Systematic Luby Transform codes as incremental redundancy scheme

Abstract

Systematic Luby Transform (fountain) codes are investigated as a possible incremental redundancy scheme for EDGE. The convolutional incremental redundancy scheme currently used by EDGE is replaced by the fountain approach. The results of the simulations performed for each incremental redundancy scheme show that the fountain approach outperforms the convolutional approach on the second retransmission when implemented on the EDGE platform. The results also indicate that if the packet sizes used by a specific platform is large enough the fountain approach will always outperform the convolutional approach.