Cryptcoding of Images for Transmission Trough a Burst Channels

Abstract

Random Codes Based on Quasigroups (RCBQ) are error-correcting codes that crypt the messages at the same time.These cryptcodes are proposed in 2007 and after that several improvements of coding/decoding algorithms have beencmade. For better performances for transmission through a binary-symmetric and Gaussian channel, Cut-Decoding and 4-Sets-Cut-Decoding algorithms were defined. Also, a modification of these algorithms (Burst-Cut-Decoding and Burst-4-Sets-Cut-Decoding algorithms) for correction of burst errors was proposed elsewhere. In this paper, we investigate performances of these algorithms for transmission of images through a burst channel. For simulation of burst errors, we use Gilbert-Elliot model. We consider two kinds of Gilbert-Eliott channel, in the first one in each state the channel isbinary symmetric and in the second one, the channel is Gaussian. In all experiments, for different values of bit-error probability (in BSC) and SNR (in Gaussian), the differences between transmitted and decoded images are considered. From the experiments can be concluded that Burst-4-Sets-Cut-Decoding algorithms gives better results than Burst-CutDecoding algorithms (i.e., clearer images) and it is much faster. Also, a filter is applied on the images (after decoding with RCBQs) for enhancing the quality of them. With the considered filter clearer images are obtained. In this paper we consider only error-correction capabilities of RCBQs, but the images decoded with these codes are also encrypted.