Broadband beam forming method of variable step length LCMV-LMS algorithm

Abstract

The invention discloses a broadband beam forming method of a variable step size LCMV-LMS algorithm, broadband signals are converted to a frequency domain through fast Fourier transform, independent beam forming is carried out on narrowband signals under each frequency point by adopting multiple groups of filters, and meanwhile, the variable step size LCMV-LMS algorithm can reduce the operation complexity on the premise that reference signals are not needed, so that the beam forming efficiency is improved. The balance between the convergence speed and the power steady-state error is realized, and the anti-interference filtering of the broadband digital signal is satisfied. According to the method, the least mean square algorithm based on the linear constraint minimum variance criterion is popularized to broadband signal beam forming, matrix inversion of each narrow band is avoided, the algorithm principle is simple, and occupation of hardware resources can be effectively reduced. The step length factor updating formula provided by the invention only comprises one division operation, the denominator is integer power of 2, and only shifting operation needs to be carried out in logic hardware such as a field programmable logic gate array.

Description

technical field [0001] The invention relates to the technical field of signal processing, in particular to a wideband beamforming method of a variable step size LCMV-LMS algorithm. Background technique [0002] In a wireless communication system, the reliable and accurate reception of signals at the receiving antenna is one of the important guarantees for the overall system performance. Beamforming technology achieves the purpose of suppressing interference and retaining the desired signal information to the greatest extent by forming the main lobe in the specified direction in the airspace to align with the desired signal, effectively improving the signal-to-interference ratio of the received signal. [0003] In recent years, thanks to the development of digital signal processing chips such as DSP and FPGA, beamforming technology has been widely used in practical engineering, from the initial conventional beamformer (Conventional Beamformer, CBF) to the minimum variance Wi...

AI Technical Summary

Problems solved by technology

[0004] Although the above-mentioned achievements have been made in the industry, there are still some problems in the current beamforming technology: most algorithms are based on linear arrays, and the shortcomings of linear arrays include small scanning range and large space occupation, so it is necessary to expand the beamforming algorithm to the current mainstream The planar array antenna; the traditional LMS algorithm performs convergence iterations based on the minimum mean square error criterion, which requires a priori condition of the reference signal, so it is not suitable for non-cooperative communication scenarios and adaptive filtering of broadband signals, application limitations bigger

However, when this criterion is directly applied to broadband signals, the deviation of the steering vector and the distortion of the array output will occur due to the large frequency difference of signals at different frequency points; the traditional LMS algorithm uses a fixed iterative step factor and cannot achieve convergence. Balance between speed and power steady state

In response to this problem, some scholars have proposed a variable step-size LMS algorithm, but the step-size iteration formula is relatively complicated; in addition, once the fixed step-size factor of the traditional LMS algorithm is selected too large, the convergence curve will diverge and the output will not meet expectations.

Images