Edge computing is the near-term future of advanced data collection and processing. It starts with the simple premise that data should be analyzed and converted to actionable information as close to the data collection point as possible. Edge computing is vital to the continued successful development of Internet of Things (IoT) networks as millions of complex IoT devices are deployed to collect billions of data points continuously. This massive increase in transmitted data threatens to jam transmission networks despite the eventual large-scale integration of 5G-based systems.
Advantages of Edge Computing
The advantages of edge computing include:
- Lower data rates—Pre-processing of data can reduce data redundancy and the transmission of data that does not provide value.
- Network efficiencies—Edge devices support more sophisticated networks on the Edge, making event-brokering possible and creating better resilience.
- Faster response—Edge processes can identify potential issues even before an event occurs and develop a course of action and a response without the latency involved in transmitting data to the cloud and waiting for a reply.
- Better User Experience—Users in remote locations can have a greatly enhanced experience when edge processes detect specific behaviors and provide desirable alterations to the environment.
Considerations for Deployment
There are several considerations that should be taken into account when looking to utilize edge computing:
- Powering Edge Devices—In remote areas, electric power may not be in abundance. The device may need to rely on consistent battery storage, possibly supplemented by alternative power such as solar energy. Transmitting wirelessly is a power-consuming process, making it even more critical that the device send only what is necessary, such as discrete event data.
- Network and storage—IoT data is inherently time-dependent. Many traditional data networks and database storage systems are not designed to deal with this type of information. Will your system be able to handle time-series data?
- Physical security—Because these devices will be deployed in remote locations, they could be vulnerable to physical attack. Bringing compute power down to the device means that the information available will be of higher quality and higher value.
- Cyber security—IoT creates more attack surfaces, and deploying any IoT approach, including Edge Computing requires attention to securing the data.
- Legacy Infrastructure—It may be difficult to integrate these new systems with legacy “dumb” sensors and data transmission networks. In many cases a complete replacement will be necessary, obsoleting a substantial amount of useful and amortized equipment.
To realize the promise of 5G enabled IoT networks and Edge Computing, systems will have to be designed to address specific use cases; smart devices attached to existing infrastructure will not be sufficient. These new systems will require security and privacy to be built-in and will feature second-level event-brokering that will help realize the efficiencies that edge computing promises. Architecting and building these new IoT networks is a challenge that will need to be met if IoT and edge computing are to succeed.
