A Wireless Sensor Network Localization System Based On State Estimation And Constrained Optimization

With the deepening study of the Internet of things technology, wireless sensor network technology has been rapid development. The localization technology is the key technology of wireless sensor network and the node location information plays an important role in monitoring activities of the whole sensor network. The wireless sensor positioning technology has a broad development space and application prospect in military affairs, environment, transportation, construction, medical and other fields. Therefore, the researching of sensor network localization technology has important theoretical significance and practical value. Therefore, the researching of sensor network localization technology has important theoretical significance and practical value.The principle of received signal strength is to calculate the distance based on the signal strength between the sensor nodes and then to calculate the node location. This method is vulnerable to environmental factors interference and positioning error is bigger. In order to reduce the ranging error and improve the precision of localization, we designed the sensor nodes to experiment and established the ranging model, then use the difference correction algorithm to reduce the distance measurement error. Further, after the difference correction RSSI ranging, a localization algorithm is proposed that based on maximum likelihood estimation with positive and negative constraints. The main innovative research work of this thesis can be summarized as follows:1. In the ranging phase, we design and produce the sensor node and then applicate this node to collect the test data to fit the curve of ranging based on RSSI using least squares method. In order to reduce the ranging error, we use error that produced in the process of the beacon nodes ranging to feedback to the unknown node localization. It can improve the ranging error fundamentally. Finally, the simulation test validates the proposed method.2. After the difference correction RSSI ranging, a localization algorithm is proposed that based on maximum likelihood estimation with positive and negative constraints. In the process of locating, if unknown nodes in communication range of beacon nodes that we adopt maximum likelihood estimation with positive constrains. If unknown nodes out of communication range of beacon nodes that we adopt maximum likelihood estimation with negative constrains. Last, we can combine the positive and negative constraints to established likelihood function, then adopt particle swarm optimization algorithm to find the unknown node. This method adopts the probability estimation to estimate the unknown node location, and introduces the information between the nodes that people often ignored to improve the positioning accuracy. The simulation results show that, the algorithm with positive and negative constraints estimation can be obtained with high precision compared with the traditional location algorithm and its stability is better.