Self-Coupling Random-Access Communications

A large share of upcoming wireless communication applications is dedicated to the Internet of Things (IoT) and Machine-to-Machine (M2M) communications. Objects of the future, such as parts of smart homes, smart vehicles, robots and automata, will exchange information without any need for human intervention and will perform various tasks jointly. Such M2M and IoT communication scenarios anticipate high density of communicating devices that send short messages on demand, without coordination and reservation of time and bandwidth resources. Energy saving, high reliability, and low latency are on the priority list of features demanded by many such applications including remote control and virtual reality. Current techniques for multiple-access communications, a scenario when a number of transmitters share communication media simultaneously, are not suitable for a massive number of devices accessing the channel without coordination, on demand, at random times, in presence of channel impairments and mobility. The goal of the proposed program and future research is to develop a data communication format for massive asynchronous random-access beyond 5G systems providing a base for M2M and IoT communications infrastructure. In terms of data packets, transmitter, and receiver design, the system needs to satisfy low-latency high-reliability requirements, operate in mmWave channel with mobility and energy constraints, be able to efficiently utilize massive multiple antenna receivers, and be optimal in terms of complexity and throughput. The research methodology includes use of spatial graph coupling techniques, where each message is represented as a graph that connects data bits and time/frequency communication resources. The graphs couple at the receiver into a graph chain that allows for efficient data extraction. To enable random access, we propose to use a self-coupling structure where each packet is represented by a mini coupled chain. When received, even at random times, the chains of individual packets couple into a single graph chain at the receiver. Message-passing algorithms for data decoding and interference cancellation can provide a way to approach optimal performance in the sense of the achievable throughput. The format enables for inclusion of iterative packet acquisition and channel equalization. The expected contribution to the field is a communication format that can be implemented in a multitude of devices and potentially standardized. We target a systematic design based on theoretic principles of optimality, supported by rigorous analysis framework and novel breakthrough techniques of spatial coupling, compressed sensing, machine learning, and channel tracking. The developed technology is expected to lay a pathway to future wireless device connectivity and have a significant impact on the Canadian economy as 5G, IoT, and M2M will serve as a vehicle for the future development of business in the information technology sector.