Tan Lei, Sweden Yi Neng Software Co., Ltd. Degrees Of Freedom Platform,Wave Simulation Platform,Shipboard Motion Simulation Bench,flight sim motion platform Suzhou Johnson Automation Technology Co., Ltd. , https://www.cn-johnson.com
**Industry Background**
In the telecommunications sector, the rapid advancement of mobile technologies and the increasing number of IoT devices have led to a significant surge in demand for 5G communication bandwidth. As illustrated in Figure 1, compared to 4G technology, 5G offers a 10-fold increase in data transmission speed and network latency reduction, while the Transmission Time Interval (TTI) has been cut from 1ms to 0.2ms. Additionally, both network capacity and base station density have seen substantial growth. This evolution demands a dramatic improvement in the data processing capabilities of 5G base stations, along with stricter real-time performance requirements for their software systems. With the emergence of Network Function Virtualization (NFV), the hardware and software architecture of 5G base stations also needs to be redefined when compared to 4G systems.
Figure 1. Comparison of key 4G and 5G indicators
**Trends in 5G Base Station Technology**
Driven by these new features, 5G base stations are undergoing significant technological transformations. One of the key trends is the adoption of C-RAN (Cloud Radio Access Network), which brings the L1, L2, and L3 layers closer to the cloud than the traditional distributed approach used in 4G base stations, where all layers were handled near the antenna. C-RAN supports different levels of centralization depending on application needs, allowing for flexible functional module partitioning between the cloud and RAN. As shown in Figure 2, 5G C-RAN can support traditional 4G-style distributed processing, fully centralized processing on the cloud, or semi-centralized configurations where only certain functions like PDCP are centralized, while others remain on the RAN side.
Figure 2. Functional structure partitioning of different centralized processing methods for 5G C-RAN
This flexibility in processing methods requires the hardware and software architecture of 5G base stations to be more adaptable, as shown in Figure 3, to meet the varying demands of different service scenarios.
Figure 3. Evolution of 4G to 5G base station technology
Another major trend is the move toward virtualization, driven by NFV technology. As shown in Figure 4, this allows BBUs at the antenna end to be virtualized into resource pools, managed and scheduled from the cloud, optimizing resource usage and configuration.
Figure 4. Evolution of 5G base station technology to virtualization
**Challenges**
As a cutting-edge technology, 5G faces numerous challenges. One of the most critical is the need to reduce the TTI from 1ms to 0.2ms while increasing data rates by 10 times. This demands a significant boost in base station processing speed and an even higher standard for real-time software performance. Moreover, the dynamic load changes in different L2 function partitions under C-RAN virtualization require more advanced scheduling and processing capabilities. These technical hurdles necessitate a highly efficient real-time operating system and superior software performance in 5G base stations.
Additionally, for NFV-based virtualization, 5G requires robust platform software that can support various virtualization services and integrate seamlessly with the L2 real-time operating system.
**ENEA’s Operating System Software Solution for 5G Base Stations**
To meet the high bandwidth and data rate demands of 5G, the high-performance functional modules in the L1 layer of the baseband must be processed using specialized hardware such as FPGAs, ASICs, and hardware accelerators. The remaining L1 modules, along with L2 and L3 layers, can run on multi-core SoC processors, executing a multi-core real-time OS for resource management, running NFV platform software, and enabling different levels of centralized processing based on C-RAN deployment models.
ENEA, a global leader in real-time operating systems and NFV platform software, provides comprehensive software solutions for 5G base stations, supporting flexible and diverse deployment scenarios across 5G C-RAN technology.
**Scenario 1: Enea OSE RTOS + Enea Linux**
In this solution, the parts of the L1 layer requiring lower computational power and the full functionality of the L2 layer are executed on multi-core processors using Enea OSE RTOS, which ensures strict real-time performance. Meanwhile, the L3 and O&M layers, which have less stringent real-time requirements, run on Enea Linux. This setup mirrors traditional base station architectures, as shown in Figure 5, and corresponds to the distributed split C-RAN deployment method in Figure 2.
Figure 5. Enea OSE RTOS + Enea Linux
**Scenario 2: Enea OSE RTOS + Enea NFV**
With the rise of NFV in 5G C-RAN, multiple virtual machines can be hosted on the same hardware to decouple network functions from physical hardware, enabling more flexible resource allocation and cost reduction. ENEA’s NFV platform perfectly meets this requirement. In this scenario, the L1 and part of the L2 layers still run on Enea OSE RTOS, while the remaining L2 and L3 layers are virtualized using Enea NFV on x86 or ARM servers, as shown in Figure 6. This corresponds to the PHY Split and PDCP Split C-RAN deployments in Figure 2.
Figure 6. Enea OSE RTOS + Enea NFV
**Scenario 3: Enea OSE RTOS as Guest OS + Enea NFV**
For the ultimate goal of full virtualization in 5G C-RAN, ENEA offers a solution where all L1, L2, and L3 layers are virtualized. This enables optimal cloud-side resource configuration and management. In this setup, Enea NFV not only virtualizes the L3 and O&M layers but also the L1 and L2 layers. Multiple VMs are created on x86 or ARM servers, with Enea OSE RTOS running as a guest OS within the L1 and L2 VMs to meet real-time requirements, as shown in Figure 7. This corresponds to the fully centralized C-RAN deployment in Figure 2.
Figure 7. Enea OSE RTOS as Guest OS + Enea NFV
**Conclusion**
As an evolution of 4G, the successful implementation of 5G relies heavily on the accumulated expertise of communication equipment manufacturers and operating system software vendors, along with deep integration between software and hardware. ENEA, established in 1968 and listed in Sweden in 1989, is a leading provider of real-time operating systems and NFV platform technologies in the mobile communications industry. It is also one of the major contributors to the OpenNFV open-source initiative.
To address the challenges and trends in the 5G industry, ENEA has developed three distinct operating system software solutions for 5G base stations, tailored to different deployment methods under various service scenarios:
• **Distributed Architecture**: Suitable for rural areas with wide coverage and low population density. It has a flat structure and is easy to deploy.
• **Fully Centralized Architecture**: Ideal for densely populated urban areas such as offices, shopping malls, stadiums, and plazas, where high network performance is required.
• **Semi-Centralized Architecture**: A middle ground between the two, suitable for sites that are moderately large and moderately crowded.