Deep OFDM Channel Estimation: Capturing Frequency Recurrence

TL;DR

This work addresses OFDM channel estimation by exploiting frequency-domain correlations within a single slot. It introduces SisRafNet, a CNN-GRU hybrid that applies bidirectional GRUs along the frequency axis to capture cross-subcarrier correlations, fed by pilot-based LS estimates. It outperforms DL baselines ChannelNet, ChanEstNet, SRDnNet and the LMS-based approach across 3GPP CDL-A and CDL-D channels over a broad SNR range, all while requiring only a single slot. It shows robustness to pilot configurations and generalizes across center frequencies, with implications for CSI compression and beamforming.

Abstract

In this paper, we propose a deep-learning-based channel estimation scheme in an orthogonal frequency division multiplexing (OFDM) system. Our proposed method, named Single Slot Recurrence Along Frequency Network (SisRafNet), is based on a novel study of recurrent models for exploiting sequential behavior of channels across frequencies. Utilizing the fact that wireless channels have a high degree of correlation across frequencies, we employ recurrent neural network techniques within a single OFDM slot, thus overcoming the latency and memory constraints typically associated with recurrence based methods. The proposed SisRafNet delivers superior estimation performance compared to existing deep-learning-based channel estimation techniques and the performance has been validated on a wide range of 3rd Generation Partnership Project (3GPP) compliant channel scenarios at multiple signal-to-noise ratios.

Figures (7)

• Figure 1: Direction of recurrent layer operations for proposed SisRafNet with recurrence across frequency.
• Figure 2: Architecture of the Proposed SisRafNet with Bidirectional GRU layers.
• Figure 3: NMSE performance comparison of SisRafNet.
• Figure 4: Performance when training with samples from different SNR combinations.
• Figure 5: Generalization across different channel models.