The promise of 5G—along with the speed and reliability it offers—is certainly nothing new. In fact, nearly all the major cellular service providers are delivering 5G, and device makers are offering compatible hardware to take advantage of those networks. However, the infrastructure for expanded 5G, particularly at the more challenging millimeter wave frequencies, is still coming together. The true coverage and speed of these networks are not fully online, even in major cities that have been developing these systems for years.
Today, we are at a point where wireless service providers are striving to differentiate themselves by demonstrating the intellectual chops of their systems. This refers to the reliability and compatibility with other networks and devices, and the actual “smartness” of their system that allows for deep connectivity with a network of devices and other systems. This is where edge computing comes into play.
Placing Computing Power Closer to the Data Source
Edge computing processes some of the most sensitive data and powers critical systems that must function reliably and safely. In other words, it refers to data processing at the “edge” of a network, or more simply, as close to the original data source as possible.
Placing computing power closer to the data source (routers, wide-area network devices, IoT devices, etc.) with edge computing ensures that the data has less distance to travel and more places to travel to versus a distant cloud infrastructure. This eases the strain on resources by processing data closer to the source and only sending “valuable” data to a remote data center.
Edge processing is particularly important for robots and automated machinery. For autonomous robotics that collect and process sensor data into outputs to actuators that control the robot’s movements and actions, this processing has virtually always taken place at the edge. However, latency from the robot to a central processor limited the use of central data processing in many use cases. Roboticists are using improvements in data size efficiency of sensors (such as lidar) that can process lidar data on-board into only the actionable data to increase the edge computing power of robots without increasing the size. This creates more capable robots that, among other advances, can safely operate closer to humans.
Realizing the Benefits of 5G with Edge Computing
Edge computing is critical to realizing all of the benefits of 5G in personal smart devices, in IoT infrastructure necessary for smart city capability, and in remote surgery, among many others. Self-driving vehicles are another example. Since the edge network is much closer to the vehicle, response time is decreased, rendering a much safer and more reliable vehicle.
When mature, communications services will need to leverage multi-modal networks with parallel fiber, terrestrial wireless, and satellite links. Optimizing which traffic is transmitted via the path will minimize latency and utilize all available bandwidth efficiently. These network management decisions—increasingly made by artificial intelligence (AI) within the network infrastructure—will be complemented by exceedingly intelligent cognitive radios within the connected device. This will allow devices to send data using the connectivity option that offers the best bandwidth and lowest latency, especially in critical applications such as connected battlespace.
Advanced devices are also moving toward increasing edge computing power for sensor data processing for immediate output to actuators and toward performance analytics that allow the device to run more efficiently. Currently, this type of analysis is considered a good candidate for transmission back to a central data warehouse for analysis by a person or AI. It is now becoming more prevalent for data to be evaluated at the edge to drive real-time performance improvement. In this way, the most advanced devices and systems will be able to determine which data should be processed at the edge and which should be transmitted to a data warehouse, continuously optimizing device performance.
Clearly, reaping the benefits of edge computing poses a great deal of complexity.
Benchmark Meets You at the Edge
Designing the systems and implementing the hardware and software into new or existing infrastructure requires integrating multiple advanced technologies and balancing trade-offs, a challenge being tackled by communications leaders worldwide. It’s no surprise that even large conglomerates are accomplishing these using partnerships rather than developing the extensive internal resources and expertise required to handle these requirements at the volume necessary to meet demand.
Fortunately, Benchmark has been involved in the 5G infrastructure and edge computing from the start through several key client relationships. In fact, leveraging our expertise with millimeter wave frequencies developed while working with the defense industry, our wireless connectivity experience even predates commercial 5G.
Benchmark is a unique partner that can come in at any point during the product development process for edge computing and help to take your designs to market, meet volume requirements, address complex technical challenges, and more. However, we provide the most value when we can collaborate at the early stages of R&D.
Our expertise allows us to help with not only the complex technology behind edge computing, but also the requirements for maintaining a powerful and efficient edge network. We can help our customers develop a deployment pipeline for getting applications to the edge. We can help to manage control of edge requirements. And we can do this all while speeding time to market, improving development efficiency, and managing costs.
Edge computing will usher in the next generation of 5G speeds, reliability, and connectivity, and Benchmark is the perfect partner to help get you there. Visit us today to learn more about our expertise in 5G and edge computing.