Introduction

Fog computing is a service started by networking giant, CISCO. It would be very difficult to define fog computing without first defining cloud computing, since fog computing is basically an extension of the cloud.

Cloud computing is the process of running ICT tasks and services and storing computer resources over the Internet. This makes it possible for people and businesses to make use of third-party hardware and software hosted online. Cloud computing makes it quite easy to access information and computer resources from anywhere so far as internet connection is available. With the all-round availability of shared/pooled computing resources, cloud computing offers advantages over traditional on-site hosted services in terms of speed, cost, and efficiency.

Though cloud computing works just fine presently, it relies heavily on the bandwidth made available, which depends on the capacity of the network service provider. With billions of users processing, sending and receiving data in and out of the cloud, the system becomes increasingly cogged up.

Fog computing (also known as fogging), on the other hand, is the extension or lowering of cloud computing capabilities to the bottom/edge of the network in order to provide faster ICT (communication, storage, software, etc.) services to the lower end users. Therefore, what distinguishes fog computing from cloud computing is its closer proximity to small end users, its wider consumer reach, and better mobility.

Rather than requiring devices to go through the network backbone infrastructure, fog computing permits devices to connect directly with their destination with ease and allows them to handle their connections and tasks any way they deem fit. As a result, fog computing improves quality of service, reduces latency, and gives a more satisfactory user experience.

Fog computing smoothly supports the emerging internet of things (IoE)—properties (vehicles, home appliances, and even clothes) that are embedded with sensors to enable them to send/receive data. Fog computing can be implemented using a basic communication system as opposed to being implemented using a heavy backbone network. As a result, it has a denser coverage. This advantage makes it easier to run a real-time, big-data operation with the ability to support billions of nodes in highly dynamic, diverse environments.

Differences between Cloud Computing and Fog Computing

Parameter

Cloud Computing

Fog Computing

Service Provided General data/application and other ICT services hosting.

Localized data/communication exchange services.

Service Provider Large Internet, network service companies Local businesses (shopping mall, logistics companies, transport coy, large vendors, etc.)
Hardware Expensive, robust and hi-tech backbone system with scalable storage and vast compute power Wireless multi-point interface
End Users General ICT services users Mobile users
Distance to Users Hosted in remote locations and can only be reached via IP networks Close to the users and can be reached via wireless (Wi-Fi) connection

Fog Computing System Architecture

Fog computing expands cloud computing by serving as an intermediate between mobile devices and the cloud. This gives rise to three layer architecture as shown below:

The middle fog layer consists of servers that are installed at the edge of IP networks. It is supported by limited data storage, computer and wireless communication tools. The basic role of fog servers is to connect the cloud directly to mobile users.

Fog servers can directly connect with mobile devices using wireless means such as Bluetooth or Wi-Fi. They are independently able to provide ready on-demand service to IoT devices without falling back on an IP network, as they have their own data processing capabilities with pre-cached contents. ISO/IEC 20248 has made available an identification system for IoT-capable devices to be detected using automated identification data carriers (AIDCs), RFID tags or barcodes that can be read and verified to enable connection to the fog.

Potential for New Application Services

Fog computing has the potential to support a new set of innovative applications and services:

  • Smart Switching Systems: Devices that require constant switching to alternative service providers in case their main service provider becomes unavailable can automatically switch to alternative service if the sensor embedded in the device feels the need to do so. For example, if a network gateway is becoming congested, the device will automatically switch to another network.
  • Smart Cities: The use of self-regulating technological systems to keep an area well organized would give rise to smart cities. For example, traffic light sensors that spot holdups on a particular road can delay/divert traffic to less congested roads in order to free up the congested road.
  • Smart Home Appliances: Sensors embedded in self-maintaining home appliances can enable them to monitor their components and environment. If they detect anomalies, they send an automatic alert to the owner with a detailed report of the problem.

With fully deployed fog computing systems, business service providers across industries can develop, run, and manage their assets directly through networked devices. This includes IP video cameras, switches, and hardened routers.

Fog computing makes it much easier to manage the large amount of data that will be generated in a fully connected world. It will result in the introduction of better products and services, such as self-regulating, self-organizing, and self-diagnosing products. For now, some parts of the world are already using reliable 4G technologies, while the rest of the world is gradually catching up. We now have mobile devices that can rival many PCs in terms of processing power.

It is therefore imperative for us to move from relying on traditional network service providers having huge data flow back and forth over networks to central data centers, in order to avoid looming bandwidth and latency bottlenecks. It makes sense to build an intermediate layer on the edge of the network to support backbone network infrastructure. Fog computing does not aim to replace, but to help reduce processing and bandwidth burden of cloud systems

Advantages of Fog Computing

Unified Expanded Coverage: IoT devices are usually distributed over large areas, including mobile environments such as vehicles, railways, and utility substations. Fog nodes can be comfortably installed on all these devices without the need for extensive configuration. This way, a large unified fog platform is developed to cover a wide area. Any device can connect to the fog so long as it is within the range of the fog node.

Efficient Data Management: Fog computing systems can be programmed to control, reduce, and organize data produced by IoT devices. Required data is collected, analyzed, and cached at the network edge, while less sensitive data is sent to the cloud for backup or further analysis.

Redundancy and Failover: Fog servers, routers, and switches are built with hi-intelligence technology. This enables network-wide intelligence and resilience to enable its scale and manage millions of new endpoints and applications. Fog computing increases reliability and availability of service for IoT devices and data.

Improved Security and Privacy: Cisco security systems are extended from the cloud down to the fog. Biometric authentication checks such as fingerprint, face, keystroke, touch-based or other authentication methods in mobile and cloud computing are also applicable to fog computing. Privacy protection cryptology such as differential privacy can be implemented between the fog and cloud to ensure non-disclosure of personal records. Hence organizations can reap benefits of the fog without sacrificing customer security and privacy.

Lower Operating Expenses: Fog data services conserve network bandwidth, as less of the data is analyzed and stored. Also, as a result of the automatic maintaining and self-troubleshooting capabilities of IoT devices, organizations’ operating expenses are drastically reduced.

New Opportunities and Innovation: Fog services will increase business opportunities by spawning development of new services. Business owners can accurately monitor their assets deployed in the field, closely study customer behavior, introduce new trends, and ultimately generate more revenue as quality of service is improved.

Conclusion

Advocates of cloud computing might boast that majority of the world’s computing operations will be done from the cloud in the near future. What they do not disclose is that sending data in and out of remote cloud servers can be very tasking and requires heavy bandwidth provided by an expensive, high-speed IP network backbone. As device manufacturers begin to churn out IoT-compliant devices, everything from baking ovens to commuter buses are being embedded with sensors. At the rate data is being generated presently, mobile networks will soon be unable to cope with demand for faster bandwidth.

Fog computing solves this problem by forming a bridge between IoT devices and the cloud. Fog computing was developed to satisfy the predictable service demands of mobile users. Fog servers are developed with three-dimensional hardware resources (communication, compute, and storage) required to discharge their duties efficiently. Fog computing also brings with it improved service delivery, better bandwidth, and network management and improved security and privacy. Therefore, fog computing is the future deal-breaker.

Reference:

*Image Credit: Prismtech on prismtech.com

*Andrew Froehlich: “IoT: Out Of The Cloud & Into The Fog”; Information Week Network Computing, 08-19-2014 Available online at: http://www.networkcomputing.com/cloud-infrastructure/iot-out-cloud-fog/345618829

*Christopher Mims: Forget ‘the Cloud’; ‘the Fog’ Is Tech’s Future; Wall Street Journal, 18-05-2014. Available online at: http://www.wsj.com/articles/SB10001424052702304908304579566662320279406

*John Apostolopolos: From the Cloud to the Fog; Cisco Technology Radar, Available online at https://techradar.cisco.com/technology/fog-computing

*Shanhe Yi, Cheng Li, Qun Li: A Survey of Fog Computing: Concepts, Applications and

Issues; Mobidata’15, June 21, 2015,

*Tom H. Luan, Zhi Li et al: Fog Computing: Focusing on Mobile Users at the Edge; arXiv:1502.01815v1 [cs.NI] 6 Feb 2015