Accelerating and Compressing Deep Neural Networks for Massive MIMO CSI Feedback
This work addresses the barrier of deploying deep learning in practical wireless systems requiring low-latency and low memory, though it is incremental as it applies existing compression methods to a specific domain.
The paper tackles the inefficiency of large deep neural networks for massive MIMO CSI feedback by proposing model compression techniques, achieving an 86.5% reduction in model size and a 76.2% reduction in inference time with minimal accuracy loss.
The recent advances in machine learning and deep neural networks have made them attractive candidates for wireless communications functions such as channel estimation, decoding, and downlink channel state information (CSI) compression. However, most of these neural networks are large and inefficient making it a barrier for deployment in practical wireless systems that require low-latency and low memory footprints for individual network functions. To mitigate these limitations, we propose accelerated and compressed efficient neural networks for massive MIMO CSI feedback. Specifically, we have thoroughly investigated the adoption of network pruning, post-training dynamic range quantization, and weight clustering to optimize CSI feedback compression for massive MIMO systems. Furthermore, we have deployed the proposed model compression techniques on commodity hardware and demonstrated that in order to achieve inference gains, specialized libraries that accelerate computations for sparse neural networks are required. Our findings indicate that there is remarkable value in applying these model compression techniques and the proposed joint pruning and quantization approach reduced model size by 86.5% and inference time by 76.2% with minimal impact to model accuracy. These compression methods are crucial to pave the way for practical adoption and deployments of deep learning-based techniques in commercial wireless systems.