As a key enabler of the Internet of Things, cellular network based communications between devices (things) have been growing rapidly in recent years, being used in a wide range of services such as security, metering, health, remote control, tracking, and so on. However, a critical issue in this communications is the energy efficiency as typically these devices are powered by batteries of low capacity. Therefore, it is of special interest to optimize the energy consumption to extend the operative lifetime of the devices.
To approach this issue, these systems require intelligent algorithms to modify behaviors of the devices as well as the overall network so they can dynamically adapt their settings to improve the energy efficiency. This idea can be achieved by making the devices aware of their states by means of contexts, which have been studied in computing systems to allow devices to adapt automatically to different situations and modify their operative modes for better performance.
To make this awareness possible it is necessary to identify such contexts. But what exactly is a context? By summarizing definitions by different sources it can be said that if a piece of information can be used to characterize a situation of a device, then that piece of information is a context; i.e. location, environmental information, settings of the application such as data reporting frequency, average packet size etc. As for context awareness, names such as adaptive, responsive, and context-sensitive are associated to this term. If the machines are taught to identify, sense, interpret and react automatically to their contexts then it can be possible to maximize their energy efficiency.
The standardization and regulator entities such as the 3GPP and ETSI are working on the standardization, but as it’s a demanding task due to the vast amount of applications, their diverse traffic characteristics, and quality of service (QoS) requirements. Even if the standards are not in place and some of the context-aware optimization require it, a lot can be done in specific solutions. For example, if a sensor only needs to report its data (e.g. temperature) once every five minutes and the reporting takes six seconds, it can put itself in sleep mode for 98% of the time and thereby extend battery life by 50 times.
Do you see other application where this is applicable?
Author: Javier Mendonca Costa
This article is based on a bigger research that can be found here:
Context-Aware Machine to Machine Communications in Cellular Networks