Edge computing

Edge computing is a form of computing for optimizing the performance of web applications and internet-connected devices by completing computational processes closer to the source of data. Edge computing reduces bandwidth consumption and latency by reducing the distance of communication between clients and servers. Edge computing acts to decentralize the processes of cloud computing centers because it changes the geographic distribution of where computation is taking place.

Edge computing is often discussed in relation to IoT. Moving computing services closer to the source of data, such as an IoT device, offers a range of benefits including rapidly analyzing real-time data, ideal for IoT sensors and devices.

A major advantage of edge computing for users is the reduction in latency times compared to cloud computing. Edge computing allows users to gather the information they need to run their code on the internet, typically using a web browser, by getting the information stored on an edge computing network that is closer to them than cloud computing and other related internet infrastructure. The reduction in latency offers consumers the benefit of accessing what they want faster.

With the rise of smart devices and more data being uploaded to the cloud a significant burden is being placed on bandwidth capacity. Edge computing offers the ability to process information locally reducing bandwidth requirements.

In theory, less data being transferred over networks reduces security needs, plus decentralizing data storage reduces the potential for a signal point of failure. Edge computing could also improve user privacy as less data is uploaded to the cloud for companies to track.

Edge computing allows users to scale their own IoT network without having to pay for costly cloud computing storage.

With edge computing, users are not dependent on having a reliable internet connection. Plus storing data locally in microdata centers ensures a more reliable connection for IoT devices. Edge computing is especially beneficial for users in remote locations with slow and unreliable internet connections.

The beginning of decentralized computing goes back to the 1990s, when Akamai launched its content delivery network (CDN). The company introduced nodes at geographical locations closer to the end-user to better deliver cached content such as images and videos.

Emerging in 2006 with the release of Amazon's EC2 service, cloud computing has gone on to widespread adoption.

In 2009, Satyanarayanan et al. published the paper “The case for VM-based cloudlets in mobile computing” discussing the end-to-end relationship between latency and cloud computing. The paper proposed a two-level architecture consisting of the unmodified cloud infrastructure (high latency) and a dispersed element known as cloudlets (lower latency). This became the theoretical basis for many aspects that would go on to become modern edge computing.

In 2012 Cisco introduced the term fog computing to describe dispersed cloud infrastructures with the aim of promoting IoT scalability by facilitating real-time low-latency applications.

Fog computing, also referred to as fog networking or fogging, is a closely related technology to edge computing. It describes a decentralized computing structure located between the cloud and the devices that produce data. This structure is flexible allowing users to place resources, including applications and the data they produce, in locations that enhance performance.

While fog computing shares similarities with edge computing, it is not taking place directly where data is generated, at the "edge" of the application's network. Fog computing is a layer between the cloud and edge computing. Instead of edge computing sending data directly to the cloud, it can instead go to a fog computing layer where relevant data is passed on, irrelevant data is deleted, or analyzed for remote access, or informing local learning models.

Demonstration of the cloud, fog computing, and edge computing.

One can think of fog computing as a mediator for various purposes between the edge and the cloud. Fog computing does not replace edge computing and edge computing can operate without fog computing.

With a range of benefits, edge computing has numerous use cases across industries:

• Autonomous vehicles
• Remotely monitoring assets for the oil and gas industry
• Smart grids
• In-hospital patient monitoring
• 5G virtualized radio networks (vRAN)
• Cloud gaming
• Traffic management
• Smart homes