Adaptive Quantized and Normalized QC-LDPC Universal Block-Parallel Decoder

As wireless communication systems evolve, new frameworks place increasing demands on terminal performance, including throughput, latency, power consumption, and size. Software-Defined Radio (SDR) systems have the potential to support multiple communication standards and enabling on-demand upgrades due to their flexibility. However, purely software-based implementations for physical layer protocols reduce efficiency, necessitating the use of external accelerators such as FPGA for specialized computation. Specifically, channel decoding contributes a lot to the physical layer’s complexity, and thus is highly demand for FPGA specialized implementation while retaining the universality of the decoding architecture in some level. This paper presents an adaptive quantized and normalized QC-LDPC universal decoder based on a block-parallel architecture with pipeline and instruction-driven approach. The proposed decoder improves throughput and reduces computational complexity while maintaining flexibility for various QC-LDPC codes. Experimental results demonstrate that the adaptive quantization and normalization scheme significantly outperforms existing methods in terms of decoding performance and maximum frequency, offering a more efficient solution for SDR systems.