A Research On The Algorithms Of Direction Finding With Uniform Circular Arrays

A uniform circular array (UCA) of antennas has a comprehensive perspective in the military and civilian applications. The research of direction finding algorithms based on UCA under the different signal models plays an important role in the practicability of the UCA of antennas.This dissertation first introduces the basic theories and algorithms of the spatial spectrum estimation and subsequently investigates the primary problems of the UCA output data model and its beamspace processing. By combining the cyclostationarity with UCA and predigesting its array manifold via narrow-band assumptions, we bring forward two efficient methods of direction finding with UCA: the beamspace Cyclic MUSIC and Cyclic-RMUSIC algorithms. Both of them have the properties of interference and noise suppression and can greatly reduce the computational complexity by exploiting FFT or rooting techniques.Then, the uncertainty of the optimal time lag and its effect on the direction-of-arrival (DO A) estimation in the conventional Cyclic MUSIC algorithm are studied in detail. The limitations of the two improved methods with great computational load are analyzed in succession. When the signal sources have different cycle frequencies, a method to estimate the optimal time lag is presented. It solves the severe problem that the conventional Cyclic MUSIC requires knowing the optimal time lag in computing the cyclic correlation matrix. A method of two-step search for DOA estimation by employing an interleaving technique and a Joint Estimation Strategy is developed to overcome the inadequency of signal sources with the same cycle frequency having no optimal time lag. We propose a modified Cyclic MUSIC algorithm according to the signal types. It improves the performance with a relatively small increase in the computational load.In the end, we discuss the direction finding algorithms of UCA for the coherent signal sources and the wide-band cyclostationary signal sources, respectively. Under the frame of forward/backward spatial smoothing, several algorithms for the coherent signal sources are united in the uniform linear array (ULA), and then extended and used in UCA. By constructing a preprocessing matrix via the phase mode excitation technique, we apply the UCA to estimate the DOA of the wide-band cyclostationary signal sources.Under the background of employing UCA for direction finding, this thesis investigates the corresponding algorithms for the stationary, cyclostationary, stationary coherent, and wide-band cyclostationary signal models and also analyzes their performance. The research will provide technique references for the design of the airborne array of antennas with super resolution.