Title:
Coding for Super Dense Networks

Speaker:
HASAN, Mohammad Nur (JAIST)

Abstract:
Multiway relay network (mRN) with capability of serving multiple users is regarded as one of solution for future networks challenge, where massive number of users are connected forming super dense networks. To avoid high complexity scheduling involving huge number of users, uncoordinated transmission scheme is considered in MWN. The users are allowed to randomly transmit their messages to the networks without having coordination with other users. The networks itself can be seen as a coding structure resembles low-density parity-check (LDPC) codes, which are usually associated with bipartite graph. Therefore, we consider a graph-based successive interference cancellation (SIC) technique to effectively decode the messages transmitted uncoordinatedly. We further proposed the use of iterative demapping (IDM) algorithm in the SIC to increase the success rate of decoding process. We derive the bound of the proposed system and confirm an achievable point using practical coding. We also evaluate the performance of the proposed MWN via computer simulations. The results confirm that a reliable communications is achievable even though the number of slots is less than number of users, yielding significant throughput improvement. The results also show that the proposed mRN outperforms the mRN employing conventional coded slotted ALOHA (CSA) in terms of throughput and packet-loss-rate (PLR) performances.