Reliability Analysis of the Joint LDPC Decoding Algorithms over the Multiple Access Channels

Abstract

The joint Low Density Parity-Check (LDPC) decoding schemes iteratively decode the received data from multiple channels. Mostly, the available data in different channels are correlated and there is kind of dependency between the links or channels. In recent decades, the graph-based codes have been considered for the communication network scenarios. The performance of these codes is close to the existing theoretical bounds and their complexity is not high which cause the possibility of real world implementation and exploitation. The Multiple Access Channel (MAC) scenario with multiple senders which aim to send correlated data to a single receiver is considered. An analysis on the reliability of the Bit Error Rate (BER) performance of the Joint Sum-Product (JSP) decoding algorithm is presented for a two-link case, which can be extended to higher number of links. The effect of parameter variations on the BER performance is studied. These parameters include: the total number of iterations, the codeword length, the total number of rounds, and the coding rate in the JSP algorithm. An optimal value of the parameters is selected during the design procedure of a communication network by considering its limitations and complexity criterion. The JSP algorithm is a reliable scheme for jointly decoding of noisy binary data from different origins.