Quantized Neural Networks for Radar Interference Mitigation
This work addresses the problem of deploying high-performance CNN-based radar interference mitigation algorithms on resource-constrained automotive radar sensor hardware, which is critical for the automotive industry.
This paper investigates quantization techniques for Convolutional Neural Networks (CNNs) used in radar signal denoising and interference mitigation. The authors analyze the quantization potential of various CNN architectures and sizes, focusing on reducing memory requirements for model storage and inference by quantizing weights and using piecewise constant activation functions.
Radar sensors are crucial for environment perception of driver assistance systems as well as autonomous vehicles. Key performance factors are weather resistance and the possibility to directly measure velocity. With a rising number of radar sensors and the so far unregulated automotive radar frequency band, mutual interference is inevitable and must be dealt with. Algorithms and models operating on radar data in early processing stages are required to run directly on specialized hardware, i.e. the radar sensor. This specialized hardware typically has strict resource-constraints, i.e. a low memory capacity and low computational power. Convolutional Neural Network (CNN)-based approaches for denoising and interference mitigation yield promising results for radar processing in terms of performance. However, these models typically contain millions of parameters, stored in hundreds of megabytes of memory, and require additional memory during execution. In this paper we investigate quantization techniques for CNN-based denoising and interference mitigation of radar signals. We analyze the quantization potential of different CNN-based model architectures and sizes by considering (i) quantized weights and (ii) piecewise constant activation functions, which results in reduced memory requirements for model storage and during the inference step respectively.