Performance Analysis of Polar Decoding Using Linear Code Mapping with Feedback Based ANN Architecture

Polar coding is being used in control channel for 5G standard. It helps to obtain channel polarisation that helps to split noisy channel into ideal noiseless channel and noisy channel. The rate of the channel code be chosen in a fine-grained manner. Successive cancellation and successive cancellation list decoding are the popular techniques that are used for polar decoding. In this project, we propose to use cascaded ANN inspired from the successive cancellation model for decoding the polar code. The performance in terms of detection rate has been compared with the corresponding values using successive cancellation technique alone. Experimental results reveal that there is a chance for increase in the detection rate under low SNR.