Key Technologies Of Large-Scale Indoor Ultra-Wideband Localization Network

Due to the narrow pulse,Ultra-wideband(UWB)has the advantages of anti-multipath,high ranging precision and penetrability.UWB is considered as a promising technology for high-accuracy localization.Especially for enterprise-level applications,the localization system often needs large-area coverage for large-scale factories or tens of thousands of square meters of logistics sorting centers to achieve seamless and precise positioning.Therefore,this dissertation mainly focuses on large-scale regional extensible UWB localization networks.For low-power Real-Time Location System(RTLS)deployments,the advantage of the Time Difference of Arrival(TDOA)solution is that the tag is located by simply sending one message,so TDOA is suitable for higher node density deployment.The UWB TDOA indoor single-area localization system has strong environmental dependence and limited effective range of localization.In order to cover larger and more complex area,the system will be extended to more than four anchors.The clock synchronization between TDOA anchors is realized by using a wireless clock signal,which facilitates multiple anchors layout and makes the layout more flexible.The major innovations of this dissertation are summarized as follows:1.The single localization area adopts the master/slave anchor mode.The master anchor sends a clock synchronization signal,which is received by the slave anchors.Once the tag localization signal arrives at the anchors,the anchor calibrates the the positioning signal arrival time according to the master-slave clock offset condition.Clock tracking can be performed by digital phase-locked loop method or Kalman filter method.The synchronization performance of the two methods tracking UWB clocks is compared.It is concluded that Kalman filter converges faster.Kalman filter strongly depends on the measured value,and the speed of the error correction is slow.For these shortcomings,the algorithm is improved and compared with simulation.The improved algorithm is more robust when large errors occur in clock synchronization.2.The large-scale system firstly clusters and selects the cluster heads as the master anchors in each cluster.The intra-cluster clock synchronization uses master/slave mode,and the inter-cluster anchors are in master-relay-master data transmission mode.An opportunity routing clock synchronization signal transmission routing method is proposed,and the vMISO-OR relay mode is adopted.The effectiveness of the proposed method is verified by the simulation of routing algorithm and the comparison of the estimated distance difference with the actual distance difference.3.For the mobile tag,it is not necessary to specify the anchors in advance,but it can dynamically select the anchors in the range of the mobile tag in each location estimation process.The proposed method clusters the anchors according their geometrical location and select the optimal anchor pair in different clusters.The included angle of the hyperbola is controlled by the included angle of the center line of the optimal anchor pair.What's more,the detection distance of the tag is taken into consideration.By considering the geometric location relationship between anchors and mobile tag,the location estimation error and localization data redundancy can be effectively reduced,so that mobile tags can be seamlessly switched between the areas.In this way,the error of position estimation and redundancy of location data are effectively reduced,and the mobile tag can be seamlessly switched between areas.