Characterizing IoT Networks with Asynchronous Time-Sensitive Periodic Traffic

TL;DR

This paper introduces a comprehensive spatiotemporal model for large-scale IoT networks with asynchronous periodic traffic and deadlines, analyzing success probabilities and latency.

Contribution

It presents a novel combination of spatial Poisson processes and Markov chains to model IoT traffic and interference, providing new insights into network performance under deadline constraints.

Findings

Strict deadlines can improve transmission success rates.

The model reveals counter-intuitive performance benefits of deadline enforcement.

Interference and timing offsets significantly impact success probabilities.

Abstract

This paper develops a novel spatiotemporal model for large-scale IoT networks with asynchronous periodic traffic and hard-packet deadlines. A static marked Poisson bipolar point process is utilized to model the spatial locations of the IoT devices, where the marks mimic the relative time-offsets of traffic duty cycles at different devices. At each device, an absorbing Markov chain is utilized to capture the temporal evolution of packets from generation until either successful delivery or deadline expiry. The temporal evolution of packets is defined in terms of the Aloha transmission/backoff states. From the network perspective, the meta distribution of the transmission success probability is used to characterize the mutual interference among of the coexisting devices. To this end, the network performance is characterized in terms of the probabilities of meeting/missing the delivery deadlines and transmission latency. The results unveil counter-intuitive superior performance of strict packet deadlines in terms of transmission success and latency.