Microwave sampling beamformer with finite pulse train

Abstract

Using microwave beamforming (MBF) reduces the cost, size, weight and battery power consumption of the adaptive beamformer structure by lowering the number of RF down-converters and analog to digital converters (ADC). However, there have been two problems associated with traditional MBF. One problem is having only one signal at the output that limits the signal processing capability of the structure. Several perturbation-based algorithms have been proposed for this purpose. Another problem the need for multiple phase shifters and amplitude control circuits which are the most expensive part of these systems. In (Farzaneh and Sebak, 2006), a new MBF structure has been proposed which independently accomplishes full range phase-amplitude control using one simple block by sampling in the microwave domain. This technique reduces the implementation cost of the MBF and also makes it possible to obtain high resolution weightings using high resolution time delay and pulse width modulation ICs. However, the structure in (Farzaneh and Sebak, 2006) is based on an infinite periodic pulse train. For adaptive beamforming (ABF) and direction of arrival (DOA) estimation, where weights are updated temporarily, the sampling pulse train is finite. In this work, the effect of the finite duration pulse train is investigated. A new limit for the sampling frequency is obtained which is a function of signal bandwidth and pulse train duration. In the next section the fundamental principle of the MSBF is introduced and the new structure is discussed. Then, simulation model and results are presented. Finally, summary and conclusions are given.