Publication Type : Journal Article
Publisher : Elsevier B.V.
Source : Integration, Elsevier B.V., Volume 66, p.80-87 (2019)
Keywords : Application specific integrated circuits, Complex processing, Decoder architecture, Decoding, Decoding algorithm, Hardware complexity, Mergers and acquisitions, Normalized throughputs, Polar codes, Processing elements, Successive cancellation, Systems analysis
Campus : Coimbatore
School : School of Engineering
Department : Electronics and Communication
Verified : No
Year : 2019
Abstract : A low-complexity design architecture for implementing the Successive Cancellation (SC)decoding algorithm for polar codes is presented. Hardware design of polar decoders is accomplished using SC decoding due to the reduced intricacy of the algorithm. Merged processing element (MPE)block is the primary area occupying factor of the SC decoder as it incorporates numerous sign and magnitude conversions. Two's complement method is typically used in the MPE block of SC decoder. In this paper, a low-complex MPE architecture with minimal two's complement conversion is proposed. A reformulation is also applied to the merged processing elements at the final stage of SC decoder to generate two output bits at a time. The proposed merged processing element thereby reduces the hardware complexity of the SC decoder and also reduces latency by an average of 64%. An SC decoder with code length 1024 and code rate 1/2 was designed and synthesized using 45-nm CMOS technology. The implementation results of the proposed decoder display significant improvement in the Technology Scaled Normalized Throughput (TSNT)value and an average 48% reduction in hardware complexity compared to the prevalent SC decoder architectures. Compared to the conventional SC decoder, the presented method displayed a 23% reduction in area. © 2019 Elsevier B.V.
Cite this Research Publication : Geethu S., Madala, L. R., Gopalakrishnan, L., and Sellathurai, M., “Low-complex processing element architecture for successive cancellation decoder”, Integration, vol. 66, pp. 80-87, 2019.