Punctured coding for forward error correction in satellite communications

It is attractive to be able to alter the coding rate, and hence the error correcting power, of a forward error correcting coder/decoder (codec) to provide a codec whose error correction capability can be matched to the requirements of the data communication system in which it is used. Punctured coding applied to convolutional encoding/Viterbi decoding offers a method of achieving this. Intelsat have specified punctured coding to provide flexibility in selecting either 1/2 or 3/4 rate coding in their Open Network Intermediate Data Rate system. This paper discusses the modification of a proprietary constraint length 7, 1/2-rate codec to operate with punctured coding at rates up to 7/8. Following a brief discussion of the requirements for variable rate coding in communication systems the practical implementation of a prototype codec and considerations relating to performance characterization are discussed. Experimental results at coding rates of 1/2, 2/3, 3/4, and 7/8 are given, together with simulation of performance at 15/16 coding rate to demonstrate the sensitivity of performance to path history length at higher coding rates. It is concluded that the method is very suitable to provide variable coding rates between 1/2 and 7/8 but suffers from increasingly severe practical disadvantages at coding rates above 7/8.