Implementation of a Parallel Viterbi Decoder for High Speed Multimedia Communications

멀티미디어 통신용 병렬 아키텍쳐 고속 비터비 복호기 설계

  • Published : 2000.02.01

Abstract

The Viterbi decoders can be classified into serial Viterbi decoders and parallel Viterbi decoders. Parallel Viterbi decoders can handle higher data rates than serial Viterbl decoders. This paper designs and implements a fully parallel Viterbi decoder for high speed multimedia communications. For high speed operations, the ACS (Add-Compare-Select) module consisting of 64 PEs (Processing Elements) can compute one stage in a clock. In addition, the systolic away structure with 32 pipeline stages is developed for the TB (traceback) module. The implemented Viterbi decoder can support code rates 1/2, 2/3, 3/4, 5/6 and 7/8 using punctured codes. We have developed Verilog HDL models and performed logic synthesis. The 0.6 ${\mu}{\textrm}{m}$ SAMSUNG KG75000 SOG cell library has been used. The implemented Viterbi decoder has about 100,400 gates, and is running at 70 MHz in the worst case simulation.

비터비 복호기는 직렬 복호 방식과 병렬 복호 방식 2 가지로 분류할 수 있다. 병렬 비터비 복호기는 직렬비터비 복호기에 비해 보다 높은 데이타율을 얻을 수 있다. 본 논문에서는 고속 멀티미디어 통신을 위한 병렬 비티비 복호기 구조를 설계하고 구현한다. 설계한 비터비 복호기는 고속 동작을 위해 64개의PE(Processing Element)를 사용해 한 클럭에 처리가 가능하도록 하였다. 또한 파이프라인 스테이지를 갖는 시스톨릭 어레이 구조의 TB(Traceback) 블럭을 설계하였다. 본 논문에서 설계한 비터비 복호기는 puncturing을 통해 부호율 1/2, 2/3, 3/4, 5/6, 7/8을 지원한다. Verilog 모델을 구현하였고 0.6㎛ Samsung KG75000 SOG 셀 라이브러리를 이용하여 논리합성을 수행하였다. 구현된 비터비 복호기는 약100,400 게이트이며 동작 속도는 worst case에서 70㎒로 기존 상용 칩들보다 빠르다.

Keywords

References

  1. Stephen B. Wicker, Error Control Systems for Digital Communication and Storage, Prentice-Hall, 1995
  2. Gennady Feygin and P. G. Gulak, 'Architectural Tradeoffs for Survivor Sequence Memory Management in Viterbi Decoders,' IEEE Trans. on Communications, vol. 41, no. 3, March 1993, pp. 425-429 https://doi.org/10.1109/26.221067
  3. Montse Boo, Francisco Arguello, Javier D. Bruguera, Ramon Doallo, and Emilio L. Zapata, 'High-Performance VLSI Architecture for the Viterbi Algorithm,' IEEE Trans. on Communications, vol. 45, no. 2, Feb. 1997, pp. 168-176 https://doi.org/10.1109/26.554365
  4. Emmanuel Boutilion, Nicolas Demassieux, 'High Speed Low Power Architecture for Memory Management in a Viterbi Decoder,' in Proc. International Symposium on Circuits and Systems (ISCAS'96 ), Atlanta, USA, May 1996, pp. 284-287 https://doi.org/10.1109/ISCAS.1996.541957
  5. C. Bernard Shung, Horng-Dar Lin, Robert Cypher, Paul H. Siegel, and Hemant K. Thapar, 'Area-Efficient Architectures for the Viterbi Algorithm-Part I: Theory,' IEEE Trans. on Communications, pp. 636-6433, vol. 41, April, 1993 https://doi.org/10.1109/26.223789
  6. Shuji Kubota, Shuzo Kato, and Tsunehachi Ishitani, 'Novel Viterbi Decoder VLSI Implementation and its Performance,' IEEE Trans. on Communications, pp. 1170-1178, vol. 41, No.8, Aug., 1993 https://doi.org/10.1109/26.231960
  7. Horng-Dar Lin, David G. Messerschmitt, 'Parallel Viterbi Decoding Methods for Uncontrollable and Controllable Sources,' IEEE Trans. on Communications, vol. 41, no. 1, Jan. 1993, pp. 62-69 https://doi.org/10.1109/26.212366
  8. David Haccoun, Guy Begin, 'High-Rate Punctured Convolutional Codes for Viterbi and Sequential Decoding,' IEEE Trans. on Communications, vol. 37, no. 11, Nov. 1989, pp. 1113-1124 https://doi.org/10.1109/26.46505
  9. Qualcomm Inc., Forward Error Correction Products Data Book, Q1900
  10. Motorola Inc., Motorola Semiconductor Technical Data, MC92300
  11. LSI Logic Inc., LSI Logic Technical Data, L64709