Raptor decoding for high overloaded wireless super-dense networks

Abstract

This paper proposes Raptor codes-structured wireless super-dense networks serving a massive number of users with passive nodes constructing a precode of Raptor codes for future wireless networks. Passive nodes do not have their own information and only forward information from the neighboring nodes to the destination node. Similar to the capability of precode of Raptor coding scheme to protect information from error, the passive nodes in the proposed super-dense networks also provide additional coding gain to the networks leading to non-dropping performance of the throughput when the networks are overloaded. The performances of the proposed wireless networks are evaluated under Rayleigh fading channels in terms of packet-loss rate (PLR) and throughput based on computer simulations. The results confirm that the Raptor codes-structured wireless networks have additional capability of performing error correction by preventing the drop of throughput when the networks are overloaded compared to wireless networks without passive nodes. The results are interesting since the passive nodes can work as bridges between random access and scheduled access schemes depending on the available time-slots and computational complexity of decoding.