Algorithms-based beamforming for a narrowband signal received by an antenna array

• Published in: Measurement : journal of the International Measurement Confederation Vol. 235; p. 114816
• Main Authors: Djungha Okitadiowo, John Peter, Lay-Ekuakille, A., Massaro, Alessandro, Isernia, T., Urooj, S., Srinivasa Rao, K.
• Format: Journal Article
• Language: English
• Published: Elsevier Ltd 01.08.2024
• Subjects: Antenna array; Beamforming; Deviation; Interference; Measurements for telecommunication; Noise; Sensors; Smart antenna; Suppression; Measurements for telecommunication; Beamforming; Noise; Suppression; Interference; Smart antenna; Sensors; Deviation; Antenna array
• ISSN: 0263-2241
• DOI: 10.1016/j.measurement.2024.114816

Maintaining a good trajectory and the transmission of signal towards the target, is a very delicate subject in the field of transmission. In particular, wireless transmission which is permitted on a sensible physical medium to guide the trajectory of the wave, is the greatest concern of radio transmission. Following this challenge, we propose approaches to the applicability of the Beamforming (BF) techniques and Convolutional Neural Network (CNN) on a signal received from an antenna array. The approaches consist of monitoring and driving the signal trajectory from the antenna output to the device target detecting and removing all interferences which may deflect the signal from the desired direction by pivoting it in another one (around 45 degrees). The work is based on experimentation with BF approaches such as the MVDR (minimum-variance distortionless-response), and the LCMV (linear-constraint minimum-variance) to optimize the transmission phase. We chose this BF technique above others because; the LCMV technique detects and reduces interference signals efficiently by utilizing calculated weight vectors to direct the powerful beam-based signal towards its intended target. The LCMV technique also ensure that each sensor outcomes are filtered with a finite impulse response (FIR) filter to satisfy narrow-band signal restrictions. For accurate signal transmission and reception with noise and interference suppression, LCMV seems promising. We use LCMV to cancel out interference and maintain the desired signal. When noise is not separable from the data (signal), an estimate of the sample covariance matrix is obtained from the data. Some approaches are taken to specify constraints, such as amplitude and derivative constraints. For example, specifying weights that suppress spurious signals from a particular direction while transmitting signals from a different direction without distortion. To avoid self-cancellation of signals, the LCMV permits multiple constraints to be placed along the target direction (directional vector). The last developed (marginal) approach is about the application of a 24-layer CNN model exploiting in depth all features and components of the Device-To-Device (D2D) signal. Three different sizes of convolution kernels are using (32, 64 and 128) to exploit the original signal which is noisy to remove the noise on the signal and to recover the desired signal. The scientific merit and novelty value of this paper is to allow us to analyse and understand how MVDR and LCMV filters identify and reduce interference in new expressions for their output signal-to-interference ratio (SIR) and signal-to-noise ratio (SNR). We showed the trade-off between noise reduction and interference rejection.