Rearch On The DOA Estimation Algorithm Of Weak Signal Under Strong Interference

DOA (Direction of Arrival) finding is an interesting topic in the field of signalparameter estimation that arises in military-and civil-applications. Spectral-basedalgorithms are able to resolve two closly-spaced incident signals, which makes thembe the mainstreams of DOA estimates, however, they are effective for only signalswith the equal power. When strong interferences and weak signals coexist in thesame environment, those approaches will degradate so significant that even failuremay occure for weak signals. With the the increasing electronic environments, thecoexistence of strong interference and weak signal is inevitable. Therefor, whenstrong interference coexist, researched for improving the estimation performance ofweak signal is of great significance.This paper firstly studies the influnce incident signals with unequal powers onthe performance of MUSIC algorithm. It is shown by numerical simulations that thecoexistence of strong interference leads degradation on the performace of theMUSIC algorithm, sometimes even gives failure for weak signal. There are twopopular ideas for estimating the DOAs of weak signals when strong interferencesexist. The first idea focus on finding the DOAs of weak signals and those of stronginterferences simutaneously, and the typical algorithms using this idea include theRELAX algorithm, DOA estimation algorithm based on noise subspace’s invariance,etc. The other idea aims to suppress the strong interference firstly, and then estimatethe DOAs of weak signal sencondly. Typical algorithms with this idea are thesubarray interference suppression algorithm, the JJM (Jamming Jam Method)algorithm, etc. Simulations of the above mentioned four algorithms are done, basedon ULAs (Uniform Linear Arrays), which show that the JJM algorithm and DOAestimation algorithm based on noise subspace’s invariance have better estimationperformance for weak signals. Howere, the JJM algorithm is based on theassumption that the array manifold matrix is of the Vandermonde structure, whichseverly limits its application to arbitrary arrays. On the other hand, DOA estimationalgorithm based on noise subspace’s invariance can not suppress interference.To overcome the drawbacks of the JJM algorithm and DOA estimationalgorithm based on noise subspace’s invariance, two modifications are proposedwhich are referred to as the modified JJM(modified-JJM, M-JJM) algorithm and themodified algorithm based on noise subspace’s invariance, respectively. The M-JJMalgorithm makes the assumption that the array is made up of two or more than twosubarrays with the same structure, where the geometries of the two subarray can bearbitrary, leading to suppress on the interference by joint processing among thesubarrays. The modified algorithm based on noise subspace’s invariance uses the basis vectors of strong interference subspace to construct a transformational matrix,with wich the steering vector is improved and the strong interference is suppressed.Simulation results verifies the effectiveness and the feasibility of the two modifiedalgorithm.Besides, since the M-JJM-and the JJM algorithms are both based on a priorinformation regarding the DOAs of strong interferences, whic may be hard toestimated accurately, a new method named the null broadening M-JJM technique isalso proposed to sovle this problem. Simulation results demonstrate the feasibilityof the algorithm.