In this post, we will first check out the basic architecture of edge computing and after that will do the comparison between edge and cloud computing in detail.
Edge Computing Architecture
In this segment of the paper, we will discuss a general architecture of the edge computing that consist of three layers: lowest layer is user devices, middle layer is edge node and upper layer is cloud server layer.
The architecture of edge computing is categorized into two segments. In first category, user devices (such as vehicles, mobile phones, laptops and sensors) having some computational power. These devices has a capability to preprocess raw data and send that data to cloud. On the other hand, if computational power exceeds and user devices are unable to handle big task then these devices will pass their data to adjacent edge node. In second category, user devices are able to offload their data to adjacent edge node instead of cloud. Furthermore, edge nodes are able to do preprocessing such as data fusion and data comparison. After processing, data will go to cloud and at this stage cloud will do final analysis. The architecture of edge computing is explained as follows:
- User Devices Layer
First layer is called as user devices layer. User devices are those devices having lower computational power and deployed at device layer to perform task in real world such as controlling and sensing. Vehicles, laptops, sensors and smart phones are the best examples of user device layer. Raw data that produced by them sent to upper layer for further processing.
- Edge Nodes Layer
The second layer is called edge nodes layer. In this layer, there are devices used to increase computational power like servers, bas stations and switches. Edge node layer can perform multiple tasks such as transforming data from cloud to users. These devices are also capable to offload some of their computational work to cloud for the sake of reducing burden. However, edge nodes can collaborate with each other to provide un-interrupted service to users.
- Cloud Server Layer
Third layer is known as cloud server layer. It is responsible for high level security, authentication, authorization, and computation and processing. Cloud server either interacts with edge node to receive data or directly with user devices for further processing. When data analytics are very large (such as country wide), then there is possibility that edge node send data to cloud for analysis.
Comparison of Edge and Cloud Computing
In edge computing there is an additional layer that is called edge node layer which makes the edge technology differed from earlier cloud computing technology. Let’s see the detail comparison of both these technologies.
i) Location Awareness:
It is an ability to track the user device or to find the geographical location of the device. It is one of the characteristic of edge computing technology whereas cloud computing does not offer this feature.
ii) Geographical Position:
In cloud computing, cloud servers are at fixed positions due to centralization and process the data from fixed positions whereas edge servers are placed at various places to process data such as on highways, super markets etc.
In cloud computing, latency rate is high because of the placement of fixed servers on specific locations that creates a long distance between IoT devices and cloud servers. This scenario is bearable for non-real time applications. If we talk for real-time applications then high latency rate must be fixed. To reduce the latency rate, edge computing is introduced and it is using for real-time application scenarios.
iv) Large Scale IoT Application Support:
Edge computing is offering a feature of supporting to large scale IoT application whereas cloud computing cannot offer this feature due to the heavy management and computational overhead.
v) Network Architecture:
In edge computing technology, there is a decentralized server architecture to process computation and data storage resources. On the other hand, cloud computing is based on centralized architecture.
In edge computing, there are heterogeneous devices are used as a hardware whereas in cloud computing technology, general devices are used.
Edge computing supports the feature of mobility whereas cloud computing does not.