspects:(1) Localization technology of WSN is summarized. Typical localization algorithm, classification standard, evaluation indicators as well as challenges faced are analyzed and discussed in detail. Sensor nodes connectivity and the probability distribution characteristics of localization errors are analyzed, which has inspiration and guiding significance for the localization algorithm design and performance testing.(2) An Optimal Distance Estimation Algorithm Based on Deployment Statistical Characteristic (ODEDS) is proposed. A network deployment model based on two-dimensional Gaussian distribution is constructed due to the large-scale, random tossed and other characteristics of WSN. Based on statistical characteristic of this model, the algorithm first uses the network connectivity information to obtain the practical value of the neighbor nodes number, and calculate the theoretical value of the neighbor nodes number, thereby establishes the error function of the practical value and theoretical value. The optimal distance estimation between the nodes is the value of the independent variable which can make the error function take the minimum value. Simulation results prove the effectivity of the algorithm. The algorithm is base on local information which does not need to interact across the network and requires less communication overhead and energy consumption accordingly.(3) A Range-free Localization Algorithm Based on Optimal Distance Estimation (LAOD) is proposed. The algorithm is divided into four stages: First, after determination of its own location coordinates, the cluster head node (beacon node) broadcasts its own location information, and the member nodes (unknown nodes) obtain the location information of their cluster head node. Secondly, according to the network connectivity information, the member nodes calculate their neighbor nodes number. Then, according to the number of neighbor nodes, uses ODEDS algorithm to estimate the distance between the member nodes and the adjacent cluster head nodes .Finally, according to the distance between the member nodes and the cluster head nodes, uses maximum likelihood estimation to calculate the location coordinates of the member nodes. Simulation results show that: LAOD algorithm has good localization accuracy and localization coverage, and does not require much to the number of beacon nodes. LAOD algorithm has broad prospects in theoretical research and engineering practice.