Energy efficient software update mechanism for networked component-based IoT devices

Abstract

Due to security issues and incremental user requirements, software in IoT devices needs to be changed frequently to improve existing functionalities or to fix bugs. Software updates have become an integral task of IoT systems to maintain effective operations. Recently, the common software architecture in advanced IoT devices is component-based, in which components can be updated at run time. In such IoT networks, devices can download updated components from neighbor nodes, enabling quick deployment of updates. In this context, there are two main issues in the distribution of software components that needed to pay attention: i) how to deliver updates to all devices in an energy-efficient way, and ii) how to quickly deploy updates to avoid long network downtime. In this thesis, we propose a mechanism that schedules updates on all devices in an IoT edge network with the goal to minimize the energy consumption, taking into account the deadline constraint for updating the entire network. Unlike previous studies on IoT component-based software update, which often focus on how a single component is replaced in the operating system, we focus on the distribution of components in the network and investigate the update process happened in the flash memory of a device, in which the order of re-written components into the memory is decisive for energy consumption. We introduce a novel energy model of the update process inside a device, focusing on the flash re-writing operation which consumes a significant amount of energy in the update process. Then, we formulate a mathematical optimization model for the problem of energy efficient update scheduling. Because of the high complexity of the problem, we then propose an algorithm called ESUS to approximate the optimal schedule for updating all devices in the network. To evaluate our scheduling algorithm, we compare the results of ESUS to the optimal solutions given by a mathematical solver. Simulation results show the efficiency of our method, which is close to optimal scheduling solution with much lower execution time compared to the solver.

Date	12 Dec 2019
Original language	American English
Awarding Institution	École de technologie supérieure
Supervisor	Kim Khoa Nguyen (Supervisor) & Mohamed Cheriet (Co-supervisor)