What is edge computing?

The origins of edge computing can be traced back to the 1990s, when Akamai introduced the content delivery network (CDN), which set up transmission nodes near end users.These nodes can store cached static content, such as images and video.Edge computing further understands this concept by allowing nodes to perform basic computing tasks.In 1997, computer scientist Brian Noble demonstrated how mobile technology could use edge computing for speech recognition, and two years later it was used to extend the life of cell phones' batteries.At the time, the process was known as "cyber foraging," and it's how apple's Siri and Google's voice recognition work.

Cloud computing is the centralization of computing services, leveraging Shared data center infrastructure and economies of scale to reduce costs in the simplest form.However, due to the number of router hops, packet delays brought by the introduction of virtualization technology or server delays in the data center are the key issues of cloud computing migration. The concept of edge computing has become increasingly hot in recent years, and edge computing is still the innovation driving force in OpenStack.

Edge computing is a decentralized computing architecture that transfers the computing of applications, data and services from the central node of the network to the edge node of the network logic.Edge computing decomposing large services that were originally handled entirely by the central node into smaller and more manageable parts and distributing them to the edge node for processing.The edge node is closer to the user terminal device, which can speed up the data processing and transmission speed and reduce the delay.

Edge computing covers a wide range of technologies, including point-to-point, grid computing, fog computing, blockchain and content delivery network (CDN). Edge computing is very popular in the field of mobile and is now almost everywhere.

The relationship between edge computing and cloud computing

In many cases, edge computing and cloud computing are symbiotic.With the development and application of Internet of things, virtual reality, augmented reality and other technologies, there will be a data explosion in the future.Relying entirely on cloud computing for data transmission and processing will cause huge network delays. Edge computing can effectively reduce data transmission and processing by processing data in edge nodes, but remote storage through cloud computing is still crucial.
Cloud computing carries the high expectations of the industry, the industry has generally believed that the future of computing functions will be entirely in the cloud.However, with the growth of access devices, the bandwidth is increasingly limited when transmitting data and acquiring information. Especially with the development of the Internet of things, cloud computing is unable to meet the demand in response to the growing tide of connected devices and smart devices.With distributed edge computing, data transmission between different devices through intelligent routing and other equipment and technologies can effectively reduce the network traffic and reduce the load of the data center.

A brief history of edge computing

The origins of edge computing can be traced back to the 1990s, when Akamai introduced the content delivery network (CDN), which set up transmission nodes near end users.These nodes can store cached static content, such as images and video.Edge computing further understands this concept by allowing nodes to perform basic computing tasks.In 1997, computer scientist Brian Noble demonstrated how mobile technology could use edge computing for speech recognition, and two years later it was used to extend the life of cell phones' batteries.At the time, the process was known as "cyber foraging," and it's how apple's Siri and Google's voice recognition work.

Peer to peer computing came into being in 1999. With the release of EC2 service by amazon in 2006, cloud computing officially came into being. Since then, many large-scale enterprises have adopted cloud computing.The virtual-machine-based Cloudlets case, released in 2009 as a summary of mobile computing, details the end-to-end relationship between latency and cloud computing.This paper proposes the concept of two-level architecture: the first level is cloud computing infrastructure, and the second level is cloudlet composed of distributed cloud elements.This is the theoretical basis for many aspects of modern edge computing. In 2012, cisco launched "fog computing", a distributed cloud computing infrastructure designed to improve the scalability of the Internet of things.
There are many concepts of fog computing that we understand as Edge computing at present, including pure distributed systems, such as blockchain, point-to-point or hybrid systems, among which the typical ones are AWS Lambda @edge, Greengrass and Microsoft Azure IoT Edge. Edge computing has become the key technology to promote the adoption of the Internet of things.
 
Scalability and elasticity of edge computing

The distributed architecture of edge computing means that with reduced latency, it can improve resilience, reduce network load, and make scalability easier.

The data processing of edge calculation has already started from the data source. Once the data processing is completed, only the data that needs further analysis needs to be sent.This greatly reduces networking requirements and bottlenecks for centralized services, and users can avoid interruptions and improve system resilience for other edge locations or the potential to cache data on the device.This reduces the need to extend centralized services because they require relatively little traffic to process, resulting in cost savings, reduced device complexity, and reduced administration.

The future of edge computing

How will edge computing evolve?As more end users use edge computing to improve performance and functionality, we will see an explosion in edge computing.Edge computing accelerates data stream generation, including real-time data processing without delay.Smart applications and devices can reduce latency by responding instantly to data as it is created.This is crucial to the development of technologies such as self-driving cars and businesses.

Edge computing can efficiently process large amounts of data in place near the source, reducing Internet bandwidth usage. Costs are reduced while ensuring efficient use of remote applications.In addition, users do not need to transfer data to the public cloud to process data, thus improving the security of sensitive data.

Edge computing can not only solve the problem of automation of networked devices, but also reduce the requirement of data transmission, which can eliminate the bottleneck of data storage and data transmission on the basis of cloud computing.In the future, with the rapid development of Internet of things and other technologies, edge computing, as its key technology, will also achieve great success.


Compiled by 
ATEL RC Dept.

 

Mobile Web Site:

Advancing Communication Worldwide