Research On Environmental-Adaptive Localization And Routing Algorithms For Indoor Wireless Sensor Networks

In recent years, Internet is now moving towards Internet of Things (IoT). Many objects in the IoT require the abilities to monitor environment and test data. Therefore, the technologies for wireless sensor networks (WSN) have become the important support for the development of IoT, and they will be increasingly used in indoor environment. However, many key technologies for WSN apply to outdoor or ideal environment, and their application effects are not ideal in complex indoor environment. The localization and routing algorithms for indoor WSN are investigated in this dissertation, and the main contents and research results are as follows:(1) The signal transmission characteristics of WSN in the indoor environment are summed up by analyzing the testing data of residential, office and chip manufacturing environment. It can provide the reliable basis for the research on key technologies of WSN in indoor environment.(2) For typical indoor environments, the characteristics of channel propagation for sensor nodes are tested, and the testing data are fitted. The fitting results show that the proposed three-order polynomial log-distance path loss model can better characterize the channel fading of WSN in the indoor environment than the classic one and two slope models.(3) When WSN is used in complex indoor environment, great propagation loss will be caused and it is very difficult to estimate adaptively the location of target nodes when environment changed. In this dissertation, two indoor adaptive localization algorithms for WSN are proposed based on one slope model and three-order polynomial model, respectively. In the two algorithms, the computer will build the parameter set of models according to the data from the fixed nodes, and can update the set with environmental variation regularly. Then, after the model parameters are decided by searching them according to the proposed method of selecting parameters, the iteration method and the maximum likelihood estimation method are applied to estimate the positions of target nodes, respectively. Through experimental tests, the better localization accuracy of the two localization algorithms is demonstrated.(4) An Energy-Balanced Steady Clustering (EBSC) routing algorithm is proposed for static ideal indoor environment. In EBSC algorithm, the number of cluster heads generated in each round is very steady, and EBSC combines the advantage both distributed and centralized clustering algorithm. For dynamic complex indoor environment, an Energy-Balanced Adaptive Clustering Routing (EBACR) algorithm is proposed. In the clustering mechanism, the path loss and the residual energy of the nodes are considered and the multiple attribute decision methods with different subjective coefficients are adopted to choose all kinds of nodes. Thus, the routing can be built adaptively. Experimental results show that the two routing algorithms not only efficiently use the limited energy of network nodes, but also balance the energy consumption of all nodes.(5) For the stringent requirements of chip manufacturers about the environment, chip manufacturing clean room environment monitoring system based on WSN is designed with the proposed EBACR algorithm. The system is tested in the chip manufacturing clean room. The results show that the system can not only monitor the changes of environment, but also ensure the effectiveness and balance of energy consumption for sensor nodes.