Research And Implementation Of Indoor Positioning System Based On ZigBee Of WSN

In the field of indoor positioning, the limitation of Global Navigation Satellite System(GNSS) becomes more and more obvious. As the integration of positioning system and communication technology, indoor positioning extends the application area of personal location information and shows more extensive application requirements.Wireless Sensor Network(WSN) usually used to gather the sensory data of nodes.Furthermore, exploring data sources, confirming the location of nodes, and identifying the position of events also become important functions of WSN with its application in recent years. According to the characteristics of WSN, the researches to achieve indoor positioning functionality of nodes in WSN without additional cost, and to expand the application of WSN, has important application value and broad development space.This paper proposes a directly solved maximum likelihood estimation positioning algorithm based on Receive Signal Strength Indication(RSSI) to fits WSN positioning system. Considering path loss and shadow fading of narrowband wireless signal, it derives analytical models of target nodes and RSSI values of referenced nodes. This helps to avoid counting the distance between sending and receiving nodes directly, and obtains lower error and high localization accuracy. The numerical simulation is used to gives the impact factors on location accuracy like numbers of referenced nodes,blocking factors, path loss exponents and so on.A WSN positioning system based on RSSI is designed and a Zig Bee radio frequency chip CC2350 is used in this design. Hardware design, on-chip programs and interface programs development is presented elaborately. Several methods are performed to improve positioning accuracy of the system, such as improving the accuracy of RSSI values indirectly, setting RSSI value threshold, ensuring the reliability of RSSI values using timer, optimizing Z-Stack protocol stack.Actual measurements are performed to verify the algorithm, hardware and software.Comparing with the method establishing nonlinear equations for maximum likelihood estimation, the algorithm proposed in this paper can improves both location stability and accuracy of target node. In the case 4 meter square area with 8 reference nodes and the case 10 meter square area with 16 reference nodes, the tests show that average location errors can be reduced to 0.36 meter and 0.91 meter respectively. According the tests,increasing the numbers of reference nodes in location area, setting appropriate received threshold value of RSSI, placing reference nodes evenly around the target, are effective ways to improve the location accuracy.