Understanding Software Defined Radio and Networks in 5G Architecture

Latest update time：2022-03-29 11:33

Reads：

Source: The content is compiled from embedded by Semiconductor Industry Observer (ID: icbank) , thank you.

The world of wireless communications is about to change as 5G finally reaches end consumers. One of the biggest promises of 5G is massive device communications, powering revolutionary IoT systems such as self-driving cars, metaverse hardware, gaming virtual reality (VR), and smart factories. Some of the 5G technologies required for this revolution include machine-to-machine (M2M) communications, massive machine-type communications (mMTC), ultra-reliable low-latency communications (uRLLC), and enhanced mobile broadband (eMBB). In this context, optimization of base stations is critical to provide low-latency connectivity, optimal sharing of spectrum and processing resources, and dense small cell deployments.

Furthermore, 5G will provide converged network communications across multi-technology networks, and open communication systems that work with satellite, cellular networks, cloud, data centers, and home gateways. In addition, 5G systems will be autonomous and able to adjust themselves based on the required QoS to dynamically handle application-driven networks. In this context, we will discuss here the orchestration of 5G mobile service-based architecture (SBA) enabled by open radio access network (O-RAN) technology. This article also explores the use of software-defined radio (SDR) and software-defined networking (SDN) in 5G, which support network function virtualization (NFV), network slicing, cloud/edge computing, artificial intelligence (AI), and machine learning (ML).

5G Network Architecture

The first component of the 5G structure is the transport network, which connects the 5G RAN to the core network. It can be divided into three structures: fronthaul, midhaul, and backhaul (see Figure 1). The distributed unit (DU) is connected to the remote radio unit (RRU) through the fronthaul network. Each DU can cover a distance from a few kilometers to more than 50 kilometers and control multiple antennas. The midhaul performs intermediate connections by linking the distributed unit (DU) to the central unit (CU). Finally, the backhaul link connects the central unit and the remote/mobile system to the core network.

Figure 1: 5G network architecture consists of three structures

In addition to the transport network, the 5G core network also contains multiple components for access and control. In the SBA architecture, the components are arranged in a set of interconnected network functions (NFs), including NF repository function (NRF), network slice selection function (NSSF), policy control function (PCF), user plane function (UPF), session management function (SMF), access and mobility management function (AMF), and data network (DN). At the user equipment (UE) side, access is controlled and performed by gNB nodes, which communicate with AMF and UPF services through NG interfaces.

The NG interface carries the user plane and control plane protocols: the user plane implements the PDU (Protocol Data Unit) session, and the control plane controls the session and the connection to the network, including service requests and transmission resources. Access and mobility management function (AMF) and data network (DN). On the user equipment (UE) side, access is controlled and performed by gNB nodes, which communicate with AMF and UPF services through the NG interface. The NG interface carries the user plane and control plane protocols: the user plane implements the PDU (Protocol Data Unit) session, and the control plane controls the session and the connection to the network, including service requests and transmission resources. Access and mobility management function (AMF) and data network (DN).

On the user equipment (UE) side, access is controlled and performed by gNB nodes, which communicate with the AMF and UPF services via the NG interface. The NG interface carries both the user plane and the control plane protocols: the user plane implements the PDU (Protocol Data Unit) session, and the control plane controls the session and the connection to the network, including service requests and transmission resources.

To better understand the advantages of 5G, let's compare it to the huge 4G/LTE technology. First, the core of 5G technology is fundamentally different, using millimeter wave, massive MIMO connectivity, cloud-native software design, and a high level of system virtualization. Second, 3GPP 5G is a service-based architecture, which means that system elements are defined as network functions (NFs) that provide services to other NFs with authorized access. The service-based nature is more attractive than the 4G/LTE implementation because it provides network slicing, function virtualization, cloud-based systems, and better compatibility with Open-RAN technology. In addition, the implementation of UPF, to decouple gateway control and user plane, and AMF, to separate session management from connection and mobility management, are not found in the 4G protocol. In 5G, the user plane and control plane are decoupled because UE traffic is 1000 times that of 4G. Finally, the 5G system allows the use of smaller and more specialized network cells, such as femto cells and pico cells.

One of the most important aspects of 5G is the decoupling and virtualization of RAN elements, which allows for smarter, dynamic and flexible networks for different applications. At the forefront of the RAN development movement is the Open RAN (O-RAN) architecture. By opening up the interfaces between RAN components, O-RAN allows operators to combine different vendors in the same system, thereby increasing flexibility and giving operators the freedom to work with the technology provider of their choice.

In O-RAN, the base station is divided into two: the centralized unit (CU) and the distributed unit (DU) (Figure 2). The CU is responsible for larger time-scale functions, while the DU is responsible for time-critical tasks. At the end of the chain, the remote radio unit (RRU) manages all RF communications and components such as modulation, coding, and interference avoidance. In terms of the protocol stack, the CU handles the high layers, the DU manages the low layers, and the RRU handles the physical layer. The open interface between the CU and DU is called the high-layer split (HLS), while the connection between the DU and the RRU consists of the low-layer split (LLS) interface.

All O-RAN applications run on the RAN Intelligent Controller (RIC). The RIC platform provides abstraction, integrated optimization and automation algorithms for RAN components. All O-RAN applications run on the RAN Intelligent Controller (RIC). The RIC platform provides abstraction, integrated optimization and automation algorithms for RAN components. All O-RAN applications run on the RAN Intelligent Controller (RIC). The RIC platform provides abstraction, integrated optimization and automation algorithms for RAN components.

Figure 2: Shows the Open RAN (O-RAN) architecture

Software Defined Radio (SDR)

Software Defined Radio or SDR is a radio system consisting of an analog radio front end (RFE), an FPGA-based digital unit and a mixed-signal interface, usually through an ADC and DAC. The RFE is responsible for receiving and transmitting the analog part of the RF signal, which is discretized by the DAC/ADC interface. The RFE is an important part of the circuit because it defines the signal range, number of channels and bandwidth.

The highest performance RFEs on the market can achieve 3 GHz of instantaneous bandwidth, using up to 16 independent channels. At the heart of an SDR is an FPGA configured with DSP functions: modulation/demodulation, up/down conversion, and data packetization. An FPGA is a fully reconfigurable matrix of digital logic, so the same system can support multiple processing algorithms, state-of-the-art protocols, and even artificial intelligence without changing the hardware. SDRs offer low latency, flexibility, high interoperability (important for the 5G physical layer), and massive MIMO capabilities—useful for beamforming and spatial multiplexing. A commercial example is Per Vices’ Cyan SDR (Figure 3), which can be used as the core of a 5G base station and test bench/emulator.

Figure 3: Per Vices Cyan can be used in 5G base stations

In a 5G environment, both the RRU and the baseband unit (BBU) can contain one or more SDR units to perform radio-related functions, providing compatibility, interoperability, and flexibility. For example, in a gNodeB 5G BBU, the connection to the RRU is achieved using eCPRI fiber.

In these cases, the SDR must include both eCPRI and Gigabit Ethernet (GBE) ports, as well as the ability to handle MIMO antennas. On the other hand, the RRU SDR needs to fit the frequency range of the application, which can fall into the FR1 or FR2 category. FR1 (Frequency Range 1) covers sub-6GHz frequencies (600 to 6000 MHz), while FR2 (Frequency Range 2) covers the 24.25 to 52.6GHz band. The FR2 band is suitable for shorter-range/higher-bandwidth applications compared to FR1. The RRU SDR must be selected and configured to operate within the required spectrum. Small cells also benefit from SDR implementation, as complete RF solutions that are lightweight, low-power, and compact are readily available in the market.

The importance of SDR implementation stems from its role in the O-RAN system. The three most important O-RAN hallmarks are disaggregation, virtualization, and software-based, the last of which is provided by SDR. Software-based is fundamental to enabling URLLC, eMBB, and mMTC capabilities. In addition, SDR-based systems are flexible, upgradeable, and interoperable, allowing operators to control the RAN without constantly replacing hardware. SDR can also comply with instructions generated by the RIC, which is critical for RAN optimization and automation.

Software Defined Networking (SDN)

Software Defined Networking (SDN) is a physical separation between control plane functions and forwarding functions. A typical SDN architecture is divided into three parts: application layer, control layer (where the SDN controller runs), and physical infrastructure. The layers communicate with each other through APIs (northbound APIs for application control communication and southbound APIs for controlling the infrastructure). SDN improves programmability and enables higher levels of network automation and optimization. It also provides cloud-like capabilities within the fabric, allowing for centralized computing and network control abstraction from the physical layer, data analytics algorithms, and system virtualization through virtual overlay networks. System virtualization supports one of the most important features in 5G:

Network slicing refers to the division of a physical network into multiple virtual networks that are unique and optimized for specific services or applications. Each virtual network or slice can be configured with only the specific resources required to perform a specific task, such as autonomous vehicles, IoT devices, and mobile services. The most obvious advantage of this technology is the optimization and adjustment of resource allocation to meet the needs of specific customers and market segments. Client services can be divided into eMBB, mMTC, and urLLC, each category has its own throughput, bandwidth, latency, and robustness requirements (Figure 4). Network slicing is achieved by combining SDN, SDR, network function virtualization, data analytics, and automation.

Figure 4: This is an image of 5G network slicing

End-to-end automation, especially to design a network slicing approach, is critical to network function virtualization (NFV). This approach enables the virtualization of RAN and core network functions that were once performed by hardware, such as routing, scaling, security, and load balancing. By implementing network functions in software, operators can continuously update network functions using state-of-the-art algorithms without replacing hardware, saving time, reducing installation costs and customer disruption. In addition, NFV allows real-time repurposing and reallocation of functions through network slicing, as well as inter-slice and intra-slice control of RAN resources.

SDR and SDN/NFV for optimizing network resources

The massive data throughput required by 5G systems can easily overwhelm state-of-the-art LTE networks. For example, a typical CPRI-based LTE fronthaul typically handles channel bandwidths of around 10-20 MHz, which translates to around 10 Gbps in a 10-channel connection. 5G, on the other hand, handles bandwidths in the 100 MHz to 500 MHz range, and with massive MIMO extensions, fronthaul throughput can reach into the Tbps range. CPRI fiber is no longer sufficient, and optimization techniques such as enhanced CPRI (eCPRI) are needed. In the eCPRI interface fronthaul, the physical layer functions are split between the RRU and DU in an optimized ratio, increasing the complexity of the RUU while reducing the load on the fronthaul. The requirements for performance optimization are not limited to the fronthaul, as the location, access, and management of resource instantiations are largely determined by the requirements of the service slice. In this case, SDR and SDN/NFV-based structures (Figure 5) can help.

There are several different types of orchestration and control for 5G optimization. For example, the software-defined RAN (SD-RAN) community is developing open source RIC controllers that are compatible with O-RAN. The SD-RAN project is focused on developing near real-time RIC (nRT-RIC) to optimize the dynamics and latency of network control, the most prominent of which is the open source µONOS-RIC. In addition to its open source nature, µONOS-RIC is also compatible with AI/ML-based applications that can be optimized for massive MIMO, self-organizing networks (SON), and intelligent radio resource management (RRM). Another recently developed optimization technology is the cross-layer controller (CLC), which is applied to resource allocation and pairing between network slices based on real-time monitored RAN conditions.

Figure 5: SDN/NFV can be applied to 5G RAN to optimize performance

In an O-RAN based architecture, the main goal of network optimization is to improve overall performance under various conditions, prevent network instability, and resolve issues with minimal service loss. It does this by constantly measuring KPIs and crowdsourcing information, and making decisions to control and adjust units accordingly. This prevents congestion, overload, and interference, and reduces latency. In O-RAN, optimization is performed through nRT-RIC. External intelligence can run on top of nRT-RIC to make decisions based on AI/ML algorithms. AI/ML driven nRT-RIC supports the use of advanced management algorithms such as dynamic spectrum sharing (DSS) and NSSI resource allocation optimization.

In O-RAN architecture, SplitOption 7-2x LLS complies with multiple optimization techniques, including beamforming optimization. Beamforming can increase data throughput and the number of parallel connections by focusing RF beams to specific locations, and improve the power efficiency and signal-to-noise ratio of the network. Massive MIMO antennas play an important role in beamforming optimization. In these systems, the controller sets a global optimization target and each MIMO cell contributes part of the beam. The SDR BBU is the basis for dynamic and coherent coordination of MIMO antennas.

Current Research and 5G O-RAN Testbed

The 5G architecture for O-RAN introduces several challenges in network design. Researchers are still trying to solve several technical bottlenecks, such as how to provide short-overhead data access for AI agents, how to design robust data-driven control loops, and what are the exact roles and requirements of each RAN component. The SD-RAN community is one of the research teams trying to solve these problems. As mentioned earlier, SD-RAN has developed an open source nRT-RIC compatible with AI/ML applications, which provides the necessary technologies and abstractions for data-driven control loops and intelligent allocation. On the other hand, the OpenRF Association aims to develop a highly interoperable 5G ecosystem, including RF hardware and software, to reduce integration costs and time to market while maintaining sufficient flexibility and customization. Neither the SD-RAN nor the OpenRF project would be viable without the use of powerful SDR and SDN.

It is impossible to discuss 5G research without discussing simulators, and specifically the Colosseum testbed. Colosseum is the world’s largest network emulator testbed, with 256 SDRs capable of emulating up to 65,536 RF channels (100 MHz). This massive system can be used with GNU Radio, MATLAB, and most DSP technologies, and provides a great test framework for AI/ML algorithms, MIMO systems, and O-RAN in general. Colosseum can also simulate path loss, multipath, and fading, providing RF conditions similar to real-life environments. Leonardo Bonati’s research team recently used Colosseum to validate the feasibility of network control using a Deep Reinforcement Learning (DRL) agent running on top of nRT-RIC via xApps. The algorithm is compatible with O-RAN and operates by selecting the most suitable scheduling strategy for each RAN slice, taking into account URLLC, MTC, and eMBB. The DRL system achieved 20% higher spectral efficiency and 37% lower buffer occupancy compared to other approaches.

in conclusion

This article discusses many aspects of 5G mobile SBA systems, including orchestration, implementation, management, and functionality, with a focus on the Open-RAN architecture. The Open-RAN community is driving the development of new 5G solutions by using open and disaggregated interface standards between RRU and BBU. In this context, SDR and SDN play an important role in the 5G revolution by providing flexibility, interoperability, softwarization, and virtualization of RNA, the fundamental tools for enabling unique 5G features such as network slicing and DSS. SDR is also highly used in new technology development and 5G research for softwarization, real-time monitoring and control, AI/ML applications, and large-scale RAN emulation.

*Disclaimer: This article is originally written by the author. The content of the article is the author's personal opinion. Semiconductor Industry Observer reprints it only to convey a different point of view. It does not mean that Semiconductor Industry Observer agrees or supports this point of view. If you have any objections, please contact Semiconductor Industry Observer.

Today is the 2992nd content shared by "Semiconductor Industry Observer" for you, welcome to follow.