UCINet0: A Machine Learning based Receiver for 5G NR PUCCH Format 0

TL;DR

This work reframes PUCCH Format 0 UCI decoding as a multi-label classification task and presents UCINet0, a compact fully connected NN that predicts phase-rotation values $\alpha_m$ for up to 12 multiplexed UEs or detects zero transmissions. By fusing 24 real-valued RX samples with L2-provided metadata, the model learns to map resource-grid observations to UCI content without explicit channel estimation. Across simulated, lab, and live OTA datasets, UCINet0 outperforms the conventional DFT-based correlation baseline in accuracy and robustness, particularly under multiplexing and hardware-induced impairments, with inference latency suitable for FPGA implementation. The results demonstrate the practical viability of ML-based receivers for PUCCH Format 0, offering improved reliability for critical uplink feedback in 5G and beyond.

Abstract

Accurate decoding of Uplink Control Information (UCI) on the Physical Uplink Control Channel (PUCCH) is essential for enabling 5G wireless links. This paper explores an AI/ML-based receiver design for PUCCH Format 0. Format 0 signaling encodes the UCI content within the phase of a known base waveform and even supports multiplexing of up to 12 users within the same time-frequency resources. The proposed neural network classifier, which we term UCINet0, is capable of predicting when no user is transmitting on the PUCCH, as well as decoding the UCI content for any number of multiplexed users (up to 12). The test results with simulated, hardware-captured (lab) and field datasets show that the UCINet0 model outperforms conventional correlation-based decoders across all SNR ranges and multiple fading scenarios.

Figures (23)

• Figure 1: Example scenarios showing how assigning different initial cyclic shifts to different UEs allows them to be multiplexed on the same time-frequency resources.
• Figure 2: Illustration of the correlation of the received Format 0 signal with the known base sequence for (a) 1 UE transmission and (b) 3 UEs multiplexed transmission
• Figure 3: UCINet0 Architecture for PUCCH Format 0 decoding with 24 neurons in the input layer, 256+1 neurons in the second layer, 256 neurons in the third layer, and 12 neurons in the output layer.
• Figure 4: Model accuracy for different FCN architectures at various values of $\Delta$. Here, the number of FCN layers and the number of neurons of each architecture are presented on the x-axis, and $\Delta$ represents the maximum metadata offset between the true value and L2’s upper bound for the number of multiplexed UEs. We use these results to arrive at the final UCINet0 architecture presented in this paper.
• Figure 5: Model test accuracy vs SNR for various values of $\Delta$. (a) No metadata given during training, (b) Metadata given during training at the input layer, (c) Metadata given during training at the first hidden layer, and (d) Metadata given during training at the second hidden layer.