Evaluating Multiple-Access Protocols: Asynchronous Pulse Coding vs. Carrier-Sense with Collision Avoidance

While the Internet of Things (IoT) is usually envisioned to support powerful functionality, like in self-driving cars, there is also increasing interest in simpler IoT applications that can be employed on massive scales at high densities, like in data gathering at meetings with large audiences. The latter vision requires low-cost devices consuming little energy, and it tends to come with a relaxed need for high-speed communication. Its realization necessitates the development of wireless protocols that are simple, yet that can effectively arbitrate multi-access to communication channels. Carrier-Sense Multiple Access with Collision Avoidance (CSMA/CA) is usually deployed in such contexts, but it tends to work less well when large numbers of nodes attempt to simultaneously access a wireless channel. Asynchronous Pulse Code Multiple Access (APCMA) has been developed with simultaneous asynchronous access to communication channels in mind by using a sparse representation of pulses to encode messages, but being relatively recent, its performance has never been systematically compared to CSMA/CA. This paper compares APCMA’s performance with that of CSMA/CA in terms of the success probability (i.e., absence of errors) of message transmissions through the use of simulations and analytical models, under the assumption of equivalence in throughput. We find that APCMA with four pulses per code word performs worse than CSMA/CA, but the roles are reversed if five or six pulses are used.