Universal Journal of Electrical and Electronic Engineering Vol. 7(1), pp. 19 - 27
DOI: 10.13189/ujeee.2020.070102
Reprint (PDF) (497Kb)


MATLAB-Based GPU Acceleration for Multiple-Input and Multiple-Output Radar Beamforming Algorithm


Mostafa Hefnawi *, Gillian Rideout
Department of Electrical and Computer Engineering, Royal Military College of Canada, Canada

ABSTRACT

Multi-input and multi-output (MIMO) radar systems possess numerous advantages, such as scanning an entire region much faster than a phased array, enhancing the spatial resolution, mitigating interference and multipath fading, and improving the probability of detection of targets. MIMO radar systems use adaptive beamforming techniques such as the minimum variance distortionless response (MVDR) to obtain the best possible estimation of the direction of arrival (DOA) of the targets. The MVDR algorithm has been thoroughly investigated for traditional phased array radars. For MIMO radar systems, however, it requires significant signal processing, which can introduce substantial latency, especially in the case of large MIMO systems where few hundred antennas are used; this can be solved by using a Graphics Processor Unit (GPU), which contains thousands of cores and can execute many operations in parallel. This paper presents a MATLAB-based approach for GPU parallelization of the minimum variance distortionless (MVDR) beamforming algorithm in a MIMO radar system. Two MIMO radar systems are considered. The first one is a simulated MIMO radar which is used for automotive adaptive cruise control (ACC), and the second one is an experimental monostatic MIMO radar that is based on a vector network analyzer (VNA). It is shown that the GPU achieved a speedup of up to 7 times while successfully detecting all targets.

KEYWORDS
MIMO Radar, GPU, MVDR

Cite This Paper in IEEE or APA Citation Styles
(a). IEEE Format:
[1] Mostafa Hefnawi , Gillian Rideout , "MATLAB-Based GPU Acceleration for Multiple-Input and Multiple-Output Radar Beamforming Algorithm," Universal Journal of Electrical and Electronic Engineering, Vol. 7, No. 1, pp. 19 - 27, 2020. DOI: 10.13189/ujeee.2020.070102.

(b). APA Format:
Mostafa Hefnawi , Gillian Rideout (2020). MATLAB-Based GPU Acceleration for Multiple-Input and Multiple-Output Radar Beamforming Algorithm. Universal Journal of Electrical and Electronic Engineering, 7(1), 19 - 27. DOI: 10.13189/ujeee.2020.070102.