Edge deployment is a difficult decision for communications service providers (CSPs). Carlos Bravo and Jitendra Manocha will discuss the five main factors CSPs need to consider – from the edge computing infrastructure to the user plane and the opportunities and challenges ahead.
Paramedics can have limited resources and equipment when arriving at an accident site or emergency scene. Paramedics are trained to stabilize patients and transport them to the hospital quickly. On the other hand, Ambulances act as a taxi service that provides little equipment or medical treatment. The average travel time to the hospital is getting longer due to increased congestion and centralization. This can lead to lost time and even death. What if the ambulance was connected to the edge of the hospital?
Imagine a tiny device connected to a smartphone that could perform remote diagnostic tests like ultrasound. This would save precious time, allow hospitals to prepare the necessary preparations, and provide immediate treatment upon arrival. Imagine a specialist providing remote support to the hospital, giving instructions and interpretations of the images, and directing the probe. This use case requires a higher network and computing performance (latency, bandwidth) than can be achieved by a regional or central cloud.
This was the exact conversation we had with a medical device manufacturer a few years back, looking to improve ambulances’ capabilities. This type of use case is not only exciting, but it can also save lives. This is just one example of the 5G-enabled enterprise market that could be worth as much as USD 700 billion by 2030.
Cloud and edge — The technology that enables 5G to live up to its promises
Edge computing uses cloud capabilities to bring compute power and storage closer to where data is generated and consumed. It will depend on what the application requires. However, it is just as essential to locate the right location for your needs. This is a complete change from the recent trend towards centralization to lower costs and keep control.
Edge computing is not a new concept. Players, including primary media streaming services worldwide, already use distributed cloud and similar technologies. 5G brings a new level of network characteristics and, in turn, opens up a new world of possibilities. Without edge computing, 5G’s promise to customers and consumers is impossible. We often say, “Without edge computing, 5G is faster than 4G.”
How can CSPs provide the end-to-end capabilities that bring the edge and network together and create a position for themselves to make the most of these early opportunities? Here are five interdependent areas to consider when defining and deploying edge computing solutions.
The edge computing infrastructure layer is in compute storage and application hosting. It’s about bringing cloud computing to the edge. Edge infrastructure is not about installing your application on many static machines or servers. It’s different from other network infrastructures. It is about creating a method of managing things, similar to cloud capabilities but at the edge in a distributed environment.
Edge applications will require reliability, so the infrastructure needs to be flexible, efficient, and automated. The infrastructure can be either on-premises or in the CSP networks. It will host its telco workloads, and 3PP OTT apps with limited local management infrastructure must support multi-cloud and hybrid clouds in order to support multiple applications.
Resource distribution and configuration are key factors in edge orchestration. It is impossible to have thousands of sites running edge workloads in one country and all of the applications deployed at every site. This would be a waste of time and expense. The edge can be resource-constrained because it is smaller than a central location. It is therefore vital to map the topology and consider the capabilities of each site in the network. This will allow you to identify the most suitable location for your application and monitor it continuously for optimal usage.
This is called “Smart Workload Placement.” It uses algorithms to determine the best places for the most capabilities. Then, they find the sweet spot where the costs of deploying an app across multi-cloud infrastructure are offset by the benefits. It is crucial for applications running at the edge to ensure that information flows and data flow in the correct places. This is especially true when working in multi-cloud environments.
Plan for the user
Control and user plane function separation (CUPS) was a technology introduced in 4G. It has been improved with the introduction of 5G. The three core packet functions now include control mobility and session management. The control plane handles access mobility and session management functions and can be centralized. However, the user plane function acts as the gateway between the application and the network – the point at which the network meets the Internet.
It is essential to distribute the user plane at the edge. You must ensure that the gateway to the application is available close to the location you are bringing it to. Also, instruct the network to retrieve the data required by the application. Operators need to be able to deploy an agile user plane function at their site that can scale to meet the application’s demands.
This is the heart of network operation. Traffic routing is a crucial area. Infrastructure and orchestration are focused on application hosting, and the environment traffic routing brings in information and awareness within the network, which CSPs have, such as the location and the user’s content. If a user requests data via a video streaming application, you will see that the CSP network has this information. The user IP session request is routed from the user’s current location to the user plane function near the streaming service’s server. There are many options to route the user’s IP session from the edge. However, traffic routing is not always within the operator’s network.
Either all traffic can be brought to the edge and a portion of it can be taken to the edge. The latter option is more efficient, as we can centrally manage the rest. The leading technologies for routing edge traffic are distributed Anchor, session breakout, and multiple sessions. How do CSPs choose the right technology?
It will ultimately depend on the intended use and application. Distributed Anchor is a simple mechanism that can be deployed on top of existing 4G or 5G networks. It would make an intelligent choice for those who want to use their existing networks. Session breakout, which requires the development of complex features within multiple products, is more complicated and specific to 5G. It is not available in 4G. Numerous sessions, which is a promising technology but not yet established due to device ecosystem dependency, will be limited to 5G. But, 5G will likely be the dominant technology by then.
We want to achieve this traffic separation at the end of each day. If you haven’t yet invested in session breakout technology, the distributed Anchor should be easy to deploy and evolve to the multiple session mechanism in the future, bypassing the need for session breakout. Before deciding on the technology, you must fully understand its benefits and costs.
Edge exposure can be viewed from two perspectives: orientation at the edge and direction for the edge. Exposure at the edge means exposing assets such as edge discovery information and UE IP-to-network identity translation information – vital information for systems that have to locate sites and connect to them.
Exposure at an edge refers to exposing capabilities at the edge to allow applications that reside there to be seen. These might include location information, quality information, or information about user equipment. These capabilities can be exposed so that you don’t need to return to a central location to access them in low-latency-dependent situations. Importantly, this information must be made available in a way that is easily identifiable by the network. It can also be converted into valuable and consumable formats by applications.
Ericsson’s Edge Exposure Server is part of Ericsson’s Cloud Core Exposure Server.
CSPs can gain a competitive advantage in emerging areas such as 5G, especially in gaming, industry, and healthcare enterprise use cases. 25 percent of emerging 5G use cases will rely on the edge next year.