Research On Range-free Localization Algorithm In Wireless Sensor Network

The development of sensor technology, microelectronic technology， computingtechnology and wireless communications technology promotes the emergence anddevelopment of wireless sensor network (WSN), Wireless sensor network is composed by alarge number of micro sensor nodes distributed in a fixed area within the network, thesesensor nodes have data-aware and data processing capabilities, and can be interconnected bywireless channel. The wireless sensor network is a new information acquisition andprocessing technology, it does not require a fixed network infrastructure, with the advantagesof rapid deployment, so it has a broad application prospect in the field of military defense,industrial control, environmental monitoring and forecasting, smart home, urban management,medical control, hazardous area remote control and so on.The research direction of wireless sensor network mainly concentrates in the nodedeployment, localization technology, routing strategy, topology control, data fusion,application-layer technology. Localization technology is the basis for the realization of othertechnologies, so people study the localization technology in-depth, and put forward a numberof different localization algorithms and techniques. In wireless sensor network, according towhether measures the actual distance between the nodes on the positioning process, thelocalization algorithm can be divided into two: the range-based localization algorithm and therange-free localization algorithm. The range-based localization algorithm has highlocalization accuracy, but has high hardware requirements to the sensor nodes, and its powerconsumption is larger. The range-free localization algorithm has low localization accuracy,but has lower hardware requirements to the sensor nodes, and its localization accuracy canmeet the common network applications, so people usually take the range-free localizationalgorithm as the direction.This paper introduces some of the typical range-free localization algorithm and analyzesthe advantages and disadvantages of these algorithms. Based on the excellent idea of thesealgorithms, we propose a new localization algorithm.Centroid algorithm and APIT algorithm is most common algorithms in wireless sensor network localization technology, the former has a features of simple implementation，the latterhas the characteristics of high positioning accuracy. This paper proposes a localizationalgorithm based on circle intersection and grid-scan. In the initial stage of the algorithm, thebeacon node sends a series of information that unknown node need to location, and then theunknown node take a different strategy to calculate its coordinate based on the collectedinformation. When unknown node has less number of neighbor beacon node than threshold N,it takes centroid algorithm calculate its own coordinate, otherwise it determines an area thatcontain it and takes the centroid the region as its coordinate. This paper proposes a newalgorithm to determine the region, which uses the grid scan thinking of APIT algorithm andeliminates the InToOut and OutToIn error of APIT algorithm, has higher localizationaccuracy.In MATLAB simulation environment, with different experimental parameters, weverified that the AIGS algorithm has high localization accuracy by comparison with centroidalgorithm and APIT algorithm. AIGS algorithm has higher positioning accuracy than centroidalgorithm and APIT algorithm in the case of network node density and network connectivityare moderate. When network has high nodes density, the localization accuracy of AIGSalgorithm is less5%. We also have verified that the factors that affect the localizationprecision.The algorithm has high positioning accuracy, but its algorithm complexity is high, so inthe future research work we want to reduce complexity of the algorithm.