Research On The Full Pass-band Signal Combining Techniques For Multiantenna Signal-reception

In recent years, human beings have made many great achievements in the field of deep space exploration. Along with the development of deep space exploration, the way of deep space communication is undergoing tremendous changes because of farther detection distance. Multiantenna signal combining technology(MSC) emerges as the time requires, which has become an important branch of reception of weak signal in the deep space communication. Compared with the reception of large and single antenna, MSC can receive weak signal more effectively and can be in more robust work mode. Due to many advantages of MSC, to which many space powers have attach importance.This dissertation focuses on the application of MSC on the multi-signal condition, researches on the problem of joint time delay and phase alignment as well as the multi-signal combining problem. The contributions obtained in this dissertation can be summarized as follows.Aiming at the weak-point of MSC in integral multi-signal combining, this dissertation firstly designs a frame of the full pass-band multi-antenna signal combining technique(FPMSC) and then an algorithm of SIMPLE based joint time delay and phase alignment for combining(SIM-JTPAC) realized in frequency domain is proposed. Since the traditional MSCT could not be used for integral combining of multiple signals, the FPMSC and its framework are proposed. On condition that the target signal spectrums are known, an algorithm of FPMSC based on spectrum is proposed. The method uses an idea of frequency domain max ration combining(FMRC) and inherent linear property of phase difference of different spectrums. And the algorithm is named SIM-JTPAC. Simulations show that the SIM-JTPAC is available to combining of multiple signals. On condition of enough high signal-to noise ratio(SNR), the SNR gain of combined output tends to the theory bound. Moreover, SIM-JTPAC algorithm is also considered as an algorithm of joint time delay and phase alignment.Aiming at the problems of phase wandering and random phase center in SUMPLE, an improved algorithm, MSUMPLE, is proposed to avoid the problems. In MSUMPLE, the weights are obtained by cross-correlation of each antenna with the sum of the other antennas, but the phase of the array will track the phase of a selected antenna. Thus MSUMPLE overcomes the weakpoints of SUMPLE. Besides, theoretical derivation indicates that MSUMPLE is identical with SUMPLE in the rate of convergence and the precision of phase alignment. Simulation shows that the convergence phase of MSUMPLE is predictable and effectively stabilize the convergence phase.Aiming at the bad performance of SIM-JTPAC in low SNR circumstances, a new algorithm of MSUMPLE based joint time delay and phase alignment for combining(MSUM-JTPAC) realized in the frequency domain, in which the idea of MSUMPLE is introduced. The spectrum of received signal is regarded as a sum of many frequency-signals, and then MSUMPLE algorithm is used for each frequency-signal, which can ensure the inherent phase property in some sense. However, the combined performance maybe not be satisfying thanks to the different alignment precision of phase of each frequency-signal. Considering that the phase of spectrum is linear in the frequency band of received signal, Least-squared method is used to fit the phase of weight vector obtained in MSUMPLE to make the phase of weight vector coincide with the phase of spectrum. We call the algorithm to be MSUM-JTPAC. Stimulations show that both MSUM-JTPAC and MSUMPLE have a good performance even in the low SNR circumstances, the former is better than the later because of the usage of inherent linear phase property.A full pass-band multi-antenna signal combining parallel processing system is designed and implemented. For testing combined performance of this system, an 8-channel test platform is set up. For large amount of data and long processing time in FPMSC, the overall scheme of parallel processing system for FPMSC is designed. The modular decomposition of the scheme is implemented and the parallel design of each modularization is implemented. We build an 8-channel test platform to verify the performance of the algorithm proposed, and the actual tests show that it has a good combined performance.