Publishing house Radiotekhnika

"Publishing house Radiotekhnika":
scientific and technical literature.
Books and journals of publishing houses: IPRZHR, RS-PRESS, SCIENCE-PRESS

Тел.: +7 (495) 625-9241


Software implementation of multithreshold decoders on GPU


V.V. Zolotarev – Dr. Sc. (Eng.), Professor, Senior Research Scientist, Space Research Institute of RAS (Moscow) E-mail: G.V. Ovechkin – Dr. Sc. (Eng.), Associate Professor, Professor, Ryazan State Radio Engineering University E-mail: P.V. Ovechkin – Ph. D. (Eng.), Associate Professor, Ryazan State Radio Engineering University E-mail:

Multithreshold decoder (MTD) is discussed. The decoder is simplest decoder of majority type and is used for decoding of self-orthogonal codes. MTD has low implementation complexity and is simple for hardware implementation. These allows to use MTD in high-speed communication and data storage systems with rates 1 Gbit/s and more. This paper presents background of MTD, comparison results of MTD bit-error rate performance and the performance of other error correction methods, known hardware MTD implementations. It\'s noted the last time telecommunications specialists pay more attention to software-defined radio (SDR). In SDR a significant part of the digital signal processing is performed on a standard personal computer. So software decoders need to be used to ensure required information decoding rate and to provide efficient error correction. The work submit high speed software binary MTD for a block self-orthogonal code. This version of MTD uses computing abilities of modern graphics processing units (GPU). During implementation of software MTD on GPU all arrays used by encoder and decoder were accommodated in the shared and constant memory. Conflict-free access to the elements of such arrays and the simultaneous decoding of several symbols within the message received from a channel used by GPU threads were provided. These allows MTD to provide parallel and independent decoding of several hundred blocks received from the channel. Simulation results are presented show the op-eration rate of the software MTD implemented on GPU GTX 970 was up to 350 Mbit/s. So, software MTD submitted in this paper can be used in high-speed software-defined radio systems based on graphics processing units.


  1. Zolotarev V.V., Zubarev Y.B., Ovechkin G.V. Optimization Coding Theory and Multithreshold Algorithms. Published in Switzerland by ITU. March 2016. 158 p. URL =
  2. Zolotarev V.V., Ovechkin G.V. Increase of reliability of data transmission and data storage using multithreshold methods of decoding of error-correcting codes // Digital signal processing. 2012. № 1. P. 16−21.
  3. Ullah M.A., Okada K., Ogivara H. Multi-Stage Threshold Decoding for Self-Orthogonal Convolutional Codes // IEICE Trans. Fundamentals. November 2010. V. E93-A. № 11. P. 1932−1941.
  4. Zolotarev V.V., Ovechkin G.V. Efficient Multithreshold Decoding of Nonbinary Codes // Journal of Communications Technology and Electronics. 2010. V. 55. № 3. P. 302−306 (ISSN 1064–2269).
  5. Ovechkin G.V., Ovechkin P.V. Ispolzovanie nedvoichnogo mnogoporogovogo dekodera v kaskadnykh skhemakh korrekcii oshibok // Vestnik Rjazanskogo gosudarstvennogo radiotekhnicheskogo universiteta. 2009. № 29. S. 7−12.
  6. Zolotarev V.V., Ovechkin G.V., Fediov V.S. Povyshenie skorosti raboty nedvoichnogo mnogoporogovogo dekodera // Vestnik Rjazanskogo gosudarstvennogo radiotekhnicheskogo universiteta. 2013. № 4−2 (46). S. 22−27.
  7. Zolotarev V., Ovechkin G., Satybaldina D., Tashatov N., Adamova A., Mishin V. Efficiency multithreshold decoders for self-orthogonal block codes for optical channels // International Journal of Circuits, Systems and Signal Processing. 2014. V. 8. P. 487−495 (ISSN 1998−4464).
  8. Ullah M.A., Omura R., Sato T., Ogivara H. Multi-Stage Threshold Decoding for High Rate Convolutional Codes for Optical Communications // Seventh Advanced Int. Conf. on Telecommunications (AICT 2011). 2011. P. 87−93.
  9. Software defined radio: architectures, systems, and functions. Dillinger, Madani, Alonistioti. Wiley. 2003. 454 p.
  10. Guohui Wang, Michael Wu, Bei Yin, Joseph R. Cavallaro. High throughput low latency LDPC decoding on GPU for SDR systems // Global Conference on Signal and Information Processing (GlobalSIP). 2013. P. 1258−1261.
  11. Pat. RF № 2377722. Sposob dekodirovanija pomekhoustojjchivogo koda / Zolotarev V.V.


© Издательство «РАДИОТЕХНИКА», 2004-2017            Тел.: (495) 625-9241                   Designed by [SWAP]Studio