Deep-Learning-Empowered Edge Computing-Based IoT Frameworks
Mithra Venkatesan, Anju V. Kulkarni, Radhika Menon · Internet of Things · 2022
Internet of Things (IoT) involve communication among millions of smart devices. The massive data produced by ubiquitous smart devices need predictive analysis. This could result in improvements in performances leading to variety of applications. The processing of mass data could be empowered by deep learning models, which could learn quickly from the sensor data and build predictive models. The IoT applications will become seamless if the deep models run on-devices. This concept of running on-devices can be accomplished by edge computing. Implementing machine learning interface on edge device has immense potential and is in preliminary stages. Hence, deep-learning-empowered edge-computing-based IoT frameworks would not only leverage deep learning to empower IoT applications but also offer the additional advantage of running them on-devices through edge computing. This chapter explores the merging of three disciplines of deep learning, edge computing and IoT which will result in a wide spectrum of innovative designs in IoT applications ranging from health monitoring, home robotics, intelligent control etc. This chapter also discusses the usability scenarios and framework for implementation. Furthermore, different issues and challenges in implementations will also be deliberated motivating future research directions in this evolving domain.