A RSSI-based WSN Localization Algorithm And Ranging Factor

With the popularity of the concept of IOT and further research, as one of coretechnology of IOT, WSN technology attracts more enthusiastic attention of researchers ineach country. Military, medical, construction, transportation and other areas ofapplications emerge endlessly, the common feature of these applications is the need toobtain the location information of events, which involves WSN localization technology.Generally, WSN localization can be divided into two types: based on the distance anddistance irrelevant. The latter has no ranging requirements and less hardware cost,however, its positioning accuracy is worse than the former. The advantage of thepositioning method based on RSSI ranging demonstrates its convenience and low costfeature. The positioning method would have broader commercial application prospect ifcertain measures can be taken to reduce errors caused by ranging and location algorithm.Specific research and innovation work is given as follows:1. This thesis introduces the concept, system architecture, features and applicationsin different areas of WSN. The significance of research on wireless sensor networkpositioning technology is pointed out. The thesis elaborates various positioning systemsand positioning algorithm of WSN, and summarizes the state of the art work overmainland or overseas.2. This paper introduces types of the radio propagation mathematical models, fits theRSSI ranging model by collecting data in experiments. And then explores two factors ofranging based on RSSI, which have certain influence. The two factors are direction of theantenna and movement speed of person. We calculates the mean and variance of RSSI byusing the experimental data, analyzes the calculation result and fluctuation range of theRSSI to determine the impact degree of the factors mentioned above.3. This paper analyzes the weighted centroid localization algorithm based on RSSI,proposes three improvements: process a large number of RSSI values by normaldistribution, determine the area which has large probability, introduce difference correction model to the distance which is obtained by RSSI ranging model, the differencemodel correct the distance by using the ranging error factor of reference node, considerthe factors that influence the positioning accuracy of the trilateral positioning method, wechoose three weighting factors: distance, angle, size and length of side, the improvedalgorithm uses these factors to get the final estimated position of the unknown node fromthe multiple approximate positions of the unknown node，which are obtained by thetrilateral positioning method.4. We compare the improved algorithm with the weighted centroid localizationalgorithm based on RSSI and difference correction centroid localization algorithm basedon RSSI, then do simulation to verify the advantage of the improved algorithm inreducing average location error through Matlab platform, we compare the averagelocation error by changing the number of beacon nodes and using different weightscombinations. Then we determine the relatively good combination.