Research On Closed-form Resource Allocation For Wireless Distributed Radars With Cognitive Tracking

Wireless distributed radar(WDR)is one of the new radar systems consisting of multiple geographically distributed radar nodes.The WDR achieves collaborative detection by using wireless communication channels and distributed fusion architectures to implement the information processing.Compared to the traditional wired and centralized radar network,wireless data transmission can provide a more extensive deployment range and higher configuration freedom for the radar network.Meanwhile,distributed information processing can significantly reduce the dimension of fusion data,thereby reducing the requirements of communication bandwidth or increasing the capacity of the networking nodes.Based on the above characteristics,the WDR system can be applied to the typical scenarios of multi-sensor collaborative detection,such as the fighter/ship formations,drone swarms,and vehicle-road collaborative automatic driving.The extensive potential applications make the WDR system significant for the military and civilian applications.The WDR system occupies a high degree of freedom in node deployment,configuration transformation,and adjustment of the operating parameters.In this case,how to effectively use the prior information of the targets to be detected/tracked to guide the adjustment of radar working parameters,estimation of target parameters,and the information feedback to realize the closed-loop scheduling of the radar system resources has become one of the popular researches in the field of multi-radar collaborative detection.In the resource scheduling work for wireless distributed radar,the resource optimization criteria,multi-radar information fusion architecture,and data transmission of wireless channels are the key factors affecting the radar resource utilization as well as the task performance.They are also the frontier and challenging issues in this field of research.This dissertation focuses on the target tracking task and carries out the systematic work,such as theoretical analysis,method research,and simulation experiments,around the above three factors.The main contributions are as follows:1.The design of the resource optimization criteria is studied under the typical mission requirements.Two criteria of differentiated resource scheduling and robust power allocation are proposed,and the objective functions of the corresponding resource allocation problem are designed accordingly.The effectiveness of the designed criteria is then demonstrated by applying them to the typical radar systems,including C-MIMO and phased array radars.The proposed criteria lay the foundation for the subsequent research of the resource scheduling for WDR.2.Two typical distributed fusion architectures for WDR,i.e.,the hierarchical and consensus architectures,are formed.The corresponding flows of the distributed signal processing are proposed to make the data fusion phase be efficiently implemented.An approximate expression of the PCRLB is derived based on the EMD produced by distributed fusion to provide an more accurate quantification of tracking performance under the distributed fusion architecture.Based on the work above,the resource allocation problems regarding the WDR's transmitted beam and power are modeled.The algorithms,including the RIDA and CW-ADMM,are proposed to solve the problems efficiently.It is shown that the proposed algorithms can achieve the rational resource allocation results regarding different mission requirements and obtain a significant improvement in tracking performance under the distributed fusion architectures.3.Under the wireless communication channels with the limited bandwidth,to efficiently estimate the target state,a closed-loop signal processing framework for cognitive tracking is proposed based on an approximate channel-aware particle filter.A design problem for data quantization thresholds of the WDR system is established and further modeled as an optimization problem.Then,according to the statistical characteristics of different channel models,the problem is solved using the analytical calculation and M2 BGD algorithm.Compared to the fixed quantization threshold,the proposed method can significantly improve the tracking performance without consuming additional transmission resources.Theoretical derivations and numerical simulations have been implemented to verify the performance of the above models and algorithms.The results demonstrate that the proposed methods can achieve a dynamic optimization for the transmitted resources of WDR and improve the overall tracking efficiency further.