Wireless Sensor Network Positioning Technology Research

Recent years,electronic and communication technology has brought wireless sensor network great impetus and broad future applications.Wireless sensor gadget has many features:low power consumption,small size,multi-functional etc.It fulfilled the need of more data collection and short-range wireless communication.All these technical features enormously expanded the creative applications of wireless sensor network.Wireless sensor network localization was one of the current attractive applications in the following fields:military,healthy care,business,Home environment and nature environment protection.The state of art research in this field had focused on the algorithm, measurement technique and application solution.Different with the conventional wireless network,the wireless sensor network's advantage lied on the ad-hoc mode within the totally random topology.In the extensive distributing network the small node could connect with each other through the wireless multi-hop network.It's a challenge to design the positioning system on such complicate environment. First,positioning algorithm has its limits on the accuracy.How to acquire more accurate position data and measurement in the current signal measurement modal should be carefully considered.How to give it more improvement based on other localization algorithm.Second,how to design the positioning protocol properly and make it more efficiency.It's very urgent to scheme out the low power consumption and low data obstruction protocol.The third aspect is working out an appropriate hardware platform to run the algorithm and protocol which many researcher and company are concentrating on.Confronting all the problems introduced above,the main work in this paper were focus on the several areas:1.The wireless sensor network localization error bound - Cramer-Rao Low Bound (CRLB) was studied.There had been proposed one-hop positioning CRLB estimation which based on the radio signal strength(RSS) modal.But the multi-hop environment is distinguishing where the error propagation should be considered.We provided an estimation method which took the propagation error into account and derived the RSS-based multi-hop CRLB formula.Further more,we applied the CRLB to the localization algorithm:Distributed Weighted Multidimensional Scaling.We used it to determine the weight updating on the process of lateration and finally got better result.The weight previously was determined only by the measurement error but not concerning about the reference point error.We could obtain better result due to the usage of reference error estimation which was CRLB estimation referred above.2.We design a localization protocol which was concentrating on its positioning coverage and convergence speed.Many nodes in wireless sensor network can't figure out their determined position because of lacking enough reference information,especially the nodes which were on the margin of the network.This paper proposed an inference method to eliminate the ambiguous of the node location.This method not only used the node-pair distances to calculate,but also inferred the nodes topology and greatly increased the localizable nodes.Besides,we also designed a fast distributing protocol to avoid the data obstruction and make it convergent more rapidly.3.The hardware platform realization.This paper chose common low power solution: Tl's MSP430 series and CC10XX series wireless module as the main board.At the same time,the power system was carefully designed to meet the need of independently currency sensing and monitoring.We also used part of TinyOS's(one kind of network operation system) module and introduced how to read the RSS register to calculate the transmission power of the wireless module.