Blind Identification of Channel Code Parameters on Burst Error Channels

Abstract

For the blind identification of channel code parameters, existing literature only focus on the channels where errors are uniformly distributed. In this paper, we investigate the code parameter identification performance on the specific Gilbert-Elliott burst error channels. We analyze the relationship among the amount of intercepted data/codewords, the channel model parameters, the code parameters, and the identification performance, and find that as the code rate increases, or the amount of intercepted available data/codewords decreases, or errors tend to be uniformly distributed, it becomes less likely to identify the code parameters. The validity of our conclusions is demonstrated by the simulations and analysis.