Overview of LDPC codes

Abstract LDPC decoding involves careful design of the sparse parity-check matrix and selection of the appropriate decoding algorithm. An LDPC matrix can be constructed using random, structured or unstructured techniques. LDPC codes generated from each of these techniques have significant impact on performance and implementation complexity. An LDPC decoding algorithm also plays a vital role in determining error correction and decoding performance. The overall performance of an LDPC decoder is measured by various parameters in terms of bit errors, throughput, and hardware resources. These parameters are measured and compared with the application requirements and benchmarks at various stages of the LDPC code design phase. It is imperative to choose appropriate techniques and design parameters to relax the hardware implementation complexity to an optimal extent. This chapter provides an overview of different types of LDPC codes and a list of related terminology that is normally used as design parameters and performance benchmarking metrics for LDPC decoders.