Wind River Systems’ Open-source Cloud Platform Designed to Support 5G vRAN at the EdgeNovember 19, 2019 by Robin Mitchell
The platform supports 5G applications at the edge through a Single Pane of Glass—the automated management of thousands of nodes.
Wind River has announced its new Kubernetes-based cloud platform designed to meet the complex needs of an edge cloud infrastructure that must support 5G, IoT, or MEC applications.
With the Wind River Systems Cloud Platform, service providers can experience the benefits of both a cloud-native and containerized architecture—namely, the ability to manage geographically separated vRAN infrastructures.
The technologies brought together in this platform use the concept of a Single Pane of Glass (SPoG), which suggests that a service provider can sit in front of a single computer screen and manage an entire distributed network.
Diagram of the Wind River Cloud Platform architecture. Image from Wind River Systems
Wind River's VP of Telecommunications, Paul Miller, remarked that "Container technology no longer uses a hypervisor context. It doesn't create a full virtual machine. It creates a lighter weight virtual environment in which you can host different applications, and this allows for far more virtualization density in a particular server."
Miller also noted that container technology, like the new platform, "creates a more performative environment, where the applications have less latency and higher throughput."
A Kubernetes-based Cloud-native Solution
At present, current core data centers may not be equipped to support the network pressures of 5G and IoT applications. The 5G network alone has multiple network types, such as micro and macro, for connection to edge devices; local and central servers communicate to connect all of these cells together.
If this network was managed in the traditional sense, each 5G-based station would have to be designed to handle maximum traffic (as well as local servers), making it an expensive option.
Wind River has combined multiple architectures (including cloud, Kubernetes, and containers) to produce a 5G-virtual radio access network. The use of common virtualization architecture allows for each element (far edge, near edge, local, and central) to be solved with a single approach. The distributed infrastructure can then be managed as a single unified element.
The evolution from traditional deployment to container deployment. Image from the Linux Foundation
The cloud platform also provides scaling solutions for network handling—from a single compute node at the network edge to thousands of nodes. The remote nodes can survive disconnection and will not only automatically reconnect, but will also resynchronize with the cloud.
Miller observes that with the trend "from 3G to 4G to 5G using this container technology" comes the need to implement servers at every cell tower to deploy 5G—an expensive and resource-intensive endeavor. "You need to manage a widely distributed implementation of containerized technology and that's what this product is geared to do," Miller explains.
The Role of Open-Source
One of the key components of the Wind River Cloud Platform is the use of Kubernetes, an open-source container orchestration system for automating application deployment, scaling, and management. Simply put, Kubernetes is what allows the cloud platform to scale, virtualize, compartmentalize, and contain all levels of the 5G vRAN deployment.
Miller describes how open-source played a role in the cloud platform's development: "What we've done is started a separate open-source initiative called Starling X that took Kubernetes as a fundamental foundation and then built a geo-distributed, operationally deployable version of it that can enable a service provider to deploy at a massive scale."
Wind River Cloud Platform logo. Image from Wind River Systems
Starling X is an open-source project of the OpenStack Foundation.
The use of open-source software is important in this platform for a number of reasons:
- Open-source software is free from proprietary property, which almost always comes with a price tag.
- Open-source software is often maintained by a community who can provide updates and patches more regularly than companies of close-source software.
- Open-source software is more accessible to the wider public, whose adoption of the software may lead open-source to become an industry standard.
The use of industry-standard platforms allows for a wide variety of solutions to be executed without having to worry about device or platform specifics. Engineers who may want to deploy different network applications can easily do so using Kubernetes. They can integrate the application either at the data center level or at the edge level.
While telecommunication is the most obvious industry in which the cloud platform might take effect, the platform also opens up applications in different verticals that are connected to a 5G infrastructure: 5G-virtualized radio infrastructure, vehicle-to-vehicle communications for autonomous self-driving vehicles, remote telemedicine, augmented reality for industrial applications, and industrial IoT, among others.
The Wind River Cloud Platform may have a variety of applications, including remote telemedicine.
This technology is applicable to design engineers in a number of capacities, considering that this Linux, cloud-based solution runs on top of hardware infrastructure, like accelerator cards and FPGAs.
Wind River Systems posits that their new cloud platform will be especially important for 5G, mission-critical cases. Additionally, Miller sees this new cloud infrastructure as another major transformation in telecommunication, comparable to the transition from TDM to IP.
He explains, "This isn't just about the evolution of 4G to 5G; it's really the transformation of the entire telecommunications infrastructure to a fully virtualized, distributed architecture."
- Tricking Autonomous Driving Systems Could Be as Simple as Subtly Altering Street Signs
- Build a Temperature and/or Humidity Controlled Fan with a Picaxe 08M2 Microcontroller
- Resolving the Signal Part 4: Understanding Effective Noise Bandwidth in Precision Delta-Sigma ADCs
- Radar Maximum Range Calculator