Fog Computing

Chaitanya Khandave
5 min readSep 15, 2021

--

Introduction

Lets start with cloud computing, cloud computing is the delivery of computing services — including servers, storage, databases, networking, software, analytics, and intelligence — over the Internet (“the cloud”) to offer faster innovation, flexible resources, and economies of scale. You typically pay only for cloud services you use, helping lower your operating costs, run your infrastructure more efficiently and scale as your business needs change.

Fig 1: Cloud Computing.

Fog computing extends the cloud computing paradigm to the edge of the network, thus enabling a new breed of services and applications. The major characteristics of a Fog Computing environment are

i) Low Latency.

ii)Widespread Geographical distance.

iii)Mobility.

iv)Very large number of nodes.

Fog Computing allows greater support and better response time to the Internet of things environment where we may have many wireless sensor nodes uploading a stream of data to the cloud. It is a highly virtualized platform that provides computation, storage, and networking services between end devices and traditional cloud servers.

Fig 2: Fog Computing.

DRAWBACKS OF THE TRADITIONAL CLOUD ARCHITECTURE

In traditional cloud architecture, the infrastructure is provided in data centers because it is cheaper. In an IOT architecture that uses the cloud, the nodes that are part of the IOT upload their operational data to the cloud. The cloud then does the processing and takes the appropriate action. The nodes are located in a large network distance from the cloud data centers, the decision-making process can be one enormous latency. This enormous latency makes the cloud unusable for time-critical applications / real-time systems. In addition, the latency increases as the number of devices that connect to the cloud increases.

BUILDING A FOG COMPUTING ENVIRONMENT

A fog computing environment can be achieved by allowing a kind of app store system on the network devices, in which the user can choose which data should be processed at the edge and which data should be processed in the cloud. This includes marking packets that are being processed on the network. When the network device receives this packet, it decodes the information contained in the packet and sends it to the appropriate application running on your system for processing. As soon as the processing results are available, the network device creates a packet with a source IP address like that of the node that sent the marked packet and the destination address as the cloud.

Fig 3: Fog Computing Setup.

Fig 3 shows a configuration in which many wireless sensor nodes continue to upload their daily cloud operations a node that only wants to send an average of its operating data to the cloud marks its packets, these packets are processed by the intermediate wireless router, which loads the wireless router after processing the data up in the cloud. Untagged packets are sent directly to the cloud without intermediate processing.

ISSUES

1. Fog Networking: Because it is on the edge of the internet, the fog network is heterogeneous. The role of the fog network is to connect all the components of the network. However, managing such a network, maintaining connectivity and providing services on it, especially in large Internet of Things (IoT) scenarios, is not easy. New techniques such as software-defined networking (SDN) and network functions virtualization (NFV) are proposed to create a flexible and easy-to-maintain network environment.

2. Quality of Service (QoS):

· Connectivity: The end-user choice of fog nodes will have a huge impact on performance. We can dynamically select a subset of fog nodes as relay nodes to achieve optimization goals of maximum availability of fog services for a given area or individual user, with constraints such as delays, performance, connectivity and power consumption.

· Reliability: Typically, regular checkpoints can improve reliability to recover from a failure, rescheduling failed tasks, or replication to take advantage of parallel execution can improve reliability, but checkpoints and rescheduling may not fit the Highly Dynamic Computing environment in Fog because of latency and inconsistency Environment can adapt changes.

· Capacity: The capacity consists of two parts: 1) network bandwidth, 2) storage capacity In order to achieve high bandwidth and efficient use of storage, it is important to study how the data is placed in the fog network, since the location of the data is very important for computation important.

· Delay: Latency-sensitive applications such as stream mining or complex event processing are typical applications that require fog computing to provide real-time stream processing rather than batch processing.

3. Privacy: There is always concern about data protection when many networks are involved. Because Fog Computing is based on wireless technology, there are major concerns about the privacy of the network. There are so many fog nodes that they can be accessed by any end user and therefore more sensitive information. switches from end users to fog nodes.

4. Security: Fog computing security issues arise because many devices are connected to the fog nodes and on different gateways, each device has a different IP address and any hacker can spoof your IP address to gain access to your personal information, which are stored in this particular fog are nodes.

5. Authentication: Authentication is one of the most troubling problems in fog computing because of the widespread availability of these services.

Conclusion

Fog computing has developed as an extension of the cloud, where it is closer to IoT elements, in which data has been stored at the cloud and Fog nodes. The incorporation of cloud computing and Fog computing in various IoT implementations would offer several advantages to users. Although, currently there are issue to take into consideration as discussed above like networking, quality of service, privacy and security. Probably, organizations would come up with solutions in the future.

References

--

--

No responses yet