A novel array signal processing technique for multipath channel parameter estimation

Many important application areas such as mobile communication, radar, sonar and remote sensing make use of array signal processing techniques. In this thesis, a new array processing technique called Cross Ambiguity Function - Direction Finding (CAF-DF) is developed. CAF-DF technique estimates direction of arrival (DOA), time delay and Doppler shift corresponding to each impinging signals onto a sensor array in an iterative manner. Starting point of each iteration is CAF computation at the output of each sensor element. Then, using incoherent integration of the computed CAFs, the strongest signal in the delay-Doppler domain is detected and based on the observed phases of the obtained peak across all the sensors, the DOA of the strongest signal is estimated. Having found the DOA, CAF of the coherently integrated sensor outputs is computed to find accurate delay and Doppler estimates for the strongest signal. Then, for each sensor in the array, a copy of the strongest signal that should be observed at that sensor is constucted and eliminated from the sensor output to start the next iteration. Iterations continue until there is no detectable peak on the incoherently integrated CAFs. The proposed technique is compared with a MUSIC based technique on synthetic signals. Moreover, performance of the algorithm is tested on real high-latitude ionospheric data where the existing approaches have limited resolution capability of the signal paths. Based on a wide range of comparisons, it is found that the proposed CAF-DF technique is a strong candidate to define the new standard on challenging array processing applications.