Enabling Transient Computing for Unreliable Internet of Things: A Paradigm Shift for Sensor Systems (PROCEEDING)

The context of the research and applications:

Technological advances and changes in the social perception of technology are fueling interest in small, flexible, wearable and implantable devices for the Internet of Things (IoT). The IoT describes the network of physical objects (things) embedded with sensors, hardware and software, and other technologies to connect and exchange data with other devices over the Internet.

These aim a range of application domains, including agriculture, home and building automation, radio-frequency identification (RFID), and assisted living and health care.

For reasons of installation location, ease, or aesthetics, many of the networked devices which underpin these applications do not have access to a wired electricity source; instead, they rely on batteries as their primary power source. However, battery-powered devices inevitably experience a direct tension between market demands requiring long lifetimes and small physical dimensions and weight.

As a result, research has looked towards replacing batteries with energy harvesting from alternative sources. However, although energy harvesting provides infinite energy in theory, its output varies and frequently fails and is overall unreliable. Its aims and objectives:

To address this challenge, the proposed research project aims to design a new class of sensor systems for IoT applications that are resilient to intermittency and adjust to the varying and unreliable supply such as harvested energy.

To achieve this, the project will first target the design and validation of methods to preserve the system state, or, in other words, to enable quick saving of the system state before a power failure occurs and restoring it once the supply recovers. Secondly, the project will concentrate on using the energy efficiently when available, which means designing and validating strategies for runtime power management to perform reliably and maximise performance and prolonge operations under varying supply conditions.

Finally, the project will focus on communication and networking for sporadic exchange between sensor devices intermittently (or transiently) powered. The problem is that sensor devices cannot receive and forward data if the power supply is unavailable. Therefore, this project will focus on designing transceivers and protocols for an efficient broadcast mechanism under varying supply conditions (energy-aware networks).