Indonesian Journal of Electrical Engineering and Computer Science Vol. 25, No. 1, January 2022, pp. 375~381 ISSN: 2502-4752, DOI: 10.11591/ijeecs.v25.i1.pp375-381  375 Journal homepage: http://ijeecs.iaescore.com A weighted group shuffled decoding for low-density parity- check codes Fatima Zahrae Zenkouar 1,2 , Mustapha El Alaoui 2 , Said Najah 1,2 1 SIA Laboratory, Faculty of Sciences and Technologies, Sidi Mohammed Ben Abdellah University, Fez, Morocco 2 Laboratory of Computer Science, Department of Physics, Signals, Automation and Cognitivism (LISAC), Faculty of Sciences Dhar El Mahraz, Sidi Mohamed Ben Abdellah University, Fez, Morocco Article Info ABSTRACT Article history: Received Mar 12, 2021 Revised Nov 16, 2021 Accepted Nov 20, 2021 In this paper, we have developed several concepts such as the tree concept, the short cycle concept and the group shuffling concept of a propagation cycle to decrypt low-density parity-check (LDPC) codes. Thus, we proposed an algorithm based on group shuffling propagation where the probability of occurrence takes exponential form exponential factor appearance probability belief propagation-group shuffled belief propagation (EFAP-GSBP). This algorithm is used for wireless communication applications by providing improved decryption performance with low latency. To demonstrate the effectiveness of our suggested technique EFAP-GSBP, we ran numerous simulations that demonstrated that our algorithm is superior to the traditional BP/GSBP algorithm for decrypting LPDC codes in both regular and non- regular forms. Keywords: Belief propagation EFAP-GSBP Girht GSBP LDPC codes Low latency This is an open access article under the CC BY-SA license. Corresponding Author: Fatima Zahrae Zenkouar SIA Laboratory, Faculty of Sciences and Technologies, Sidi Mohammed Ben Abdellah University Fez, B.P. 2202–Route Imouzzer, Fez, Morocco Email: fatimazahrae.zenkouar@usmba.ac.ma 1. INTRODUCTION In the literature, authors have often used the concept of belief propagation (BP) and the concept of sum and product (SPA) method to control the parity of low-density parity-check codes (LDPC) with good performance. But the concept of sum and product (SPA) decoder has several limitations such as fast divergence and the great difficulty to implement this code. In the literature some authors have proposed solutions to solve these problems. Thus, the authors of the paper [1] have proposed a method based on serial BP decoding. Their algorithm focuses on the variable nodes of the bipartite graph by dividing them into several groups. In the paper [2] the authors focused on the control nodes (CN) of the bipartite graph by dividing them into several groups. Other authors [3]-[6] have used the concept of serial and parallel decoding through sequential groups of messages to achieve reliable decryption of extrinsic messages within an iteration. Moreover, the authors of the article [7], [8] have shown that the decryption method based on the horizontal group shuffling principle (HGSBP) is the easiest algorithm to implement. Thus we will also use this algorithm as a development of our proposed decoding method. Indeed, we will associate each control node to a group. In this case, each decryption operation will consist of more sub-operations. These sub-operations will initialize the log-likelihood measures generated at the virtual node (VN) variable nodes that are linked to the CNs of the same group in parallel in each group. Moreover, in this proposed form, the messages will flow within the two-part sub graph consisting of the group's CNs and all the VNs that are connected to its CNs. Thus, considering the CNs and VNs as belonging to the