AI/ML based Joint Source and Channel Coding for HARQ-ACK Payload

TL;DR

This work addresses the bias inherent in HARQ-ACK bits by proposing a joint source-channel coding framework for uplink HARQ-ACK using a transformer-based encoder, per-codeword power shaping, and a Neyman-Pearson-inspired bitwise decoder to unequal-protect NACK over ACK. The proposed approach, combined with NR-compliant fading-channel receiver designs (including a low-complexity coherent variant), yields 3–6 dB reductions in average transmit power and 2–3 dB reductions in maximum transmit power relative to NR baselines, offering substantial coverage gains. The results demonstrate robustness to distribution shifts via Free-Lunch training and show practical gains under realistic NR uplink conditions, including DMRS shaping and channel estimation. Overall, the paper provides a principled, end-to-end JSCC framework for HARQ-ACK that leverages source statistics to enhance uplink power efficiency and reliability in 5G NR and beyond.

Abstract

Channel coding from 2G to 5G has assumed the inputs bits at the physical layer to be uniformly distributed. However, hybrid automatic repeat request acknowledgement (HARQ-ACK) bits transmitted in the uplink are inherently non-uniformly distributed. For such sources, significant performance gains could be obtained by employing joint source channel coding, aided by deep learning-based techniques. In this paper, we learn a transformer-based encoder using a novel "free-lunch" training algorithm and propose per-codeword power shaping to exploit the source prior at the encoder whilst being robust to small changes in the HARQ-ACK distribution. Furthermore, any HARQ-ACK decoder has to achieve a low negative acknowledgement (NACK) error rate to avoid radio link failures resulting from multiple NACK errors. We develop an extension of the Neyman-Pearson test to a coded bit system with multiple information bits to achieve Unequal Error Protection of NACK over ACK bits at the decoder. Finally, we apply the proposed encoder and decoder designs to a 5G New Radio (NR) compliant uplink setup under a fading channel, describing the optimal receiver design and a low complexity coherent approximation to it. Our results demonstrate 3-6 dB reduction in the average transmit power required to achieve the target error rates compared to the NR baseline, while also achieving a 2-3 dB reduction in the maximum transmit power, thus providing for significant coverage gains and power savings.

Figures (15)

• Figure 1: E2E FD pipeline for uplink SIMO PUCCH HARQ-ACK transmission per Rx antenna.
• Figure 2: Transformer-based Architecture for encoding PUCCH HARQ-ACK
• Figure 3: Entropy and Learnt PS as a function of $\pi_m$
• Figure 4: Transition probabilites for consecutive HARQ-ACK bits
• Figure 5: ACK and NACK BER vs. SNR for $p=0.9, \rho = 0$.

Theorems & Definitions (5)

• Remark 1
• Remark 2
• Remark 3
• Remark 4
• Remark 5