Abstract
•The integration of edge computing and Green IoT significantly improves the network performance.•The ISEC model manages energy and information security for data transportation.•The proposed model generates optimal features using deep learning for data routing.
Internet of Things (IoT) consists of a huge number of sensors along with physical things to gather and forward data intelligently. Green IoT applications based on Wireless Sensor Networks (WSNs) are developed in various domains, such as medical, engineering, industry, and smart cities to grow the production. To increase the performance of sustainable cities, communicating nodes are interconnected autonomously to observe the environment, where they need to be more energy-efficient. Edge computing operates in a distributed manner and improves the response time with the least latency through various edge servers. Although the integration of edge computing and Green IoT significantly improves the network performance in terms of computation and data storage, low powered sensors have constraints in terms of battery power, low transmission range, and security aspects. Therefore, adopting an emerging solution is needed to offer energy services with secure data delivery for sustainable cities. This paper presents an intelligent and secure edge-enabled computing (ISEC) model for sustainable cities using Green IoT, which aims to develop the communication strategy with decreasing the liability in terms of energy management and data security for data transportation. The proposed model generates optimal features using deep learning for data routing, which may help to train the sensors for predicting the finest routes toward edge servers. Moreover, the integration of distributed hashing with chaining strategy eases security solutions with efficient computing system. The experimental results reveal the improved performance of the proposed ISEC model against other solutions for energy consumption by 21 %, network throughput by 15 %, end-to-end delay by 12 %, route interruption by 36 %, and network overhead by 52 %.