research trends in fog and edge computing

Research Trends in Fog and Edge Computing

Spread the love

This blogpost is about the recent trends in fog and edge computing research. Recently, due to the excessive growth in data and applications generating heterogenous type of data, the need to perform data analytics efficiently has also increased. Consequently, in addition to the cloud computing, the concepts of fog computing, edge computing, and Internet of Things (IoT) have been materialized. With the inception of the aforementioned paradigms, the scientific community and researchers have started benefitting from their potential to overcome the limitations of the conventional models and paradigms. Particularly, fog and edge computing have gained popularity because of their ability to bring the computation and analytics tasks at the edges of the networks or closer to the locations where data is being generated. Consequently, fog and edge computing result in reduced traffic or data transmission to and from the cloud and hence the network usage is minimized by utilizing the fog services. Moreover, processing the data in the fog or near the place of its origination results in minimizing the latency as compared to the cloud.

Because of the above benefits of fog and edge computing, their application in different domains is also increasing. Therefore, in this article we highlight the recent research trends in the areas of fog and edge computing. We hope that this article will be helpful for researchers particularly for the graduate research students to conceive the ideas for carrying out the research.

Machine Learning Models for Edge Devices

Due to the capability to process information efficiently, recently researchers have started working on the development of machine learning models for edge devices. This not only brings the intelligence closer to the network edge but also minimizes the associated problems of processing the data at the cloud. Moreover, lightweight techniques that allow edge devices to train the machine learning models locally and offline are also being worked on by the researchers.

Energy Efficiency in IoT and Edge Computing

Most of the devices connected in the IoT environment are battery powered with limited battery time. In addition, frequent exchange of information among the sensors further increases the energy consumption. Few interesting methodologies capable of selectively offloading computations into the edge devices have been proposed. Therefore, investigating about the development of energy efficient techniques for the fog and edge computing environment is a promising research direction.

Trust, Security, and Privacy Issues in Fog Computing Systems

Trust, security, and privacy of the data and information in the distributed environments are among the key issues. Several lightweight schemes for resource constrained IoT devices have been proposed. Further possible directions for research to pursue include Privacy Disclosure, Statistical Methods for Inferring Patients’ Information in Networks, Anonymity and Differential Privacy, Attacks and Privacy-preserving Mechanisms, Models of Information Sharing, Users’ Privacy Risk, and Management of Privacy Settings.

Scalability and Fault Tolerance in Fog Computing Environment

Various scalable and fault tolerant techniques for fog computing have been proposed to handle the scalability and reliability issues. Therefore, researchers can opt to investigate further about fault tolerance, reliability, and availability of on-demand information in distributed environments.

Load Balancing and Scheduling on Fog Servers in the Industrial Environment

In data intensive environments, in the wake of excessive load, the fog nodes might contain too many service requests. Consequently, the applications may suffer unnecessary processing delays. To handle such issues, methodologies capable of balancing the load on fog devices are needed. Therefore, in this regard, the researchers may attempt to come up with some suitable techniques to solve the problems of dynamic load balancing, resource allocation, and task offloading.

Fog Computing for Smart Homes and Buildings, and Industrial Environment

The concept of smart homes and smart buildings has recently been materialized with the emergence of IoT solutions. Therefore, to manage multiple home devices generating data continuously, fog-based solutions are a better alternative to administer the delay sensitive data. Similarly, fog computing has demonstrated its potential in Industrial Internet of Things (IIoT) by connecting the industrial processes, devices, and machines with the Internet to efficiently and effectively oversee and manage the manufacturing processes. Therefore, researchers may find several interesting research problems in the IIoT domain.

Fog Computing Architectures for Remote Health Monitoring

With the evolution of Healthcare 4.0, the trend for remote healthcare delivery services and remote health monitoring has increased. The doctors can observe the patients remotely with the help of smart IoT devices. To offer the efficient delivery of healthcare services, data analysis and feature extraction tasks can be performed at the gateways. Performing these tasks near the gateways will not only reduce the network bandwidth utilization but will also minimize the latency.

Fog Computing for Smart City Applications

The IoT phenomenon has also greatly impacted the concept of smart cities because the availability of several types of sensors and services have made tasks, such as surveillance, traffic monitoring, identification of hazardous events etc. relatively easier. The massive amount of big data in smart cities can efficiently be processed by deploying the fog and edge computing services. Therefore, proposing the new architectures and methodologies to ensure sustainable smart cities are one of the potential areas for the researchers to work on.

Efficient Urban Surveillance using Fog Computing 

Developing the architectures for urban surveillance to analyze the traffic in real-time from video streams using the advanced machine learning techniques near the network edges or gateways to identify the criminals or suspicious activities etc. can definitely contribute towards the development of safe communities. Analyzing the collected data close to the location of its creation on edge or fog devices can help in rapid dissemination of important information regarding any undesirable incident, crime, or suspicious people.

Likewise, there are several other interesting problems domains that can be explored, for example, (i) virtual learning environments supported by fog computing, (ii) virtualization in fog computing, (iii) fog-based architectures for smart grids, (iv) experimental evaluation of data intensive fog frameworks in industrial settings, (v) employing fog computing with Intelligent Transportation Systems (ITS) in the Internet of Vehicles (IoV), and (vi) efficient resource allocation techniques for Internet of Things (IoT) and fog networks.

Although the topics and research trends described in this article are not detailed and exhaustive but for sure it can provide readers a good starting point toward the fog and edge computing research.

Read about virtualization in cloud computing.