A Study of Neural Polar Decoders for Communication

TL;DR

This work extends Neural Polar Decoders (NPDs) from memoryless or synthetic channels to real-world memory channels encountered in 5G-like systems, enabling end-to-end decoding for both OFDM and single-carrier waveforms without pilots or cyclic prefixes. By introducing rate matching, higher-order modulation support, and robustness across diverse channel conditions with a single trained model, the NPD delivers consistent BER/BLER and throughput gains over the standard 5G polar decoder. The combination of a learned channel embedding, puncturing-aware rate matching, and MI-driven code design yields a practical, pilotless decoding solution that performs well under Doppler, delay spreads, and nonlinear PA effects, with additional advantages for SC in terms of PAPR. These results suggest substantial performance and spectral-efficiency improvements for 5G control channels and beyond, along with a viable path for hardware-aware, scalable neural decoding architectures.

Abstract

In this paper, we adapt and analyze Neural Polar Decoders (NPDs) for end-to-end communication systems. While prior work demonstrated the effectiveness of NPDs on synthetic channels, this study extends the NPD to real-world communication systems. The NPD was adapted to complete OFDM and single-carrier communication systems. To satisfy practical system requirements, the NPD is extended to support any code length via rate matching, higher-order modulations, and robustness across diverse channel conditions. The NPD operates directly on channels with memory, exploiting their structure to achieve higher data rates without requiring pilots and a cyclic prefix. Although NPD entails higher computational complexity than the standard 5G polar decoder, its neural network architecture enables an efficient representation of channel statistics, resulting in manageable complexity suitable for practical systems. Experimental results over 5G channels demonstrate that the NPD consistently outperforms the 5G polar decoder in terms of BER, BLER, and throughput. These improvements are particularly significant for low-rate and short-block configurations, which are prevalent in 5G control channels. Furthermore, NPDs applied to single-carrier systems offer performance comparable to OFDM with lower PAPR, enabling effective single-carrier transmission over 5G channels. These results position the NPD as a high-performance, pilotless, and robust decoding solution.

Figures (13)

• Figure 1: 5G Polar code system. Yellow, red, and orange blocks are implemented in downlink, uplink, and both, respectively.
• Figure 2: Block diagram of an end-to-end communication system integrating the proposed NPD. The diagram of the system presents both OFDM and single-carrier waveforms, which differ only in their Waveform block.
• Figure 3: A visualization of NSCLoss for $N=4$.
• Figure 4: Pilot patterns used in simulations.
• Figure 5: BER and BLER over a TDL-A channel with 100 ns delay spread and $0\text{--}8\,\text{m/s}$ mobility. The system uses OFDM with BPSK modulation and code length $N = 1024$. The NPD is compared to the 5G Polar code with LS estimation and Perfect CSI. Results are shown for varying information bit lengths.

Theorems & Definitions (5)

• Definition 1: Wireless channel model
• Definition 2: Puncturing Set Based on Bit-Reversal Permutation
• Definition 3: Channel Embedding
• Definition 4: NPD SC functions
• Definition 5: Channel Embedding for Communication