Research On Cooperative Localization Algorithm Of Wireless Network

A Wireless network is made up of multiple networks connected together,whose special feature is that the connection is realized by wireless communication technology instead of the traditional network line.The appearance of wireless network makes the construction of network get rid of the physical constraint,so it has strong flexibility and expansibility.Because of the convenient way of formation,the Wireless network has grown rapidly.It mainly includes Short-distance Bluetooth Network,WLAN and Wireless Sensor Networks(WSNs).WSNs are composed of micro sensors with radio frequency communication capability and multi-hop routing protocol.The micromotor system enables the mechanical part of the sensor to be embedded in the very tiny chip,and the digital electronic technology can realize the network communication and the compression and integration of the collected data.The sensor nodes not only have the characteristics of small size and convenience,but also have the ability to perceive environmental parameters and process data.In view of the fact that WSNs has better technical support and application requirements,this paper mainly analyzes and studies the composition and positioning methods of WSNs.However,the existing positioning technology is mainly for the ideal environment of high-precision positioning scheme,committed to improve the positioning accuracy while no practical significance.In order to overcome these limitations,under the application scenarios of regional rejection and dynamic active sensor network,this paper deeply analyzes the crux of errors caused by each classical algorithm and proposes an improved positioning algorithm.The specific contents are as follows:(1)The composition of WSNs and the hardware inside the sensor are described.The principle,precision range,application scenarios and advantages and disadvantages of the classical algorithm are reviewed.In addition,the error sources of the multi-lateral localization algorithm are analyzed in detail,and the localization process of the Multi-dimensional Scaling(MDS-MAP)algorithm in centralized,distributed and cluster-oriented situations is elaborated.(2)In the case of regional denial of GPS signal,in order to avoid the error accumulation caused by multi-layer positioning of multilateral algorithm,this paper proposes a hybrid positioning algorithm(MA-MDS),which combines the advantages of multilateral algorithm and MDS-MAP.This algorithm not only avoids the problem of error accumulation in multilateral algorithm,but also increases the number of anchor nodes in MDS-MAP algorithm.This algorithm converts relative map to absolute map by using Procrustes Analysis(PA)algorithm.A lot of simulation experiments show that the proposed algorithm is more accurate and efficient than the existing algorithm.(3)In the scenario of large-scale dynamic active WSNs,a hybrid localization algorithm(HLBC)based on clustering is proposed under the inherent defect of resource limitation.The key of HLBC algorithm is the selection of cluster-head node and the filtering and correction of "bad nodes".The algorithm constructs an interest function to select a reasonable number of cluster head nodes with uniform distribution.This function considers not only the connectivity of the network,but also the residual energy of the nodes.Then,a matching degree function is proposed to filter the "bad nodes" in the network and to correct them with the multilateral algorithm.The simulation results show that HLBC algorithm is not only more accurate than MDS-MAP algorithm,but also can increase the lifetime of the network.