Sub-band Adaptive Beamforming Array Techique for Wideband Signals without Decimation/interpolation Features

A novel method to implement a sub-band adaptive beamforming array (SABA) is proposed. The lower and the higher sub-band signals are generated from the incoming full band signal and both the subband signals are optimised directly instead of decimation or down sampling. After optimisation the outputs of both the sub-bands are combined without interpolation or up sampling to generate the desired signal. This SABA reduces the cost and circuit complexity, while maintaining better performance over the full band adaptive beamforming array. Introduction: Adaptive beamforming arrays (ABA) are known to be very powerful in aiding the performance enhancement for mobile and other communication networks. These arrays help in concentrating the intelligence in the direction of the radio frequency (RF) energy from the transmitting antenna, while suppressing any interfering signals from any other directions. It has been established that the error introduced (compared with the original signal) by the full band beamforming array (FABA) can be reduced by processing the received wideband signal by dividing them into a number of sub-bands [1, 2]. This sub-banding is done using the Quadrature Mirror Filter (QMF) banks for wideband applications. It has recently been proposed that the QMF sub-band adaptive beamforming array (SABA) can also be performed by optimising the higher sub-band output and then, multiplying the signal output with a multiplication factor that generates the lower sub-band output [3]. In this Letter, a new schematic for the SABA configuration is investigated. A reduction in the cost and the hardware complexity, the mean square error (MSE) and improved radiation responses have been expected for this configuration for wideband applications. Why QMF: The cancellation performance of a FABA can be corrupted if wideband signals are dealt with. This drawback can be eliminated if beamforming is applied over a number of frequency bins or sub-bands. The currently used, Fast Fourier Transform (FFT) filter has a drawback which results in poor cross correlation between these adjacent sub-bands which results in aliasing, due to the slower roll of the FFT filter. An alternate method to combat this undesirable effect is proposed by the introduction of QMF banks. These filter banks can be used to provide a sharp cut off of the frequency spectrum [4]. Proposed configuration: The outputs of the antennas of a FABA system are generally fed to the optimiser through a number of taps to increase the signal to noise ratio of the received wideband signals. The properties of these adaptive arrays are employed by varying the weight of each individual tap which is computed iteratively from the received data. However in a conventional SABA using the QMF Banks, the output of each antenna is connected to the input of the QMF Bank and the lower and higher sub-band outputs are decimated (down sampled) by a sampling or a decimation factor (which is equal to the number of the total sub-bands) and then as in the case of FABA they are fed to a number of taps and the optimisation is carried out individually for each sub-band to find the weight of the individual taps. After the optimisation, both the sub-bands are interpolated (up sampled) by the same factor and then combined at the output to generate the desired signal. In this proposal, modifications have been suggested to the above SABA configuration to remove or to bypass the down sampling and the up sampling features, so that the output of the QMF bank can be connected directly for optimisation to find the weight of the individual taps. This results in the reduction of the number of the total components needed for the conventional SABA and hence results in the reduction of the total cost of the system. If one wants to keep the down and up samplers, with the SABA hardware system as an option for other applications, then the removal or bypassing of the down and up samplers can be done mathematically by