Quantum ML-Based Cooperative Task Orchestration in Dew-Assisted IoT Framework

The proliferation of Internet of Things (IoT) devices has resulted in an overwhelming surge in data generation, posing a challenge to traditional Cloud computing infrastructures. However, relying solely on this network for delay-sensitive and compute-intensive tasks is impractical due to the potentially catastrophic consequences of even the slightest delay. Additionally, task scheduling poses a significant challenge in these computing paradigms, requiring efficient allocation to computing nodes while considering factors such as deadlines, power utilization, violation of delay, and more. Moreover, challenges arise in managing the cost of services in these paradigms. To mitigate this issue, the presented article incorporates Fog computing layer, offering computing resources at the network’s edge through collaboration with Edge nodes, which are positioned closer to the IoT devices. The Edge layer comprises intermediary computing nodes which are responsible for executing local operations. Consequently, Dew-enabled IoT devices are integrated to ensure uninterrupted services within the network. Consequently, the proposed approach emphasizes conducting most computations in the Edge–Fog layer rather than the Cloud layer. An improved approach by hybridizing Deep Reinforcement Learning (DRL) and Deep Q-Network (DQN) that leverages the principles of Quantum Machine Learning (QML) algorithm called Q-FogSched has been presented in this article to allocate tasks to the Edge, Fog, or Cloud layer. Moreover, a networking model called Dew-enabled IoT-Fog-Cloud-of-Things (DIFCoT) has also been proposed for the implementation of the suggested approach. Using a standard benchmark data sample, the algorithm was evaluated using a Dew-assisted IoT-Fog computing environment using iFogSim simulator. The experimental evaluations show that the proposed methodology significantly improves the efficacy for various Quality-of-Service (QoS) parameters like power utilization, rate of delay, cost of service, violation of delay, and delay of service with an overall rise in efficiency of 49%, 34.3%, 29.2%, 16.2%, and 8.21%, respectively, when compared to some earlier baselines.