An Adaptive CSI Feedback Model Based on BiLSTM for Massive MIMO-OFDM Systems

TL;DR

This work tackles the challenge of adaptive CSI feedback in massive MIMO-OFDM under varying input lengths and feedback budgets. It introduces ABLNet, a BiLSTM-based autoencoder, with a plug-in FBCU that allows flexible output bitrates, and a BNA algorithm to stabilize SGCS at a target level while minimizing overhead. A UE-first separate training strategy is proposed to decouple encoder and decoder across manufacturers, reducing complexity and enabling a common gNB decoder. Experiments show robust performance across subband counts, effective bitrate adaptation, and comparable results to joint-training baselines, highlighting practical applicability in heterogeneous deployments.

Abstract

Deep learning (DL)-based channel state information (CSI) feedback has the potential to improve the recovery accuracy and reduce the feedback overhead in massive multiple-input multiple-output orthogonal frequency division multiplexing (MIMO-OFDM) systems. However, the length of input CSI and the number of feedback bits should be adjustable in different scenarios, which can not be efficiently achieved by the existing CSI feedback models. Therefore, an adaptive bidirectional long short-term memory network (ABLNet) for CSI feedback is first designed to process various input CSI lengths, where the number of feedback bits is in proportion to the CSI length. Then, to realize a more flexible feedback bit number, a feedback bit control unit (FBCU) module is proposed to control the output length of feedback bits. Based on which, a target feedback performance can be adaptively achieved by a designed bit number adjusting (BNA) algorithm. Furthermore, a novel separate training approach is devised to solve the model protection problem that the UE and gNB are from different manufacturers. Experiments demonstrate that the proposed ABLNet with FBCU can fit for different input CSI lengths and feedback bit numbers; the CSI feedback performance can be stabilized by the BNA algorithm; and the proposed separate training approach can maintain the feedback performance and reduce the complexity of feedback model.

Figures (14)

• Figure 1: The massive MIMO-OFDM system model and adaptive CSI feedback framework.
• Figure 2: Proposed ABLNet architecture.
• Figure 3: BiLSTM structure.
• Figure 4: CSI feedback with different lengths of subband eigenvectors.
• Figure 5: CSI feedback scheme with FBCU.