| The opportunities and challenges for the development of future communications technologies are provided in the rapid development of wireless communications and its huge market.In the wireless communication research,many problems need to be solved urgently,such as resource limitation,wireless channel interference,multi-user access,navigation and positioning.Wireless signals of high-resolution and high-accuracy pose great challenges to both software and hardware of wireless communication systems.Ultra-wideband systems,while transmitting large amounts of data at high speeds,are prone to multi-path interference and high bit error of the transmitted signal,which eventually results in signal attenuation.In the thesis,the performance of Ultra-Wideband(UWB)systems from three aspects: UWB channel communication detection,UWB multiple access and UWB positioning are studied.In the thesis,the UWB communication systems are divided into three layers,namely the physical layer,the middle layer,and the application layer.Regarding the physical layer,the data sharing problem and the performance parameter estimation problem of multi-cluster channel signals are studied,and a multi-task compressive sensing method combining hierarchical Dirichlet process and Markov chain algorithm are proposed and verified.Regarding the middle layer,the multi-user access and anti-interference performance of UWB systems are mainly studied,and a hierarchical shrinkage Bayesian compressive sensing algorithm and multi-user interference algorithm are proposed and verified.Regarding the application layer,the UWB systems are applied in the research of indoor location and the accuracy and anti-interference of indoor location estimations are analyzed and detected.A multi-task compressive sensing algorithm is proposed to reconstruct the sampled signal,and the weighted least square multiplication is used to estimate the location.The main research results of the thesis are as follows.1.Physical layer: An UWB channel estimation method based on multi-task compressive sensing has been proposed.In the physical layer,concerning the problem of transmission signal error and multicluster signal sharing in UWB channel,the performance of UWB channels is estimated and detected based on multi-tasking Bayesian compressive sensing technology,and the signal is reconstructed thereby.First,both the architecture of the UWB compressive sensing systems and a compressive sensing feature dictionary are es-tablished and the signals are reconstructed through the feature vector.Second,the hierarchical Dirichlet process is used to solve the data sharing issues among multiple tasks in UWB system communication,and the signal performance parameters,namely the peak noise ratio and the normalized mean square error of the signal,are calculated,while reducing the signal transmission time.Finally,the channels of the UWB communication systems are simulated using standardized IEEE802.15.4a channel model in the line-of-sight and non-line-of-sight environments,respectively.Given the measurement of the signal,both the peak noise ratio and the mean square error obtained by the proposed algorithm have been improved compared with previous algorithms.The experimental results verify that,through the application of this algorithm in detection and analysis of the channel transmission signal,the receiver can receive more accurately the signal from the sender,and the time required for signal transmission is less than that of other algorithms.2.Middle layer: A performance detection algorithm for UWB systems based on hierarchical reduced Bayesian compressive sensing is proposed.In the middle layer,in view of the issues of multi-path access and user noise inference,a multiple address access and noise interference algorithm combining the Monto Carlo algorithm and compressive sensing is proposed.First,establish the measurement modeling equations and apply the hierarchical reduction algorithm of tree structure and the Dirichlet process to simplify the wavelet coefficients and reduce the computation complexity.Second,the UWB system model of TH-PPM(time hoping-pulse modulation)and the multi-user interference model are constructed.The hidden Markov chain and the Monto Carlo algorithm are combined to detect the anti-interference performance of the UWB multi-user inference systems,and the error rate and mean square error of signals are calculated.Third,by using of the above bit error rate and mean square error,the performance of the proposed algorithm is analyzed and compared with other algorithms.The simulation results show that its better performance on the noise suppression than other non-tree structure compressive sensing algorithms.As the number of users increases,the proposed algorithm performs much better on the anti-interference and noise suppression than other algorithms.Through comparisons with different cases of multiuser access to the UWB systems for 20,50,100,300,500,and 1000 users at the same time,the algorithm has been achieved better performance in anti-interference and lower error probability than other algorithms.3.Application layer:An UWB indoor multi-path positioning algorithm based on multi-task compressive sensing is proposed.In the application layer,on the problem of the UWB indoor positioning,the multitask compressive sensing technology is used to reduce the signal sampling cost and improve the accuracy of positioning.First,the sampled signals are recovered and reconstructed using the intrinsic sparsity and the multi-task Bayesian compressive sensing algorithm of multi-path UWB channels.Second,the time difference arrival algorithm and the maximum likelihood algorithm are used to estimate the channel impulse response and delay time,and the distance between the target node and the reference node is detected.Third,we simulate the UWB indoor positioning algorithm,in which the multi-task compressive sensing algorithm uses the different signal samples as the measurement matrices.Under the indoor multi-path non-line-of-sight environment in the rectangular region,the UWB signal model is established,and the positioning accuracy of the proposed algorithm is analyzed and compared with other algorithms.The simulation results show that the signal with lower sample rates can approximate the initial signal of the UWB channels.The positioning error accuracy of the proposed algorithm is lower than other positioning algorithms.Moreover,a square area of 100 m ×100m is established,and the indoor positioning accuracy of the proposed algorithm is compared with the least squares method.The simulation results show that as the ranging error increases,so does the average positioning error of the least squares method,while the average positioning error of our algorithm decreases as the ranging error increases and,moreover,is inversely proportional to the number of reference nodes.Therefore,the indoor positioning accuracy of the proposed algorithm is higher than that of the least square method. |
