The Research Of UWB Based Positioning Techniques

In recent years, the location service industry develops rapidly, and has a broad market space in various fields such as the Internet of things, the wis-dom of city construction, disaster relief and disaster mitigation. At present, the global positioning system (GPS) has solved the positioning problem in outdoor environment. However, to realize high-accuracy positioning is still a great chal-lenge for the GPS-denied indoor environment. Ultra-wideband (UWB) tech-nology is a key technology for high-accuracy positioning, due to its advantages of strong penetrating power, high ranging accuracy, and anti-multipath capabil-ity. This paper studies the UWB based localization methods, analyzes the lim-itation of current UWB positioning technologies and the problems which still need to solve. In this paper, two improved TOA estimation algorithms are pre-sented for indoor multipath channels. Both simulation analysis and theoretical analysis prove the effectiveness of the algorithms. In addition, two efficient co-operative positioning methods are proposed in this paper, considering the lack of reference nodes in wireless sensor network scenario. The proposed methods improve the positioning accuracy with fewer links and achieve effective man-agement of the node energy and the network traffic. The main contributions are as follows:Firstly, this paper focus on TOA estimation for UWB ranging system, presents the deficiency of existing TOA estimation algorithms, analyzes the key factors that constrained TOA estimation accuracy, show the characteristics of received signal. Then, the checking window based TOA estimation algo-rithm and the error characteristic based TOA estimation method are proposed. Different from traditional TOA estimation algorithms, which acquires decision results with each energy sample comparing with a threshold independently, the proposed checking window based TOA estimation algorithm makes full use of the signal sample correlation in time dimension, and improves estimation accu- racy according to joint detection over multiple energy samples. Moreover, it is a kind of universal technique which can be combined with other TOA estimation methods. From another point of view, the proposed error characteristic based TOA estimation algorithm establishes the error type function by analyzing the relationship between the threshold and the detection errors. According to the error type function, an adaptive threshold is obtained by iterative processing. As compared with traditional TOA estimation algorithms, the proposed algo-rithm acquires an adaptive threshold through decision feedback, and it does not need a priori CSI and shows good anti-interference performance.Secondly, for the implementation of the network localization, cooperative positioning technology is further discussed in this paper. As is known to all, cooperative localization improves the positioning accuracy caused by insuffi-cient reference nodes. However, current cooperative localization algorithms show limitation in practical application, due to the high complexity, large en-ergy consumption. This paper decomposes the localization problems into two sub-problems:location estimation and information control. Particularly, an in-formation propagation controlled cooperative positioning scheme is proposed. According to the bootstrap percolation, only sensor nodes that satisfy the acti-vation conditions can implement location estimation and broadcast positional information. As a result, an effective management of energy and network traffic is realized. Moreover, because the proposed scheme reduces error propagation, improved positioning accuracy can be acquired.Finally, this paper extends the bootstrap percolation strategy to bayesian networks, and a hierarchical NBP localization algorithm is proposed. We refor-mulate the framework of bootstrap percolation controlled cooperative localiza-tion problem, and introduce the hierarchical graph model. Then, an information propagation rule is investigated based on hierarchical graph model. Finally, we present the corresponding hierarchical NBP localization algorithm. Simula-tion results verify the effectiveness of the proposed algorithm. According to traditional cooperative positioning algorithms, each sensor node estimates its position and broadcast its positional information without control, thus the al-gorithm convergence speed is slow or cannot be guaranteed. By controlling the node behavior and message propagation, messages flow in order with the proposed algorithm. Specifically, information always flow from sensor node of high confidence to sensor node with low confidence, thus effective energy management is achieved and error propagation is supressed.