Research On Weighted Centroid Localization Algorithm Based On K-means Clustering Point Density

With the rapid development of mobile communication technology, indoor positioning technology based on Wireless Sensor Networks(Sensor Networks Wireless, WSNs) has received more and more attention. Wireless sensor network is composed of a large number of intelligent sensor nodes randomly deployed in the monitoring area, the sensor nodes form a multi hop dynamic self organizing network, which is designed to carry out cooperative sensing, acquisition and processing of the information in the network coverage area. As one of the important technologies in the automatic acquisition of information, wireless sensor network technology integrated wireless communications, embedded technology and sensor technology and other technologies, its application is very wide, such as underground personnel positioning?fire rescue? safety monitoring?environmental monitoring and so on.Although, now today's science and technology more and more developed, but indoor relief work of mining accidents?fire due to the inability to precisely locate to victim location, and that increases the risk, that is a problem to be solved. In this paper, based on the traditional RSSI ranging and non ranging of centroid location algorithm, taking the advantages of the two fusion, proposed a weighted centroid localization algorithm based on K-means clustering point density(KCPD-WCLA). The main work of this paper is as follows:(1) Firstly, the correlation algorithm of indoor location is analyzed and studied, divided into two kinds of localization algorithm based on distance and non ranging. The main focus of this paper is on the RSSI based ranging and centroid algorithm for non ranging.(2) Due to the application of WSNs, the number of anchor nodes is generally more. When the positioning is carried out, the large amount of data is processed to face the problem of large amount of computation and high computational complexity, and it is necessary to find a suitable algorithm to deal with a large number of data sets. However, there are many kinds of clustering algorithms which have been processed by a large number of data sets, this paper focuses on the research of K-means clustering algorithm based on dividing method.(3) Using the RSSI measure obtained with a large number of unknown nodes and anchor nodes distance. The original grouping method is to use the 3grouping method for the number of anchor nodes, and then use the three edge location method to get a lot of estimates of the true position of the n. However, due to the n value is generally relatively large, the use of 3 calculation is very large, it will increase the computational complexity. In this paper, a new method for computing complex and smaller packets is proposed.(4) The location algorithm based on RSSI distance and weighted centroid fusion, and combined with some characteristics of K-means clustering, The application of wireless sensor network in indoor positioning. A weighted centroid localization algorithm based on K-means clustering algorithm(KCPD-WCLA) is proposed. Because of its simple, easy and effective way to reduce the complexity of the WSNs algorithm, compared with other existing indoor positioning algorithm, the positioning accuracy is improved.Finally, this paper uses MATLAB7.0 simulation tool to simulate the KCPD-WCLA algorithm? MLA algorithm and WCLA algorithm. The experimental results show that:the proposed algorithm has obvious improvement in positioning accuracy than MLA and WCLA. It has the advantages of strong operability and high positioning accuracy, that consistent with the WSNs general application scenarios, with universal applicability.