The IoT( Internet of Things) is a real event. It is a real part of the cloud computing scenario. The key challenge faced by the users is the way of achieving data processing from the myriad of devices that are available. As per the predictions of Cisco Systems, the cloud traffic might observe a rise to nearly four-folds by the time of 2020. It is likely to increase at a rate of 3.9 ZB (zettabytes) per year. This figure might reach up to 14.1 ZB by the year 2020.
As a result of these observations, the users can have the perfect storm of cloud computing from the growth and expansion of the IoT (Internet of Things). In the end, IoT is all about the processing of the data that is device-generated and yet tends to be meaningful. At the same time, cloud computing is all about the utilization of the data from the centralized storage and computing. It is assumed that the rapid growth rates of both cloud computing and the Internet of Things would eventually become unmanageable.
So owing to the risk ahead, the question comes up “what should be done?”. The answer to this question is quite simple –“Edge Computing”. There is no denying the fact that edge computing tends to push much of the date processing work out to the edges of the particular network. This phenomenon tends to close the possible sources of data, and then comes the point of bringing about the division of the data processing between the centralized system and the edge. This implies the fact that a public cloud system like Microsoft Azure, AWS (Amazon Web Services) or Google Cloud would achieve the same.
The overall concept of edge computing might appears to be similar to that of the client-server architecture. The processing of edge computing is also concerned with the decision about that processes to be carried out at the client as well as at the server end. For any highly distributed applications like that of the IoT (Internet of Things), the users are essentially provided with a client-network along with edge-architecture going on simultaneously. In case, the devices of the users are not able to do any processing by themselves, then a network edge-server architecture is applied.
The main goal with edge computing is to achieve the processing of the data near the device that the system would be needing quickly, such that proper action is taken care of instantly. There are several use cases in which the reaction time is the significant value point of the entire IoT system. These use cases are also used for sending the data back to a centralized cloud in a consistent manner such that it would prevent the particular value from happening.
The users can still make use of the cloud computing even after the advent of the edge computing. The users can use the cloud for the purpose of data processing that is either not time-sensitive or is not needed by the particular device like that for the data analytics on all the available data from the various devices.
There is another dimension to this approach. Both the cloud computing as well as edge computing are two different concepts. There is no point one replacing the other in any form. It is no more true than claiming that PCs would be replacing the data-centers.
It would make a perfect sense for the creation of a purpose-built application that is based on edge computing. The application could be like the one that places the data processing activities in a sensor such that it quickly processes the reactions to alarms. However, if you are not going to place the inventory-control data along with your applications at the hands of edge computing, then moving all of the computing to the edge would lead to an undistributed, unmanaged and insecure mess.
All the public cloud providers have the functioning of IoT technology stacks and strategies that aim to include the concepts of edge computing. Both the cloud, as well as edge computing, can function together in a coordinated manner. However, edge computing is specifically meant for the purpose-built applications and systems that would need specific needs and requirements. On the other hand, cloud computing is more of a general-purpose platform that can function well with any other purpose-built infrastructures or systems with the old client-server model.