RSSI Wireless Positioning Technology Based On The Revised Environmental Factors

With the maturity and development of communication technology, people increasing demand for wireless positioning. To provide users with wireless positioning, such as GPS, GSM positioning,outdoor exercise and automobile navigation,etc have been widely used. But in indoor, GPS, GSM positioning performance cannot be satisfactory, So the indoor location,as the research hotspot of positioning technology get the attention of people more and more.Indoor wireless channel has the following characteristics:complex environment, serious decline phenomenon, multipath effect and reflection phenomenon obviously, weak signal, that make indoor wireless location technology different from that of ordinary positioning system characteristics. So it lead to location harder, At present,the commonly used method is in indoor positioning within the scope of a small network.First, this paper introduces the indoor wireless location technology, in which the research background and significance,the research present situation;Secondly, this 4 kinds of measurement methods, that the RSSI, TOA, TDOA and AOA, this paper made a description and comparison, and research focus and elaborated the present stage. Key research based on RSSI localization algorithm simulation, and compares their advantages and disadvantages. As we applied the algorithm in the design of the indoor wireless location, the algorithm can significantly reduce the number of training through computer simulation, improve the accuracy and its generalization performance is superior to the traditional algorithm in transmission network. And mainly to do the following research work:(1)In non-line-of-sight (NLOS) environment, the received signal strength (RSSI) with large error, which caused the positioning accuracy is not high and on the question of analysis.(2)Based on RSSI data source characteristics, this paper focus on algorithm,and would improve the network is divided into two parts:A:to select representatives after a rough classification and put them into the input layer,which is easy to activate more neurons as much as possible.B:to make full use of each neuron competition layer by using particle swarm optimization algorithm, then iteractive update to get the optimal solution. The alternative to traditional network algorithm wins in the competition mechanism. The improved algorithm is not easy to fall into local optimum, improve the convergence speed when training and avoid competition layer selection of winning neurons randomness, etc.(3)Give a method dealing with a complex nonlinear characteristics, which was cuased by the uncertainty of the unknown. under the condition of that Non line-of-sight environment received signal strength was low, the method was used to processing and prediction. Furthermore, use the maximum likelihood estimation to solve it. It is concluded that this method get good positioning result.