Massive Connectivity and Low-Latency for Next-Generation Internet of Things: A Filtered OFDM-based Deep Learning Approach

Abstract

Abstract The Internet of Things (IoT) is one of the numerous services offered by sixth-generation (6G) mobile communications. It is necessary to meet heterogeneous criteria for large connectivity and low latency to serve the various kinds of IoT applications. In order to concurrently provide enormous connectivity and meet the low-latency conditions in the uplink IoT network, we proposed filtered orthogonal frequency division multiplexing (FOFDM) based on a service group adopting the deep learning (DL) technique. The proposed FOFDM-DL platform focuses on two key areas: first, it works for the concurrence of different time-frequency granularity appropriate for distinct service-based grouping, and secondly, it facilitates low latency and massive connectivity to deliver dependable communications. The suggested FOFDM-DL architecture may accommodate the needs of massive machine-type communication referred as mMTC and ultra-reliable and low-latency communications known as uRLLC concurrently for the next-generation communication systems. However, the uRLLC and mMTC requirements can only be supported separately by the new radio (NR)-5G, beyond 5G (or 6G). In comparison to the traditional scheme, simulation results demonstrate that the suggested FOFDM-DL platform works surprisingly well in the next-generation IoT network.