The primary objective of fog computing is to minimize the reliance of IoT devices on the cloud by leveraging the resources of fog network. Typically, IoT devices offload computation tasks to fog to meet different task requirements such as latency in task execution, computation costs, etc. So, selecting such a fog node that meets task requirements is a crucial challenge. To choose an optimal fog node, access to each node's resource availability information is essential. Existing approaches often assume state availability or depend on a subset of state information to design mechanisms tailored to different task requirements. In this paper,OptiFog:a cluster-based fog computing architecture for acquiring the state information followed by optimal fog node selection and task offloading mechanism is proposed. Additionally, a continuous time Markov chain based stochastic model for predicting the resource availability on fog nodes is proposed. This model prevents the need to frequently synchronize the resource availability status of fog nodes, and allows to maintain an updated state information. Extensive simulation results show thatOptiFoglowers task execution latency considerably, and schedules almost all the tasks at the fog layer compared to the existing state-of-the-art.

In this paper, we propose RedgeX, a meta-learning based approach for generating analytical models in a distributed edge intelligence network. The approach involves training a meta-learning model on a large dataset of edge device information and performance metrics to predict the optimal analytical model for a given task and available resources. An edge controller, which has the status of all the edge devices, can then deploy the optimal model to the most suitable edge devices based on their available resources. The RedgeX improves the efficiency and effectiveness of edge intelligence systems by dynamically generating analytical models based on the specific requirements of each task and the available resources in the edge devices. The performance evaluation of the proposed scheme shows better utilization of resources, improved performance, and reduced latency in edge intelligence systems.

Internet of Things (IoT) with edge computing capability enhances efficiency, availability, and improves latency of an industrial automation system. However, to provide dynamic services at the resource-constrained edge device, reconfiguration of services is necessary. This article proposes a programmable edge network to (re)configure different services of industrial IoT, that employs programmable layers at the edge for reconfiguring the sensor/actuator network and application services. The lowermost layer allows reconfiguring the communication-related parameters and the middle layer consists of a software-defined networking (SDN) controller that can dynamically program different modules and handles actuation decisions from the edge. An interfacing protocol between the layers is proposed to provide reliability by considering the required configuration parameters among layers. At the top layer, a priority forwarding mechanism is designed for SDN core (control loop) communication when sensor and actuator are on different edges. The proposed architecture significantly improves the actuation latency and is highly energy efficient compared to the existing state-of-the-art.

In a fog-computing assisted Internet of Things network, end-devices typically offload computation and storage-intensive tasks to fog devices. It is primarily done to meet the latency requirements of tasks, and QoS requirements of the network. In addition to providing localized computing and storage services, the fog network also needs to support end-device mobility while handling offloaded tasks, especially, to mimic the ubiquitous availability of the cloud. Most of the existing works in this direction either recommend task migration or offloading tasks by predicting the device's location. Both these approaches are shown to have their respective limitations, and, thus a mobility-aware task offloading scheme is crucial to meet end-device task requirements. In this paper, we present an approach to handle the mobility of end-devices for effectively handling offloaded tasks. The proposed mechanism is simple, effective, and is not constrained by a device's location, thereby lowering the costs associated with mobility. Especially, the proposed scheme entirely eliminates the cost induced during migration, since effective task offloading can lessen the necessity to attempt task migrations. The simulation result of the proposed scheme reduces execution latency by 44%, saves upto 68% of network usage and 62% of computational cost at the cloud compared to the state-of - the-art.

Due to the multi-hop, long-distance, and wireless backbone connectivity, provisioning critical and diverse services face challenges such as low latency and reliability. This paper proposes ioFog, an offline fog architecture for achieving reliability and low latency in a large backbone network. Our solution uses a Markov chain-based task prediction model to offer dynamic service requirements with minimal dependency on the Internet. The proposed architecture considers a central Fog Controller (FC) to (i) provide a global status view and (ii) predict the type of tasks at the Fog Nodes for intelligent offloading decisions. The FC also has the current status of the existing fog nodes in terms of their processing and storage capabilities. Accordingly, it can schedule the possible future offline computations and task allocations. ioFog considers the requirements of individual IoT applications and enables improved fog computing decisions. As compared to the existing offline IoT solutions, ioFog reduces service time significantly and service delivery ratio up to 23%, compared to the existing relevant architectures.