Position Techniques In Mobile Wireless Networks And Position-based AD HOC Routing Strategy

In recent years, position location of a target in wireless communication systems has received significant attention. In military area, position location of a target is widely used in reconnaissance, weapon attacking, radar countermeasure, photoelectronic countermeasure, etc. Position location can provide accurate location information to the remote fire control system of army fighting vehicle, artillery, tank, aircraft, naval vessels and guidance system of cruise missiles and ballistic missiles. Thus appropriately utilize the limited electronic warfare resource and improve the operational efficiency of communication countermeasure, radar countermeasure and photoelectronic countermeasure. In the field of civil application, position location originated from the U.S. Enhanced emergency call service 911 (E-911), which required that for 67% and 95% of the cases the positioning accuracy must have a maximum average error of 100m and 300m, respectively. Nowadays, positioning techniques have become an attractive topic for both researchers and wireless service providers. There are many applications with mobile position, such as mobile yellow pages, navigation services, location-sensitive billing, restaurant/hotel finding, and so forth. Moreover, position location systems can provide many other public interest services, such as fleet management, roadside assistance, traffic routing, network design and resource optimization, etc. Besides, location positioning techniques are widely used in the field of"The Internet of Things", which will have a great influence on the world's future economy.Conventional positioning technique in wireless communication is generally based on radio communication and optical communication. Positioning parameters include the distance, pitch angle, received signal strength (RSS), time of arrival (TOA) or time difference of arrival (TDOA) and angle of arrival (AOA), etc. And the positioning parameters from multiple measurement terminals in line-of-sight (LOS) environment can be organized by employing some sophisticated positioning algorithms, such as triangulation positioning, circular positioning and hyperbolic positioning to give a final location estimation of the target. However, optical-based positioning parameters are sensitive to clouds, fog, spray dust and smoke screen, etc. When there is an obstacle between the measuring terminal and the target, it is difficult to get the positioning parameters by optical communication. Whereas the characteristics of positioning parameters in wireless communication depend heavily on the radio channel between the target and the measuring terminal. In a dense urban environment, there may be dense buildings blocking the LOS propagation path from a transmitter to a receiver. So signal may actually travel extra distance due to reflection and diffraction. Hence the position estimation will make the target appear further away from the measuring terminal than it actually is. The measured positioning parameters will accordingly have NLOS errors on the order of hundreds of meters and the location estimation of a target will also have a significant bias. Specifically, when the target is moving, NLOS errors may be intermittently introduced in the range measurements, which will provide great challenge to the robustness of location algorithms.There are four main positioning issues to be solved in wireless communication. First, the non-line-of-sight (NLOS) errors in parameter measurements need to be suppressed or mitigated. When the transmitted signal reaches the receiver by reflection or diffraction, the measured distance from the transmitter to the receiver may appear further away than it actually is. Compared with the system measurement noise, the NLOS errors usually have greater mean value and standard deviation. Therefore, NLOS error suppression and mitigation is an effective way to improve the positioning accuracy. Second, positioning parameters from different systems needs to be integrated. Positioning systems have different kinds of information, such as time, distance, distance difference, angle, signal strength, etc. And the location estimation of the target may exhibit significant differences when informations are explored with different algorithm. Therefore, data fusion of positioning informations from different systems is another efficient way to improve the positioning accuracy. Third, positioning technique in wireless sensor networks is a great challenge because of Complex geometry structure of node distribution, distance measurements error and sparse reference nodes, which are main concerns in wireless sensor network positioning. And it is difficult to obtain analytical solution of the node location by conventional positioning algorithms. Therefore, positioning of nodes in wireless sensor networks is an urgent problem. Finally, AD HOC routing algorithms need to be developed with node location information. Wireless AD HOC networks usually have dynamic network topology, while the route establishment is relatively slow and the network topology information will soon be rendered useless. Existing mechanism for routing searching, establishing and maintaining needs a lot of network overhead, which may lead to network congestion and packet loss. Therefore, constraining the routing space and reducing the network overhead with positioning technique is an important application in the field of wireless communication.This dissertation first introduces the application area, state-of-the-art research and key issues of positioning in wireless communication. The positioning techniques with electromagnetic communication and optical communication are discussed. And the corresponding positioning algorithms based on TOA, TDOA, AOA, RSS and optical range measurements are also investigated. The main factors affecting the accuracy of positioning are analyzed and some main performance evaluation indexes are given. Then two main error models in NLOS environment are studied. And a parameter reconstruction algorithm based on orthogonal polynomial fitting together with a positioning optimization algorithm is proposed. The performance of our proposed algorithms in urban, sub-urban and rural environment is also simulated and analyzed. At the same time, a NLOS error mitigation scheme using data fusion method and a total least square optimization technique are proposed. And the performance with various numbers of measuring points and different level of system measurement noise deviation is also simulated and analyzed. Moreover, positioning technique in wireless sensor networks is studied and finite element method is introduced for the first time to give location estimation of nodes, which is divided into three phases, i.e., initial positioning, location refinement and location optimization. Performance comparisons with different number of reference nodes are also given. Finally, the application of positioning technique in AD HOC wireless routing protocol is studied. The routing protocol structure, route searching and maintaining method combining location information and received signal strength are proposed and a routing optimization algorithm is given. Performance comparisons with other existing routing algorithms are simulated.The main contributions of this dissertation are:1. An error identification technique, a measurement reconstruction technique and a positioning optimization algorithm are proposed to suppress the errors in NLOS environment based on wireless location parameters. First, the error identification technique is proposed depending on a time series of range measurements. Then, the orthogonal polynomial fitting technique is used to reconstruct the measurements according to the results of error identification. Finally, the positioning optimization algorithm in NLOS environment is presented. Simulations are performed to calculate performance indexes, such as GDOP, MEE and Cramer-rao lower bound and give comparison results with other positioning algorithms.2. The data fusion positioning method based on optical parameters and electromagnetic parameters is studied and a NLOS error mitigation and location optimization algorithm is proposed. When there is a NLOS propagation path between the measuring point and a target, optical measuring platforms, such as portable individual distance measuring instruments, airborne range finder or laser radar are first used to give an estimation of the distance to the target. Then their location infromation are measured using electromagnetic-based measuring technique to eventually give a final positioning result of the target. A total least square method is also exploited to optimize the positioning results and improve the location accuracy.3. Positioning technique in wireless sensor network is studied. A positioning algorithm based on finite element method (FEM) and global optimization is proposed, which includes three phases, i.e., initial positioning, location refinement and location optimization. First, network nodes are initially located in global coordinate system by signal measuring model and range measurements. Then whole wireless networks are divided into several discrete elements and displacement model is established in each element. Thus location estimations of the network nodes are given by an area coordinate system and the positioning accuracy is improved. After that, a global optimization object function is constructed by weighted optimization method or residual optimization method and solved by gradient descent method to give an accurate location estimation of the target.4. Application of positioning technique in AD HOC wireless routing strategy is studied. Routing protocol structure, route searching method and maintaining method based on location information and received signal strength are presented to constrain the searching and maintaining space of a desired route and therefore reduce the communication overhead. Simulated annealing (SA) method is also explored to get a routing optimization. Finally, position-based routing protocol algorithm is simulated in a dynamic AD HOC wireless network and compared with existing on-demand routing algorithms.