Angles Of Arrival Estimation Based On Cyclostationarity Of Signals

Direction of Arrival (fort short AOA) estimation is a primary subject in spatial spectrum estimation field. It is applied markedly to military and civil economy fields, such as electronic countermeasures, mobile communication, sonar, etc.As the most classical DOA estimation method, MUSIC algorithm could estimate the DOA of signals agonicly with great efficiency, when the signals are no-coherent, narrowband and based on exact model. However, its resolving ability would decline, even be invalid, if the signals are correlative, coherent in especial, too close, or in a low SNR. The cyclostationary is one of the methods that can be used in the signal circumstance metioned above.Since most communication signals are cyclostationary, the signals which have the same cyclic frequency could be cyclic cross-correlated, whereas others could be cyclic cross-uncorrelated. Using the virtues of better signal selectivity, better resolution power, and being irrelevant to statistical characteristic of noise, the expected signals could be identified, and the coherent ones and noise could be shielded, with the cyclic correlation method.However, all of the Cyclic MUSICs algorithms need to search spatial spectrum in all areas. When the signals are in small areas, quantity of operation will be huge, and it's going to be worse, while estimating two-dimensional angles of arrival (fort short AOA).In this thesis, to advance the estimation efficiency of cyclic MUSICs algorithm, and to improve the estimation performance of two-dimensional AOA, following work has been done:Firstly, compaired with MUSIC algorithm, the virtue of the cyclic-MUSIC algorithm has been validated by the simulation results.And then, an improved cyclostationary fast algorithm with great resolving ability is presented, getting rid of the disadvantage of Cylic MUSICs algorithm, whose resolving ability for signal is not very ideal, and whose time of searching spatial spectrum pinnacle are too long. Taking advantage of FFT, such as lesser samples, rapidness, only the limited field where the signal exists is considered. Then, aftersearching spectrum pinnacle within the limited area, the DOAs could be obtained instead of searching spatial spectrum in all areas. Its good efficiency and great performance are illustrated by the simulation results.At the end, a three-dimensional array structure is designed, to simplify the complicated problem of the two-dimensional AOA estimation into three one-dimensional DOA estimation. Then, the simple two-dimensional AOA could be got through parameter matching. The algorithm based on FFT and cyclostationary is proposed. Simulation results show that the algorithm performs well in reduceing the complexity and improving efficiency.