Efficient Decoder Design For Error Correcting Codes PDF Download

Error correction codes (ECCs) have been widely used in communication systems and storage devices. Nowadays, the rapid development of integrated circuit technologies makes feasible the implementation of powerful ECCs such as turbo code and low-density parity-check (LDPC) code. However, these high-performance codes require complex decoding algorithms, resulting in large hardware area and high power consumption. Furthermore, some of these decoders require an iterative decoding process, which leads to a long decoding latency. Therefore, low-complexity, low-power and high-speed very-large-scale integration (VLSI) architecture design for the ECC decoder is of great importance. This dissertation focuses on efficient VLSI implementation for the decoders of convolutional codes and two advanced coding schemes based on convolutional code: trellis-coded modulation (TCM) and convolutional turbo code (CTC). The first part of this dissertation is dedicated to low-complexity, low-power decoders design for a 4-dimensional, 8-ary phase-shift keying (4-D 8PSK) TCM system. We propose a low-complexity architecture for the transition-metric unit (TMU) to reduce the hardware area without performance loss. Then, a power-efficient scheme by applying T-algorithm on branch metrics (BMs) is proposed for the Viterbi decoder (VD) embedded in the 4-D 8PSK TCM decoder. Unlike the conventional T-algorithm, the proposed scheme does not affect the clock speed of the decoder. Finally, a hybrid T-algorithm is developed by applying T-algorithm on both BMs and path metrics (PMs), which reduces significantly more computations than the conventional T-algorithm applied on PMs. The VLSI design for VDs has been an active research area for decades. In the second part of the dissertation, we extend our research to a more general topic of VDs, where novel architectures are explored to efficiently reduce the power consumption, while still maintaining a high decoding speed and a low decoding latency. CTCs are constructed from parallel convolutional encoding of the same message in different sequences and have the error-correcting capability near the Shannon bound. Practical decoding schemes normally require an iterative decoding process employing the soft-in soft-out (SISO) decoder. The third part of this dissertation is focused on the SISO decoder design for double-binary (DB) CTCs. We propose a low-complexity, memory-reduced architecture by partitioning BMs into two independent portions: information metrics and parity metrics. Furthermore, high-speed recursion architectures for logarithm domain maximum a posteriori probability (log-MAP) algorithm are proposed to increase the decoding speed by algorithmic approximation and bit-level optimization.