For those working or interested in working in technology, it’s essential to understand the concept of fog computing applications.
It might be used in various contexts, from factories and warehouses to medical institutions.
However, it is crucial to understand fog computing, how it operates, and its advantages and disadvantages. In the following sections, you will learn more about them-
What is Fog Computing?
Fog computing is a horizontal, physical, or virtual resource paradigm between intelligent end-devices and conventional cloud computing or data centers in the March 2018 edition of NIST Special Publication 500-325, Fog Computing Conceptual Model.
The definition takes on a lot of Cisco’s marketing jargon. Scalable, layered, federated, distributed computation, storage, and networking are all possible with this architecture, making them ideal for latency-sensitive, vertically-separate applications.
In the theoretical paradigm, fog computing application nodes are located between edge nodes and centralized cloud computing and provide a similar purpose to both.
There is no agreed-upon definition of even the most fundamental architectural terms, such as “smart,” and the distinction between “fog computing” and “edge computing” is murky at best.
Fog computing may service a wider area and is closer to end users than traditional cloud computing functions.
In addition, examples of fog computing architecture in IoT pave the way for unhindered information, processing, and communication between end-user gadgets and remote servers.
History of Fog Computing
In January 2014, Cisco introduced the concept of fog computing application.
Fog computing uses nodes near the host and cloud since fog is associated with low-lying clouds. The goal was to physically have the system’s processing power near the host computer.
When Cisco, Microsoft, Dell, Intel, Arm, and Princeton University came together in 2015 to create the OpenFog Consortium, Cisco was among the founding members.
Besides the companies mentioned earlier, Foxconn and Hitachi contributed significantly to this partnership.
The consortium’s primary objectives were to spread awareness of and set criteria for fog computing. In 2019, the group joined forces with the Industrial Internet Consortium (IIC).
Characteristics of Fog Computing
Heterogeneous nodes and networks, numerous geographically scattered nodes, sensitivity to a physical location, the necessity for mobile device support, and low latency define fog computing architecture in IoT.
Below are the major characteristics of the Fog Computing system-
- Low Lag Time
- Multi-Tenancy
- Help with Mobility
- Fast Communication
- Context Awareness
- Connectivity
- Real Time Analytics
- Commercial Use
- Sensor Networks
- Interconnected Nodes
Low Lag Time
The time required to receive, process, and complete a computational request. Due to their closeness to edge devices, fog nodes may speed up the processing of tasks and the analysis of answers.
End-user assistance that is not uniform
As the requested IoT device gets closer to the processing node, end-user support improves.
Multi-Tenancy
Describes a setup where numerous users may access and use the exact program copy. Commonly used infrastructures are known as “shared systems.”
Since the fog platform is decentralized and heavily virtualized, it naturally took to this.
Help with Mobility
It is a feature that lets Internet-of-Things devices move freely between access points without losing registration information.
Due to the potentially disastrous consequences of data loss or delay when the device is in motion, mobility support is a crucial feature for mobile IoT systems.
Therefore, direct communication between the fog node and the IoT gadget is necessary.
Fast Communication
A system is said to be “real-time” if it must provide an answer within a specific time. Real-time e-health, traffic transmission, airlines, and industry-critical process monitoring systems are examples of this sort of fog computing application.
Furthermore, fog computing architecture in IoT favors real-time transmission over batch processing to provide QoS for its users.
Context Awareness
By obtaining data about their networks, fog nodes may better understand how to operate and make informed judgments.
Moreover, extensive geographical coverage is also a great point for fog computing.
The architectural framework of fog computing architecture in IoT enabled the paradigm to support widespread geographical distribution, guaranteeing the provision of quality service.
Connectivity
Increase fog computing architecture in IoT transparency and interoperability by adopting open standards and providing third-party systems.
Especially with the ability to execute models and processes through web service calls and share the results with other services.
Real Time Analytics
Instead of transferring data to a central location, fog computing processes and uses data in real-time.
Lightweight computer units at the data source at the network’s edge do this without a central data center.
These devices may do initial processing on the data locally before sending it to the cloud, where it can be further analyzed. This allows for more efficient data processing, less waiting time, and less data sent.
Commercial Use
This feature in fog computing applications allows for more efficient and effective communication and connectivity across sectors.
The fog node layer may also be used to stream data.
Fog nodes at the bottom of the hierarchy, such as those running on a single computer, may evaluate data, identify anomalies, and take action to fix problems on their own if given the green light.
Sensor Networks
The goal is to keep an eye on the fog’s surroundings, and the Smart Grid is another example of a distributed system that needs decentralized processing and storage.
Interconnected Nodes
This is shown by sensor networks in general and the Smart Grid in particular.
It also refers to the large number of nodes on the fog’s network, which is spread out over a large area.
How Does Fog Computing Work?
Writing or porting Internet of Things applications at the network edge for fog nodes using fog computing software, a packaged fog computing program, or other tools is integral to putting fog computing into action.
Edge nodes, located closest to the network’s periphery, collect information from other edge devices, such as routers and modems, and forward that information to the best possible place for analysis.
Fog computing’s primary purpose is to enable a group of nodes to receive data from the Internet of Things devices in almost real-time.
At regular intervals, the nodes will upload a summary of their data analytics to the cloud.
The collected data from the numerous nodes are then analyzed by a cloud-based application that aims to provide valuable insights.
Different scenarios call for different connectivity solutions. If you want to install an Internet of Things sensor in a factory, you can utilize a wired connection.
But a resource that is always on the go, like an autonomous car, or in a remote location, like a wind turbine in the middle of a field, will need a different connection method.
A high-speed connection is essential to evaluate data in near-real time, making 5G an attractive choice.
Connecting fog and cloud computing networks requires managers to determine which data is the most time-sensitive before proceeding.
Very time-sensitive data should be examined as near to its point of generation as feasible, preferably inside closed and confirmed control loops.
Based on the aims of the analysis, the nature of the data, and the user’s current requirements, the features of fog computing mandate that each kind of data defines which fog node is the optimal site for analysis.
Types of Fog Computing
Fog computing refers to bringing characteristics of cloud computing and related services to the network’s periphery in a business setting. It enables the distribution of data, apps, and other resources closer to, or even on top of, the consumers who will use them.
Listed below are the four most common varieties of fog computing.
- Device-level Fog Computing
- Edge-level Fog Computing
- Gateway-level Fog Computing
- Cloud-level Fog Computing
Device-level Fog Computing
Low-power gadgets like sensors, switches, and routers may use this. It may collect information from these gadgets and upload it to the cloud.
Edge-level Fog Computing
Network edge servers or appliances use this software. These devices can process data before uploading it to the cloud.
Gateway-level Fog Computing
Devices that mediate communication between the edge and the cloud host the software. This way, only necessary data is delivered to the cloud, and traffic is managed.
Cloud-level Fog Computing
This operates on remote servers or cloud-based equipment. These gadgets may process information before transmission to the final recipients.
Benefits of Using Fog Computing
Industrial control systems, video monitoring, and autonomous cars employ fog computing examples for fast reactions.
It may also offload computationally intensive activities from centralized servers or offer backup and redundancy in the case of a network failure.
Fog computing examples are implemented for several reasons-
Reduce Delays and Maximize Output
Fog nodes, often put near the network’s edge and are therefore closer to the IoT devices themselves, can drastically cut down on processing time and boost performance for low-latency applications.
Offer Better Judgment
Since fog computing architecture in IoT enables real-time data gathering and analysis from IoT devices, it may aid in improving real-time decision-making.
Saves Money
Data storage and analysis costs are two areas where fog computing might assist.
Fog computing’s migration of computing and data storage closer to the network edge reduces the amount of data transmitted back to a central point for processing.
Components of Fog Computing
The deployment of a fog computing system may take several forms. Below, we break down the standard components across all of these designs.
- Physical & Virtual Nodes
- Fog Nodes
- Monitoring Services
- Data Processors
- Resource Managing
- Device Protection
- Applications
- Managing Information
- Record Keeping
- Connectivity
Physical & Virtual Nodes
Application servers, edge routers, end devices like mobile phones and smartwatches, and sensors all function as points of interaction with the physical environment.
These gadgets are information generators whose capabilities may cover a wide range of technologies.
However, This implies that their underlying software and hardware and their storage and processing capacity may vary.
Fog Nodes
Fog nodes are standalone devices capable of picking up the resulting data. There are three fog node types: devices, servers, and gateways.
These nodes operate as data repositories, whereas fog servers do the computations required to determine the next steps.
Connecting fog devices to fog servers is standard practice. When communicating with many fog devices or servers, a gateway is required.
Moreover, this layer controls the processing and information transfer rates, making it crucial.
Fog node setup requires familiarity with various hardware configurations, the devices they control directly, and network protocols.
Monitoring Services
APIs are often a part of monitoring services since they track metrics like system uptime and resource use. All endpoints and fog nodes must be up and able to communicate without interruption; therefore, monitoring mechanisms must be in place.
Occasionally, contacting the cloud server directly is cheaper than waiting for a node to become available.
The monitor will handle any such occurrences. Using metrics, administrators may assess the system’s present state and plan for the future regarding resource allocation.
Data Processors
Data processors are apps that function on fog nodes to process data. They cleanse, compress, and even recreate corrupted information from the endpoints.
A data processor may keep information in a local fog server or send it to the cloud for more permanent storage.
These processors standardize data from many sources to make it more portable and transferable.
That is why all parts of the system need to have a consistent and adaptable interface.
Specific computers can fill in the blanks using primary data if one or more sensors fail. This eliminates the possibility of an unsuccessful application.
Resource Managing
In fog computing, each node is autonomous but must coordinate its actions with the others.
The resource manager organizes data transfers between nodes and the cloud and allocates and reallocates resources to different nodes.
It ensures there is no data loss by handling backups automatically.
Fog components require high availability since they consume part of the cloud’s service level agreements.
The manager of available resources coordinates with the observer to pinpoint times and locations of peak demand. This eliminates the possibility of data and fog server duplication.
Device Protection
Incorporating security measures from the ground up is essential for fog components because of their close interaction with unprocessed data sources.
Since most interactions occur via wireless connections, encryption is essential. In certain circumstances, end users make direct data requests to fog nodes.
Therefore, the security measures in fog computing include the administration of users and their permissions.
Applications
End users get tangible benefits from applications. Information from the fog computing system ensures efficient and effective service delivery at a reasonable cost.
An abstraction layer must control these parts by making a standard set of protocols and interfaces available for all of them to use.
In fact, developers often turn to APIs and other web service forms to do so.
Managing Information
Edge devices do it themselves instead of sending data to a centralized place for processing. The end effect is faster response times and less lag.
Record Keeping
Instead of transferring data to a centralized server, edge devices may save it on their local storage.
Moreover, this not only shortens response times but also increases security and privacy.
Connectivity
For fog computing to function, edge devices must be able to communicate quickly with the rest of the network.
Either wired or wireless methods can do this.
Examples of Fog Computing
While cloud computing has grown commonplace, fog computing has recently emerged to solve the latency problems that have plagued Internet of Things gadgets.
Intelligent Houses
Smart homes are one of the most popular applications and examples of fog computing applications.
Smart lighting, programmable blinds, and sprinklers, innovative intercom systems to connect with people inside and out, and an intelligent alarm system are all components of a modern “smart home.”
Alongside climate control systems that are operated by technology, such as the Nest Learning Thermostat.
The usage of fog computing applications allows for the development of unique alarm clocks. Sprinklers, for example, may be activated automatically depending on time and temperature.
CCTV
Given the size and complexity of continuously streaming movies via networks, video surveillance is likely the most common use of fog computing examples.
The latency issues and network difficulties stem from the characteristics of the data at hand. The storage of media material also often incurs considerable costs.
Many cities and towns have installed video surveillance systems in public spaces like malls and streets.
As an example of a fog computing application, fog nodes may alarm the proper authorities if crowd behavior deviates from the norm.
Smart City
Every aspect of life in a “smart city,” from waste collection to traffic control, must be mechanized. Fog computing architecture in IoT is beneficial for managing traffic flow.
Sensors can identify pedestrians, bikes, and cars at traffic lights and road barriers.
Their speedometers can tell them how close they are to colliding with another vehicle. For data collection, these sensors use cellular and wireless networks.
Depending on the data these sensors collect, traffic lights may automatically change to a red or extended green state.
Healthcare
Hospitals and other healthcare providers must comply with HIPAA, making healthcare one of the most regulated industries.
In this field, people continually try new things and react quickly to crises like a dip in vitals.
Health apps and wearables can detect early signs of physical discomfort.
Since even a few seconds of delay might significantly impact a life-or-death event like a stroke, latency problems should never affect this data.
Others
Fog computing architecture in IoT is also used in the hospitality, government, oil and gas, and military sectors.
Smartwatches and other wearables are only two examples of the many platforms on which personal assistants like Siri and Alexa are accessible.
As a result of its adaptability and pervasiveness, the fog computing architecture in IoT is expected to have a significant impact on various markets.
A company’s cloud infrastructure has to include fog computing examples if it intends to provide real-time solutions to its customers.
Issues with Fog Computing
Many devices at many gateways provide security risks in fog computing environments.
Due to the unique nature of each device’s IP address, hackers may easily impersonate you to access whatever data you may have saved in a given fog node.
Educate yourself on the challenges of fog computing-
Authenticity Issue
Since fog computing services are provided on a massive scale, authentication is a major challenge.
Cloud service providers, ISPs, and even end users all have the potential to offer fog service.
Such adaptability, however, undermines confidence in the fog’s overall structure. A rogue fog node is a fog gadget that acts legitimate in order to trick users into connecting to it.
After being connected, it may simply launch assaults by manipulating the user’s cloud-to-user communications.
Privacy
When dealing with a large number of networks, privacy considerations inevitably arise. Network security is a major issue for fog computing since it relies on wireless technologies.
Due to the high density of fog nodes, sensitive data is increasingly being sent from end users to the fog nodes.
Security
Many devices at many gateways provide security risks in fog computing environments.
Due to the unique nature of each device’s IP address, hackers may easily impersonate you to access whatever data you may have saved in a given fog node.
Misty Servers
Fog servers should be strategically placed so that they can best serve their users.
Before deciding where to put a fog node, the business should assess the expected workload and the current demand.
Use of Resources
Since there are so many active fog nodes in a fog environment, computing in a fog environment consumes a lot of power.
It is in the best interest of businesses to reduce the power consumption of fog nodes as much as possible.
Future of Fog Computing
Services and applications that are very sensitive to latency, such as real-time gaming, real-time streaming, and augmented reality, cannot fully use cloud computing.
Due to the proximity of cloud data centers to the main network, round-trip delay is introduced when data is transported from end devices to the data center.
When it comes to problems like mobility support, geographical dispersion, and context awareness, cloud-based solutions have fallen short, too.
In light of these limitations, this computing system has emerged as a practical and effective option.
The impact of cloud platforms has already been seen in the market, and they have permeated every industry. Since adopting a Cloud-based business solution may have a significant impact on operating costs, every company is on the lookout for one.
Cloud-based systems will get a fresh lease of life thanks to Fog Computing architecture in Iot, which will allow the Cloud service to expand into new markets.
Some companies are unable to send private information via the Internet due to business policy and regulatory prohibitions. This highlights the need of adopting Fog Computing examples.
Fog computing works in tandem with a cloud by allowing for the processing of time-sensitive data close to the network’s periphery.
By processing data closer to its point of origin, we may reduce the load on the central processing unit and save time.
Technology in the “Fog” will eventually replace “Cloud” systems. It will allow Cloud platforms to expand into hitherto inaccessible areas.
Final Thoughts
Fog architecture in IoT is an excellent answer to the issue of data processing in IoT. It uses edge devices more potent than end devices and closer to these devices than the more robust cloud resources to decrease application latency.
In this chapter, we presented a reference architecture for the Internet of Things and spoke about research and development now underway in academia and business to make the fog-computing vision a reality.
However, security, resource conservation, and energy efficiency remain challenges. Other areas for future study that will make fog computing more appealing to end users include open protocols and architectures.
Instead of transferring data to and from a central server, processing occurs closer to the input. That is why fog computing is faster and more responsive than traditional cloud computing.
Edge computing may help devices acquire results faster by evaluating device data in real-time. Fog computing stores just the most important data from the device’s massive data set by sending only the filtered data to the cloud.
Fog technology will power future computers and networks. Researchers are studying fog/edge computing and networking to maximize virtualized, pooled, and shared resources.
