followers 0 popularity
0
following 1

following  view all

Groups
macandrew is not in any groups
macandrew

Edge Computing

Feb 10th 2020 at 3:56 AM

 

Introduction

In recent years we have seen a great rise in cloud native computing. Many computers use cloud computing to offload its heavy tasks and get the results directly. In the early days of cloud computing only computers and other end user hardware were connected to the cloud. Only these devices used the facilities of cloud computing.

Gradually as other smart devices came on the rise, they too started using cloud computing. These smart devices were smart air conditioner, smart watches, smart phones etc. They too started taking advantage of the cloud to offload their heavy work. These devices collectively are called IoT (Internet of Things) devices. IoT is basically collection of smart devices and sensors that are connected to some cloud that processes their request and returns some response.

As more and more intuitive IoT devices came into existence, the problem of bandwidth arose. Security cameras that were connected to cloud were constantly required to send footage to be analyzed for image recognition.

The problem was heavy data transfer that caused other devices on the same network to slow down. This was a big problem as the security cameras required constant sending and receiving of information between the cloud.

So either you had to have a very high speed internet connection, or have a separate internet connection especially for transfer of data for the cameras. This is just one example of the drawbacks of cloud computing.

Let’s look at another example of limitations of cloud computing before diving into the real topic at hand. It is necessary to understand the limitations so we can truly appreciate the beauty of its solution. Take autonomous vehicles as an example. These self driving cars rely on sending and receiving data between the cloud. They send real time information and the network speed should be very fast to deliver the data sent by the car.

The cloud also needs to be very fast to calculate the request of the car and send a valid response to the car. This works fine but what happens if it starts to rain and obstructs the network connectivity, or slows down the connectivity? If you are on a highway and suddenly lose connection to the cloud, what happens then?

Well a solution is proposed and is called Edge Computing. What exactly is edge computing and how does it overcome the drawbacks of cloud computing exactly? We’ll dive deep but first let’s see what Edge Computing basically is.

What is Edge Computing?

Gartner defines edge computing as “a part of a distributed computing topology in which information processing is located close to the edge – where things and people produce or consume that information.

Don’t worry if you don’t understand yet. Let’s elaborate the definition. Edge Computing basically is a networking philosophy, which earns to bring computing and processing of data, as close to the source of data as possible.

Meaning instead of sending data to a remote location for processing, the Edge Computing brings the computational factors and elements as close to the source of data as possible. In a nutshell, edge computing means running as few possible processes in the cloud, and moving those processes to local place, such as user’s device or on an edge server.

Why would we even want this? Well following factors will make it more evident to why bringing the computation close to the source of data is beneficial.

Reduced Latency:

The first benefit of edge computing over traditional cloud computing is the drastically improved latency. With cloud computing, latency was a major drawback which sometimes caused a problem in the system.

Some systems required a response rate of very few milliseconds which was not possible as data had to be sent to the cloud, then processed and then sent back over the same network. This took time and as a result, increased latency. With edge computing, the computation is right next to the data generation.

This means as the data is generated, it is also processed simultaneously with very little latency. As the computation is embedded in the system or placed very close to the system, latency becomes an issue of the past. The data is processed within milliseconds by the edge computer.

Bandwidth Use

Another benefit of edge computing is the saving of data. The bandwidth usage greatly decreases as very little data is required to be sent to the cloud for computation purposes and majority of the processing is done locally. The processing is done locally on user’s device itself or on the edge server which is placed closely to the system generating data.

This allows data to be generated and processed simultaneously without using a lot of bandwidth. Like we previously talked about the security cameras requiring to send data constantly to the cloud. With edge computing you can process that data within the camera itself. As a result you don’t need to send a lot of data and all the processing is easily done on the camera itself.

Similarly like we talked previously about autonomous vehicles requiring an internet connection to send analytics to the cloud. With edge computing, all the processing can be done locally without the hassle of sending large amounts of data over the network. The processing can be done locally, and results can be generated locally. As a result, the decision is also taken locally by the car. This greatly reduces the bandwidth usage. This is another reason why edge computing covers the limitations of cloud computing.

Network Edge

Network edge is where the device, or the local network containing the device, communicates with the internet. The network edge is a bit blurry term. An example is the user’s computer or the processor of an IoT device can be considered a network edge. The main point is that the edge of the network is very close to the device or system unlike cloud servers which are usually very far away from the devices they communicate with.

Example of Edge Computing in Action

Consider mobile manufacturing unit that uses edge computing technology. Many mobile phones are manufactured per hour, are packed and sent out to be sold. Occasionally there are faulty mobile phone. They need to be identified and removed, before they are packed and shipped away. As faulty phones will cause a bad reputation for the company.

What the company has done is implied edge hardware to each of its monitoring cameras. What the camera does, is constantly record footage of mobile phones rolling out and at the same time analyzing the footage. The edge hardware is directly connected to the camera which mean all the processing is happening in real time. As soon as the camera identifies a faulty phone, it send that data to the cloud server and the mobile phone is removed immediately.

The benefit of this implementation is that the cameras need not send all the footage to the cloud server but instead they analyze it locally. This reduces latency and allows to identify faulty mobiles immediately. Another advantage is that it save bandwidth usage as only footage of faulty mobiles need to be sent to the cloud to be saved. This is just one of many examples of how edge computing can be used.

Drawbacks of Edge Computing

Like every technology in the world that has tremendous benefits, it also comes with some drawbacks and security threats. Similarly edge computing also has its own drawbacks and security threats.

When data at the edge is being handled by different devices, it may not be secure as compared to cloud storage. It can also increase the chance of malicious attacks. With increasing IoT devices that have robust built in computers, the opportunity for malicious attack on these devices increases greatly.

One solution is to use encrypted data on edge hardware. Another is to use VPN tunneling for access control methods. With some tweaks and precautions, you can make your data safe on edge computing.

How does 5G play its role?

With many countries staring to deploy their 5G technology, that is sure to offer 10 times high bandwidth as compared to traditional 4G and very minute latency rates. It’s started to roll out in some countries and is bound to be deployed in the rest of the countries as well.

With such high data transfer speeds you would expect the companies to just resort to traditional cloud computing right? Well you’re wrong. 5G can be considered as a booster to the already evolving edge computing. Many carriers are incorporating edge computing strategies into their 5G deployments for faster real time processing of data for connected cars and autonomous vehicles.

Conclusion

The edge is emerging in many places and is bound to expand with enterprise organizations leading the way, delivering the IT infrastructure to support it. It’s evolving at a very high rate with people still mind boggled to define what it is and where it resides. Will edge be a bigger hit than cloud computing, well on the future holds the answer!

Visit our site BYTES FUTURE

1 comments
Please to comment
Feb 10th 2020 at 3:57 AM by macandrew
Bytes Future is providing exceptional services for marketing & advertising for all the industries including, Medical & Healthcare, Financial Institutions & Banks, Restaurants, Hotels, Automotive, Industrial, Government, Technology, Telecom, Oil & Gas, Entertainment, Events, Exhibitions & Conferences.
   

sign in

Username
Password
Remember Me


New to IM faceplate? join free!

Lost Password? click here