5G, edge computing and IoT poised to retool networking
The combination of 5G and edge computing could introduce several changes for network operations. Find out why the network infrastructure -- along with fog computing -- is so important.
5G offers wireless cellular connectivity with much higher bandwidth, lower latency and greater device density than previous generations of cellular service. In addition to speeds up to 10 Gbps and latency of 1 ms, 5G will support thousands of devices per square kilometer.
5G also supports network slicing, which enables carriers to divide the 5G radio access network into virtual segments, each of which can be customized to support different types of applications. Network slicing has the added benefit of securing one segment's traffic from that of other segments.
5G uses a variety of frequencies, ranging from existing 2.4 GHz bands to millimeter wave (MM wave). The MM wave frequencies support higher bandwidth channels, albeit with reduced range due to radio frequency absorption. This means more access points will be deployed, which will reduce the number of endpoints that associate with each AP and increase the bandwidth available to each endpoint.
5G edge computing, fog computing and IoT
Edge computing integrates data processing into edge devices, which are often data collectors or process controllers. Edge computing processes raw sensor data quickly, eliminating the need to transport data to the host application. This reduces both transmission latency and bandwidth costs by processing raw data close to or at the edge device.
A good example is the programmable logic controller, which collects and processes locally generated data. Summary data is then typically uploaded over a network connection to a comprehensive monitoring application that performs additional data processing and archiving.
Edge computing improves the efficiency of the upstream data transfer and provides real-time control of IoT edge devices.
Edge computing improves the efficiency of the upstream data transfer and provides real-time control of IoT edge devices. In edge computing, the processing power is installed within, or close to, the edge device.
Fog computing offers another alternative as it places compute and storage capabilities near, but not within, the edge devices. It provides local device-to-device communications, offers greater control system resiliency and sends summarized data to cloud-based applications.
The 5G edge computing combination creates opportunities for new and improved applications -- among them IoT deployments -- thanks to the standard's bandwidth latency and real-time control capabilities.
Security and improved battery life are important 5G use cases related to IoT and edge computing.
5G's latency of 1 ms -- versus 10 ms for 4G -- supports real-time applications that could not tolerate 4G's latency. A common example is autonomous cars, where vehicles communicate with each other over 5G, sharing sensor data and driving intentions that enable each car to make informed decisions regarding its intended path.
Fog computing, combined with 5G's low latency, will support real-time applications that cannot be supported by high-latency cloud-based applications. Application resilience can also be improved by using distributed architectures in which the fog computing infrastructure is installed near IoT devices.
With the right architecture, applications based on edge and fog computing can continue to function at a local level even when network connectivity to the cloud has failed.
Security use cases underpin 5G edge computing
Edge computing security is enhanced by using private 5G network slices that are protected by firewalls, built on network functions virtualization infrastructure. Containerization, meanwhile, will make it easier to deploy custom software to edge or fog compute systems. Large, solid-state storage systems will enable the storage of significant amounts of data. The combination of networking, compute and storage opens the door to a lot of interesting and powerful systems.
Consider a building's environmental, lighting and security system. With 5G and a fog computing infrastructure in place, low-cost sensors transmit data via a local, secure network. The fog computing system makes its control decisions based on configuration data downloaded from the cloud. The low latency and high density of the 5G network mean it can collect data in real time from hundreds of thermostats, room occupancy sensors, security scanners and ambient light monitors.
The fog computing system sends commands via the 5G network to building lighting, door locks, and heating and cooling systems. If there is an interruption in cloud connectivity, the system continues to operate. Security is fortified because the 5G network slice segregates the sensor and control data from other network users.
An assembly plant control system is another good example. Robots communicate with each other to pass parts among themselves, relying on 5G edge computing to communicate. A fog-based infrastructure would add another level of control between adjacent robots. You can easily extend this concept to a chemical plant process control system. 5G's network slicing would be used in both examples to provide increased security.
