点击次数：

论文类型：期刊论文

发表时间：2018-06-01

发表刊物：IEEE INTERNET OF THINGS JOURNAL

收录刊物：SCIE

卷号：5

期号：3,SI

页面范围：1625-1636

ISSN号：2327-4662

关键字：Atomic norm; compressed sensing; intelligent computing; Internet of Things (IoT); optimization; target tracking

摘要：In the new era of integrated computing with intelligent devices and system, moving aerial targets can be tracked flexibly. The estimation performance of traditional matched filter-based methods would deteriorate dramatically for multiple targets tracking, since the weak target is masked by the strong target or the strong sidelobes. In order to solve the problems mentioned above, this paper aims at developing a joint range-Doppler-angle estimation solution for an intelligent tracking system with a commercial frequency modulation radio station (noncooperative illuminator of opportunity) and a uniform linear array. First, a gridless sparse method is proposed for simultaneous angle-range-Doppler estimation with atomic norm minimization. Based on the integrated computing, multiple work-stations or servers of the data process center in the intelligent tracking system can cooperate with each other to accelerate the data process. Then a suboptimal method, which estimates three parameters in a sequential way, is proposed based on grid sparse method. The range-Doppler of each target is iteratively estimated by exploiting the joint sparsity in multiple surveillance antennas. A simple beamforming method is used to estimate the angles in turn by exploiting the angle information in the joint sparse coefficients. Simulation result and real test show that the proposed solution can effectively detect weak targets in an iterative manner.