Intel and Canonical have been collaborating to deliver advanced infrastructure that can run the networks of the future. More and more, we see edge cloud systems deployed closer to end users and systems to meet the ever-increasing demand for faster edge networks that can rapidly deliver business services and boost productivity. This requires highly efficient technical integration of multiple complementary technologies at the network’s edge, including operating system, cloud infrastructure control and management systems, and hardware that enables fast information processing.

Our goal has been to deliver exactly that: an ultra-reliable edge cloud with extremely low latency in information delivery between business software and devices. In this blog post, we talk about how the Intel FlexRAN reference architecture with Real-time Ubuntu can boost performance at the telco edge.

Ultra-reliable low latency communications (URLLC)

5G mobile networking standards published by the Third Generation Partnership Project (3GPP) define a set of distinct telecommunication use cases. One of these is called URLLC, short for ultra-reliable low latency communications.

URLLC was introduced as a separate use case in 5G to address the performance requirements of latency-sensitive business applications that require highly-reliable communications. Business applications in multiple industry sectors benefit from URLLC, such as manufacturing, healthcare, automotive, and energy, to name a few.

These applications provide the necessary set of operational functionalities to business devices, such as control and actuation, as well as remedial features such as monitoring, alerting and healing. In these systems, there is tight coupling and time-sensitive control loops between devices and software applications, which require a fast and reliable communication network.

The need for mobile networking in URLLC scenarios

A fast and reliable network can be set up with multiple technologies, which can be wired or wireless. In the specific setting of industrial facilities where a large number of devices operate at multiple locations, it is impractical to deploy and maintain wired connectivity. This is because devices may be mobile or their locations could be altered frequently, and the operational environment is often not suitable for cabling for many sensors and actuators. This calls for wireless networking, which is much more suitable as opposed to the inherent deployment complexity and inflexibility of wired communication. There are plenty of wireless technologies available in the market today, but you need to choose the right one.

Mobile communication networks are the most suitable wireless technology for edge environments. They can provide high levels of security and reliability with well-defined protocols that have evolved over generations of 3GPP releases. What’s more, URLLC support is now possible in telecom systems thanks to the advancements in 5G, especially in the radio interface and the system architecture.

Mobile networks as software for URLLC

The 5G system architecture enables URLLC thanks to its modern structure based on microservices. Each component of the mobile core and the radio access networks (RAN) is highly specialised and deployable as a microservice on cloud-native infrastructure. This is a result of the telecom industry’s gradual shift from fixed and rigid systems that run as all-in-one appliances to flexible, modular and scalable architectures.

A flexible and modular architecture is advantageous in many aspects. Deployment and operational maintenance of each microservice can be governed separately. It allows for configuration and orchestration of the lifecycle and runtime parameters of each microservice as necessary in order to meet the performance requirements set by service level agreements (SLA).

The operational flexibility with microservices makes it possible to have different flavours of network services offered to customers by means of different sets of configurations that fit the particular characteristics of different data traffic types and use cases. This means, an operator can now offer a dedicated network service as offered to URLLC business applications.

Virtual radio access networks at telco edge clouds

Microservices run on cloud-native software infrastructure, so that they can be deployed, configured, removed, scaled and migrated on demand, providing greater flexibility to operators. Operational costs are also reduced, as now it is possible to deploy and maintain systems as software.

A network running as microservices is a particularly attractive proposition for telco edge clouds, which need low maintenance and cost-efficiency. What remains though is performance - edge clouds need fast packet processing at hardware, and with minimal delay at every layer of the software stack.

The system architecture based on microservices has evolved how we run modern radio access networks (RAN). Softwarisation has enabled RAN to be decomposed into separate modular components that can be run as software. This has brought the concept of virtual RAN, where RAN systems are operated on telco clouds as distributed software pieces.

Since the protocol stack runs as software, it is possible to distribute it across systems. There are multiple ways to split the stack. The key is to ensure that any part of the stack that requires close interaction with end devices over the air interface receives fast processing support.

The lower layers of the communications stack require this support for real-time processing. Layer 1 (L1) performs low-level wireless signal processing tasks and Layer 2 (L2) provides wireless channel access functions with resource sharing and scheduling capabilities. These parts of the stack can be conveniently placed at the telco edge and frequently interact with radio units in short time frames.

URLLC with Intel FlexRAN reference architecture and Real-time Ubuntu

Performance with low packet latency greatly matters at the telco edge cloud for URLLC applications. 5G systems have multiple enhancements in the air interface protocols and physical transceiver technologies to boost service delivery performance. However, for true end-to-end latency reduction between end devices and a URLLC application running at an edge cloud, the edge cloud itself requires real-time performance. This must be delivered at both hardware and software levels.

Intel and Canonical have been working in close partnership to provide an effective solution for fast and ultra-reliable processing suitable for URLLC use cases at telco edge clouds. It is a combination of complementary technologies at the silicon and operating system levels. Together, Intel’s cutting-edge silicon and the Ubuntu real-time kernel provide a robust stack from hardware to software which can address the needs of URLLC in modern radio access networks.

Next generation Intel CPUs with Intel Time Coordinated Computing (TCC) deliver optimised computing performance with low latency, supporting IEEE 802.1 Time Sensitive Networking (TSN). TCC prioritises real-time workloads at the hardware level when accessing cache, memory, and networking resources. With support for TSN, Intel TCC ensures timeliness through synchronisation across network nodes. Intel CPUs with FlexRAN reference software integrate advanced signal processing, hardware acceleration and highly efficient input/output (I/O) modules for real-time performance.

Ubuntu’s real-time kernel accompanies Intel TCC and FlexRAN reference software by prioritising real-time tasks at the software level, and ensures their execution is uninterrupted, delivering minimal response time. It provides the necessary bounded time delay guarantees in the operating system to URLLC applications powered by FlexRAN applications.

Intel FlexRAN Reference Architecture

Intel FlexRAN is a reference implementation for cloud-native wireless access virtual network functions, compliant to 3GPP standards. It allows mobile operators to run their L1-L2 stack as virtual functions with high efficiency. The reference software includes kernel modules optimised for the Intel Architecture (IA) instruction set, which can efficiently run both 4G and 5G workloads.

The 4th Gen Intel Xeon Scalable processors with Intel vRAN Boost deliver up to twice the capacity for vRAN workloads compared to the prior generation¹. Intel vRAN Boost is a unique architectural design that integrates vRAN acceleration directly into the CPU, which provides around 20% additional compute power savings¹. The integration includes advanced signal processing, hardware acceleration on advanced embedded silicon-based accelerators, highly efficient input/output (I/O) with DPDK and reduced interface delays with SR-IOV.

FlexRAN reference software includes a task controller, which distributes computation tasks across multiple CPU cores by means of real-time scheduling and scaling. This enables processing of information collected from many radios across many CPU cores, making it an ideal solution at the edge cloud where distributed vRAN functions are run.

The FlexRAN solution scales in different telco edge use cases. In URLLC, the prime deployment scenarios for FlexRAN reference software are in RU and DU environments where the L1-L2 mobile networking stack, 5G user plane function (UPF) and multi-access network (MEC) applications require real-time performance.

Real-time Ubuntu

Canonical provides optimised, secure, and efficient processing to telco workloads with support for real-time kernel as an enterprise-grade operating system. Real-time Ubuntu provides agile processing capabilities at the operating system kernel, perfectly complementing Intel Xeon and Core processors at the telco edge.

Real-time Ubuntu is silicon-optimised to work with 4th Gen Intel Xeon Scalable processors, ready for production systems that run URLLC applications as well as virtual radio access network workloads at telco edge clouds. It provides optimal performance and bounded low latency to these delay-sensitive workloads by means of thread prioritisation and core pinning.

Canonical offers real-time kernel with Ubuntu Server 22.04 LTS, via Ubuntu Pro - Canonical’s comprehensive enterprise security and compliance subscription. Besides support for real-time kernel, Ubuntu Pro includes security features for telco infrastructure and applications, with a 10-year expanded security maintenance commitment.

Besides the URLLC application workloads that run on server hardware at telco edge clouds, delay-sensitive edge devices can also benefit from support for real-time kernel processing. URLLC devices can run Ubuntu Core with a real-time kernel, and benefit from real-time kernel processing. Ubuntu Core is the fully containerised Ubuntu variant optimised for edge devices, such as those at the telco edge cloud.

Summary

Telco edge clouds have multiple scenarios where business applications require ultra-reliability and low latency in compute and networking. 5G mobile telecommunication networks can deliver URLLC support to sectors such as automobile, manufacturing, automation, healthcare, and smart city services, thanks to its support for microservices based cloud-native networking at 5G core and radio access networks. However, besides the system architecture and the protocol stack support for URLLC, it is essential that low-latency with high reliability is achievable at the silicon and operating system levels.

Intel’s and Canonical’s cutting-edge products together deliver a complete solution for real-time support of URLLC business applications and the 5G networking protocol stack at the telco edge. It is a combination of ultra-fast and efficient CPU architecture in Intel Xeon and Core processors, which is accompanied by Intel Time Coordinated Computing, Intel FlexRAN reference software and the silicon-optimised real-time Ubuntu operating system kernel delivered by Canonical for Intel FlexRAN reference architecture. Canonical makes it easy to get Real-time Ubuntu for URLLC applications running on Intel silicon with an Ubuntu Pro subscription, which brings security and support for your telco edge infrastructure.

Contact us

Canonical provides a full stack for your telecom infrastructure. To learn more about our telco solutions, visit our webpage at ubuntu.com/telco.

Further reading

• Reduce latency in telco edge clouds with real-time Ubuntu
• Learn more about real-time kernel on Intel SoCs for industrial applications
• Download Ubuntu for your Intel IoT hardware

• ¹   For workloads and configurations visit www.Intel.com/PerformanceIndex. Results may vary.

Intel, the Intel logo, Xeon, Core and FlexRAN are trademarks of Intel Corporation or its subsidiaries.