Study On Localization Algorithm Based On Compressive Sensing

With the development of wireless communication network technologies, the location-based services have attracted more and more people’s attention. A large number of wireless location technologies appeared continuously, in which the method based on received signal strength(RSSI) is widely used in the industrial field owing to the principle of simplicity and low cost. Traditional RSSI positioning technology based on channel transmission model to calculate energy distribution, suitable for open areas. In indoor and urban multipath environment, channel transmission is complex and change over time, so the traditional RSSI positioning technology localization accuracy declines seriously. In recent years, the emerging theory of compressed sensing applied to location field, can combine sampling with compression process, and reduced the most redundant data processing, thus improving the effect of wireless network localization.This paper, on the basis of the existing RSS localization algorithm, aiming at the application of positioning requirements,do the following research:① In-depth analysis of indoor wireless location technology. Mainly studied infrared positioning technology, ultrasonic positioning technology, ultra-wideband positioning technology, Zig Bee positioning technology, and positioning technologies based on RFID. Secondly, studied some common algorithms for indoor positioning including TOA, TDOA, AOA and RSSI algorithm. Last, for the complexity of the indoor environment, we introduce several kinds of indoor channel propagation model, and analysis the feasibility of using wireless signal for positioning mobile terminal.② Describes in detal the basic theory of compressed sensing, mainly including sparse representations of signals, signal measurement and signal reconstruction. For indoor wireless positioning system, this paper describes the positioning process of using compressed sensing algorithm in detail mainly including the establishment of the system model, the establishment of the fingerprint feature library, the design of the observation matrix, orthogonal preprocessing and algorithm implementation. Several kinds of common compressed sensing recovery algorithm including OMP, St OMP, SP are simulated in MATLAB, and comparatively analysis of the positioning performance of each algorithm.③ Propose a positioning scheme using compressed sensing theory based on power field reconstruction algorithm. Traditional localization algorithm based on compressed sensing is mainly through the establishment of fingerprint, but indoors or multipath environment to establish and maintain the fingerprint are more complex. Compared with the existing fingerprint methods, the proposed method in this paper does not need to establish fingerprint database, only need to measure the received signal strength of a few grid points in the positioning area, and then recovery the whole area power field with compressed sensing technology through the observation matrix and sparse basis, regarding the coordinates corresponding to maximum power field as the estimate target position. This method reduces the complexity of the algorithm, improves the work efficiency, and can obtain higher positioning accuracy under the same conditions. By optimizing the observation matrix improves the stability of positioning system, and avoids accidental error due to randomness and there will be more practical in the actual location, and has a good application prospect.