Edge computing refers to the process of allocating and getting data to the correct place at the right time. This is done by location computing and storage resources at the edge of a network and moving around the data. Edge computing ensures decentralization of the data process and avoids non-essential data transmissions.
There are many definitions of edge computing, however, the one shown below is one the Technology Exchange believes fits the services that we provide:
“Cloud-like capabilities located at the infrastructure edge, including from the user perspective access to elastically-allocated compute, data storage and network resources. Often operated as a seamless extension of a centralized public or private cloud, constructed from micro-data centres deployed at the infrastructure edge”.
Typical network topology for an edge computing implementation involved three major layers:
- Internet of Thing devices: IoT devices are usually connected to an edge device, whereby they can communicate through diverse communicate protocols with edge computing acting as a data source
- Edge nodes: use close to real-time data analytics to process data. It enables diverse communication routes for data acquisitions and act as a temporary data storage for IoT devices.
- Cloud services: can create management facilities for edge computing, as well as IoT devices, while performing as a long-term data storage and analytics. They also act as a point of integration for other systems.
Edge computing comes with many advantages:
- Privacy which can keep data from being sent and stored on cloud systems.
- Connectivity that is a cost-effective method as you do not stream or send data to the cloud.
- Advances to responsiveness and reliability happen when maximizing activity at the edge so there is less dependency on the cloud
- Wide physical distributing because the data on edge computing are decentralized and sent by nature.
- There is a need for real-time interactions as edge computing aims to enable IoT assets by bringing data closer to the data sources. This, therefore, requires an immediate response time.
- Autonomy in edge computing is subject to data center boundaries.
- Heterogeneity is present in both the variety of devices that connect to edge computing programs and in the diversity of networks and security protocols they are exposed to. Interoperability and federation because typically edge installations work together with a federal cloud offering. For this both edge and clous environments interoperate for operations such as edge device management for long-term data storage.