Minglun Gao

Multi-Gb/s LDPC Code Design and Implementation

Low-density parity-check (LDPC) code, a very promising near-optimal error correction code (ECC), is being widely considered in next generation industry standards. The VLSI implementation of high-speed LDPC decoder remains a big challenge. This paper presents the construction of a new class of implementation-oriented LDPC codes, namely shift-LDPC codes. With girth optimization, this kind of codes can perform as well as computer generated random codes. More importantly, the decoder can be efficiently implemented to obtain very high decoding speeds. In addition, more than 50% of message memory can be generally saved over conventional partially parallel decoder architectures. We demonstrate the benefits of the proposed techniques with an application-specific integrated circuit (ASIC) design (in 0.18-mum CMOS) for a 8192-bit regular LDPC code, which can achieve 5 Gb/s throughput at 15 iterations.

Area-efficient reed-solomon decoder design for optical communications

A high-speed low-complexity Reed-Solomon (RS) decoder architecture based on the recursive degree computationless modified Euclidean (rDCME) algorithm is presented in this brief. The proposed architecture has very low hardware complexity compared with the conventional modified Euclidean and degree computationless modified Euclidean (DCME) architectures, since it can reduce the degree computation circuitry and replace the conventional systolic architecture that uses many processing elements (PEs) with a recursive architecture using a single PE. A high-throughput data rate is also facilitated by employing a pipelining technique. The proposed rDCME architecture has been designed and implemented using SMIC 0.18-mum CMOS technology. Synthesized results show that the proposed RS (255, 239) decoder requires only about 18 K gates and can operate at 640 MHz to achieve a throughput of 5.1 Gb/s, which meets the requirement of modern high-speed optical communications.

An improved scaled DCT architecture

This paper presents an efficient architecture for computing the eight-point 1D scaled DCT (discrete cosine transform) with a new algorithm based on a selected Loeffler DCT scheme whose multiplications are placed in the last stage. The proposed DCT architecture does not require any scaling compensation in the computation. Furthermore, a multiplication approximation method is developed, which is more efficient than traditional CORDIC (coordinate rotation digital computer)-based algorithms. Compared to the latest work (Sun et al., 2007), the proposed approach can save 14% addition operations for the same precision requirement and the path delay can be significantly reduced as well.

Efficient Shuffle Network Architecture and Application for WiMAX LDPC Decoders

In this brief, a new algorithm that can efficiently generate all the control signals for the shuffle network used in flexible low-density parity-check (LDPC) decoders is proposed. Employing the proposed algorithm, the hardware complexity of the controller of shuffle networks using the Benes network structure can be significantly reduced. In addition, a low-complexity reconfigurable shuffle network architecture for flexible LDPC decoders is developed. Both the Benes network and the controller can be tailored to fit specific applications. Consequently, an efficient shuffle network for WiMAX LDPC decoders is presented. Synthesis results demonstrate that with the SMIC 0.18-mum complementary metal-oxide-semiconductor process, the total gate count of the proposed shuffle network is only 16 000. The area saving is between 26.6% and 71.1% compared to related works in the literature.

Decoder Design for RS-Based LDPC Codes

This brief studies very large-scale integration (VLSI) decoder architectures for RS-based low-density parity-check (LDPC) codes, which are a special class of LDPC codes based on Reed-Solomon codes. The considered code ensemble is well known for its excellent error-correcting performance and has been selected as the forward error correction coding scheme for 10GBase-T systems. By exploiting the shift-structured properties hidden in the algebraically generated parity-check matrices, novel decoder architectures are developed with significant advantages of high level of parallel decoding, efficient usage of memory, and low complexity of interconnection. To demonstrate the effectiveness of the proposed techniques, we completed a high-speed decoder design for a (2048, 1723) regular RS-LDPC code, which achieves 10-Gb/s throughput with only 820 000 gates. Furthermore, to support all possible RS-LDPC codes, two special cases in code construction are considered, and the corresponding extensions of the decoder architecture are investigated.

An FPGA Implementation of Array LDPC Decoder

Low-density parity-check (LDPC) code is one kind of prominent error correcting codes (ECC) being considered in next generation industry standards. This paper presents an FPGA implementation of array code based low-density parity-check code decoder. The advantages of the proposed architecture as compared to the fully parallel or partially parallel architecture are: less memory requirement, avoidance of complex global interconnects and its satisfying throughput. These advantages are derived from exploiting the well-defined structure of the parity check matrix of array code based LDPC codes

Efficient Decoder Implementation for QC-LDPC Codes

Channel coding is an important building block in communication systems. Low-density parity-check codes is one kind of prominent error correcting codes being considered in next generation industry standards. This paper presents a memory efficient, very high speed decoder architecture suited for quasi-cyclic low-density parity-check codes using modified Min-Sum decoding algorithm. In general, about seventy percent of message memory can be saved over conventional decoder architectures, and the decoding speed can be largely accelerated because of the highly efficient VLSI architecture. Consequently, the proposed approach facilitates the applications of LDPC codes in area/latency sensitive communication systems.

LDPC decoder design for high rate wireless personal area networks

This paper presents two efficient decoder designs for the low-density parity-check codes in IEEE 802.15.3 standard proposal. These decoders feature by efficient hardware usage, low message memory requirement and code rate flexibility. The highly parallel level design can reach a throughput of 3.6 Gbps, which fulfills the standard requirement by processing 72 columns and 72 rows in parallel. The low cost design offers another tradeoff which significantly reduces the area and power consumption while maintaining necessary data throughput required by specific applications. Furthermore, both decoders support three different code rates by employing flexible check node processing units.