Abstract-This paper addresses the problem of direction of arrival and time delay estimation, and derives multi-invariance MUSIC (MI-MUSIC) algorithm therein. The proposed MI-MUSIC, which only requires one-dimension searching, can avoid the high computational cost within two-dimension MUSIC (2D-MUSIC) algorithm. It means that MI-MUSIC algorithm has better performance than that of ESPRIT and MUSIC, and also can be viewed as a generalization of MUSIC. Simulation results verify the usefulness of our algorithm.

Abstract-This paper derives its link of direction of arrival (DOA) and direction of departure (DOD) estimation problem for multiinput multi-output (MIMO) radar to the trilinear model. With the exploitation of this link, we present a trilinear decompositionbased blind algorithm for angle estimation in MIMO radar. After the algorithmic presentation and discussions for identifiability results, trilinear decomposition-based angle estimation under different array conditions has also been investigated. Our proposed algorithm works well without either spectral peak searching or pair matching, has better angle estimation performance than ESPRIT, and even supports small sampling sizes. Simulation trials testify that the merits of our algorithmic performance are in collaboration with theoretical findings.

Abstract-Blind direction of arrival (DOA) estimation algorithms of coherent sources using multi-invariance property is presented in this paper. ESPRIT-like algorithm in [23] can estimate DOA of coherent signal, but its performance is without satisfaction. We reconstruct the received signal to form data model with multi-invariance property, and then multi-invariance ESPRIT and multi-invariance MUSIC algorithms for coherent DOA estimation are proposed in this paper. Our proposed algorithms can resolve the DOAs of coherent signals. They have much better DOA estimation performance than ESPRITlike algorithm. Meanwhile they identify more DOAs than ESPRIT-like algorithm. The simulation results demonstrate their validity.

No abstract

This paper investigates the joint detection method for TD-SCDMA system, and derives an improved approach to eliminate inter-symbol interference (ISI) in condition of multipath channel. Analysis for the received signals exhibits its characteristics of PARAllel profiles with LINear Dependencies (PARALIND) model. The proposed blind PALALIND algorithm utilizes no channel parameter for detection, and performs better with snapshots increasing. Simulation results indicate that its performance is quite close to space-time minimum-mean-squareerror (ST-MMSE) and also robust in array error condition.