Energy-Efficient Algorithms for Edge Al on lOT Devices

Abstract

Internet of Things Devices: Data Processing, there exist many Internet of Things devices in our life. So we need to implement data processing to Internet of Things devices. We can use Edge Artificial Intelligence in place of cloud computing all the time. Edge Artificial Intelligence can be used for executing data processing to Internet of Things devices. Then it will be more effective to reduce the delay, network usage and privacy. However, most of the Internet of Things devices have limited battery, processing power and memory. So, the Artificial Intelligence algorithm can consume energy in order to run in Internet of Things devices. Then it will reduce the life time and energy efficiency to Internet of Things devices. This research is about that importance of energy efficient algorithms to Edge Artificial Intelligence in Internet of Things environment. This research is about that conventional artificial intelligence models consume very large power. This research is also about that we need to create the algorithms which can run in limited resource environment. We can do many things to do that, for example: creating power efficient model, compressing model, removing unnecessary part of the model, representing data by numbers, and so on and so forth. All of these can reduce the energy consumption while maintaining the accuracy. This research is about maintaining the accuracy of the model while reducing the energy consumption. This is of critical importance in relation to their real-world feasibility. We have an Energy- algorithms device for the Internet of Things that can predict and anticipate what will happen. It can also respond quickly. It can do this without having to replace or recharge its batteries many times.