Empirical Delay Models for 802.11 under Deterministic Convergecast Traffic

Abstract

Wireless sensor and actuator networks as well as industrial Internet of things networks use protocols such as IEEE 802. 15.4e TSCH, which uses deterministic scheduling, in order to collect data. The data collected from multiple sensor nodes is convergecast to a gateway. Since the scheduling is deterministic the data arrival at the gateway follows a periodic deterministic pattern. Often, in an industrial setting it is useful to take this collected data from multiple gateways and transfer it for cloud processing or storage. We consider an architecture where the convergecast data is communicated over WiFi to the cloud. Although real-time latency requirements are not imposed on this cloud transfer, average latency requirements still apply for use-cases like long term control of parameters or for updation of dashboards. In this context, we analyze the average latency performance of the 802.11 MAC layer with deterministic input traffic. We propose a data driven empirical model for the average delay of 802.11 under deterministic traffic.