Abstract

Noise immunity and speed are two vital issues for designing encoding-decoding system for wireless communication. Convolutional coding is widely used in wireless communication system for its error correction property. For the decoding purpose of Convolutional coding Viterbi decoder is used. Core module of Viterbi decoder is Adder-Comparator-Selector (ACS) which takes approximately 70% of total power consumption. So, Adder-Comparator-Selector (ACS) module is transformed into Comparator-Selector-Adder CSA) module for power saving. Reduction of Hamming Distance Logic Circuitry for branch metric calculation also saving power but enhances the packing density of the circuit. In this paper the comparison between ACS and CSA is not only described in terms of power reduction and area but also speed and noise immunity are compared. Basically there are three types of Viterbi decoders: namely Register Exchange, Shift Update and Selective Update. These decoders do not follows the parallelism and pipelining concept but folding cascaded designing of Viterbi Decoder supports parallelism which enhance the speed of the system. This paper gives a new idea of logic reduction of Viterbi Decoder as well as comparison of different Viterbi decoders in different aspects.

Highlights

  • As an extension of the previous paper [1], in this paper the main discussion point is speed, noise immunity and types of Viterbi decoders

  • This paper proposes a CSA module implementation which optimizes Viterbi decoder circuits

  • Reduced circuit Hamming distance calculation and first comparison method for path metric calculation at CSA module improves the power consumption up to 10% compromising the total area of the circuit

Read more

Summary

Introduction

As an extension of the previous paper [1], in this paper the main discussion point is speed, noise immunity and types of Viterbi decoders. In the paper [1], comparison makes through only Adder-Comparator-Selector (ACS) and Comparator-Selector-Adder (CSA) modules of the Viterbi decoder, but whole system comparison is not described. The encoder output of Convolutional coding is depends on present input and depends on previous input which enhances the capability of error correction by assumption of present output using previous inputs. This characteristics differs Convolutional coding from Block coding.

Motivation
Working Principles of Viterbi Decoder
Working Principle of ACS Module in Viterbi Decoder
Generation of branch metric in conventional way
Generation of branch metrics in new architecture
Transformation of ACS unit to CSA unit in new architecture
Types of Viterbi Decoder
Comparison of Error rates in Different Viterbi Decoding Approaches
Test Results and Analysis
Noise Immunity Test Result
Speed Test Result
10. Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call