A Research On The Localization Technology Of Wireless Sensor Network Based On RSSI Value

Wireless sensor network as a new field of interdisciplinary has been paid muchattention to by people. Wireless sensor network location technology, as one of thekey technologies of WSN, has played a fundamental role in the research andapplication of WSN. What’s more, it has become a hot issue around the scholars.This paper introduces the wireless sensor network location technology researchsituation at home and abroad, expounds the wireless sensor network positioningsystem composition, and makes a detailed description of the localization algorithmand performance evaluation standard. What’s more, it introduces some existinglocalization algorithms based on range-free and range, compares their advantagesand disadvantages, applicable scopes. The main research contents are as follows:1. According to the traditional localization algorithms based on RSSI(Received Signal Strength Indicator), the multipath transmission has an effect onthe ranging accuracy. This article puts node in different heights, and does a lot ofexperiments to obtain the relationship between the RSSI value and the distance d.At the same time, this article analyzes the influence of different node heightstowards to RSSI-d curve;2. This article adopts the least square fitting method and subsectionprocessing method to contrast and analyze the RSSI-d curves. The processingmethod of subsection can more accurately reflect the relationship between RSSIvalue and the distance d in the real environment. At the same time, in order to dealwith the problem that RSSI-d curve hasn’t monotonicity, this article puts forwardthe mobile step length calibration method. The result of calibration methodindicates that the relative error of the node location has been improved. We also cansee the position of node real-time and the trend for monitoring;3. According to the measured RSSI-d curve data, this article introduces EMDalgorithm to obtain signal decomposition. This article insists residual signal as thetrend of the RSSI-d curve and compares it with least squares fitting curve. On thisbasis, we make use of EMD decomposition to establish ranging model cushion, andcompare it with the traditional curve fitting processing method. Experimentalanalysis shows that, as the relative error is concerned, the ranging accuracyintroducing EMD algorithm processing is superior to the traditional processingmethod;4. Based on the improved location algorithms, this article completes wirelesssensor network localization algorithm simulation, and makes a comparative analysis with the traditional location algorithm based on RSSI.The simulation results show that the proposed methods can effectively improvethe accuracy of the positioning.