Sampling Rate and Bits Per Sample Tradeoff for Cloud MIMO Radar Target Detection

Zhen Wang,Qian He,Rick S. Blum
DOI: https://doi.org/10.23919/EUSIPCO.2019.8903103
2019-01-01
Abstract:In this paper, target detection is studied for a cloud multiple-input multiple-output (MIMO) radar system, where each receiver communicates with a fusion center (FC) through a backhaul network. To reduce communication burden, local measurements at each receiver are quantized before they are sent to the FC. Under a bitrate constraint for each local sensor, we derive the detection probability of the cloud radar and analyze effects of the sampling rate and bits per sample on the detection performance. The quantizer output is initially modeled using direct analysis (DA), and then the Gaussian quantization error approximation (GQEA) method is employed to facilitate theoretical analysis. We verify that these two methods lead to close enough detection performance for large enough number of bits per sample. The tradeoff between the sampling rate and bits per sample is presented analytically and numerically.
What problem does this paper attempt to address?