Preface – This post is part of the Edge Computing series.

Introduction

It’s a time when we can say that Edge computing will replace Cloud computing in the future. Several organizations are adopting edge computing due to its solution power and feature benefits. After cloud computing is rocking the IT world, now is the time when edge computing is picking up pace. So the understanding starts from knowing about edge computing. Edge computing is very clear from its name, which says computing that takes place at the edge of internet networks. The edge in edge computing is where end devices access the rest of the networks supporting intelligent devices like phones, laptops, industrial robots, and sensors. The edge used to be where these IoT devices connected so they could transmit data to and receive instructions from a distributed data center and Cloud.

What is Computation Offloading?

In digital technology, offloading is meant to transfer data from one device to another digital device. Computation offloading transfers resource-intensive computational tasks to an alternate processor or an external platform like a grid or Cloud with the objective of improvising computation with low latency while saving energy.

The offloading Work

Let’s understand offloading term with a simple example,

Everyone has a smartphone which is a kind of miniature computer, and you can do a lot of things with it. But users do not use all the resources of smartphones. Now in this new era where we have tiny computers such as phones, smartwatches, and much more all around, and these all are connected using technology such as wifi, Bluetooth, etc.

If you look at smartphones alone, on average, a person actively uses a smartphone for about five hours a day; this means that competing resources are being wasted for the rest of 19 hours a day. There will be around two billion+ smartphones worldwide by 2017, and 1 million could provide up to a petaflop of computing power. About 2000 pedo flops of computing power are available worldwide for 19 hours daily. If we collect the whole data and calculate the resource wastes in numerical form, then we will get a result of a high wastage number.

The point is there are plenty of pile resources on small devices, which can be scaled up for better efficiency. Scaling the resources can provide a lot of computational power, and some entities can then be used with a lot of work to do, reducing tremendous jobs in weak mini devices.

Benefits

• More efficient power management
• Fewer storage requirements
• Higher application performance.
• Less latency and high customer satisfaction.
• High efficiency and increased customer satisfaction.

Conclusion

Edge computing is arising as one of the strategic technology that will redefine the future computing paradigm for its promise of lower latency, less bandwidth usage, and data privacy protection. Computation offloading is critical to make the promise reality in various applications, from connected autonomous vehicles to smart homes. Multiple aspects of computation offloading, including energy consumption minimization, Quality of Services guarantee, and Quality of Experiences enhancement, are surveyed. Moreover, resource allocation approaches, gaming theory, and heuristics-based computation offloading optimization of system performance and overheads for computation offloading decision-making are also analyzed.