Convolutional Neural Network based Multiple-Rate Compressive Sensing for Massive MIMO CSI Feedback: Design, Simulation, and Analysis

TL;DR

A multiple-rate compressive sensing neural network framework to compress and quantize the CSI, which not only improves reconstruction accuracy but also decreases storage space at the UE, thus enhancing the system feasibility.

Abstract

Massive multiple-input multiple-output (MIMO) is a promising technology to increase link capacity and energy efficiency. However, these benefits are based on available channel state information (CSI) at the base station (BS). Therefore, user equipment (UE) needs to keep on feeding CSI back to the BS, thereby consuming precious bandwidth resource. Large-scale antennas at the BS for massive MIMO seriously increase this overhead. In this paper, we propose a multiple-rate compressive sensing neural network framework to compress and quantize the CSI. This framework not only improves reconstruction accuracy but also decreases storage space at the UE, thus enhancing the system feasibility. Specifically, we establish two network design principles for CSI feedback, propose a new network architecture, CsiNet+, according to these principles, and develop a novel quantization framework and training strategy. Next, we further introduce two different variable-rate approaches, namely, SM-CsiNet+ and PM-CsiNet+, which decrease the parameter number at the UE by 38.0% and 46.7%, respectively. Experimental results show that CsiNet+ outperforms the state-of-the-art network by a margin but only slightly increases the parameter number. We also investigate the compression and reconstruction mechanism behind deep learning-based CSI feedback methods via parameter visualization, which provides a guideline for subsequent research.

Figures (11)

• Figure 1: Overview of the multiple-rate bit-level compressive sensing CSI feedback framework. UE compresses CSI matrix with a selected $CR$, quantizes the measurement vector, and then transmits it. Once BS receives the transmitted bitstream, it dequantizes bitstream and then decompresses measurement vector.
• Figure 2: Overview of CsiNet+ architecture. The left module is an encoder at the UE, compressing the CSI matrix. Meanwhile, the right module is a decoder at the BS, reconstructing CSI matrix from the received compressive measurements.
• Figure 3: Proposed bit-level CsiNet+ framework. The original CSI is first compressed at the encoder (UE), and then quantization is adopted to generate a bitstream. At the decoder (BS), the received measurement vectors are first dequantized and then fed into several neural networks.
• Figure 4: Series multiple-rate compression framework. The key idea of SM-CsiNet+ is that high compression measurement vectors can be generated from the low ones.
• Figure 5: Parallel multiple-rate compression framework.