ORAN NEAR-RT RIC: COMPARISON OF OPENSOURCE SOFTWARE SOLUTIONS ORAN SC AND SD-RAN

5G is being developed with the understanding that the network needs to be adaptive and dynamic. Specifically, it needs to optimally support a variety of new business verticals that require low-latency, mission-critical, and massive machine-type communications, not just voice and broadband data. Consequently, there is a huge movement around virtualized networks, and the operators are transforming their deployment strategies based on disaggregation.

Disaggregated RAN, based on NFV, SDN, and cloud principles, along with network artificial intelligence (AI), is a critical factor that will allow rapid innovation and onboarding of new services while lowering CAPEX and OPEX.

As a result, telecommunications operators and network equipment providers (NEPs) have set up the O-RAN Alliance with a charter to establish standards around the disaggregated RAN intelligent controller (RIC), allowing inherent multi-vendor interoperability of different components from different vendors.

This blog describes the open-source RIC initiatives and their positioning regarding near-Real-Time RIC (near-RT RIC) and xApps and provides recommendations for implementation.

RIC is a key element to enable best-of-breed Open RAN to support interoperability across different hardware (RU, servers) and software (DU/CU) components, as well as ideal resource optimization to deliver the best subscriber quality of service (QoS).

Today, the disaggregated RAN ecosystem includes cloud service providers, communications service providers, and network equipment providers. Figure 1 shows how these service providers and equipment providers are working together in this highly collaborative ecosystem. Two of the key initiatives – the O-RAN Alliance and Telecom Infra Project (TIP) – have complementary missions. Both focus primarily on deploying end-to-end disaggregated telecom infrastructure in various environments. Also, two key open-source initiatives – ORAN-SC, supported by the Linux Foundation, and SD-RAN, supported by ONF – are focused around the near-RT RIC along with xApp SDK and other xApps.

Figure 1. The many participants in the ORAN ecosystem

Source: ONF SD-RAN Techinar 2021¹

Near-RT RIC enables near-real-time control and optimization of RAN elements and resources via fine-grained data collection and actions over the E2 interface. Hence, near-RT RIC plays the most important role in enabling intelligence as part of the RAN.

O-RAN Software Community

The O-RAN Software Community (SC) is a collaboration between the O-RAN Alliance and Linux Foundation with the mission to support the creation of software for the Radio Access Network (RAN). (See Figure 2.) O-RAN SC released its third official release on December 12, 2020, announcing the Cherry release and the current ongoing release planned for the second week of June 2021.

Figure 2. The O-RAN SC architecture

Source: ORAN- SC Confluence²

SD-RAN

The SD-RAN Open Source Alliance for O-RAN implementation, led by the Open Networking Forum (ONF), provides various renowned open-source networking initiatives such as ONOS, AETHER COMAC, P4, CORD, etc. On January 25, 2021, ONF released its first SD-RAN release. SD-RAN v1.0 is a cloud-native exemplar platform for software-defined RAN based on ORAN specifications. Figure 3 shows the high-level SD-RAN software architecture based on ONF open-source components like µ-ONOS, Aether, and others.

Figure 3. The SD-RAN architecture

Source: ONF SD-RAN³

Comparison of O-RAN SC and SD-RAN

This section compares the O-RAN SC and SD-RAN platforms based on eight key aspects: (See Figure 4.)

1. Contributor/community members
2. Activities and source code contributions
3. Key feature details
4. Compliance with O-RAN standards
5. License type
6. Membership cost
7. Flexibility
8. Roadmap

Figure 4. Comparison of SD-RAN and O-RAN SC

#	Key aspects	SD-RAN	O-RAN SC
1	Contributor/ community members	Aarna NetworksAirHop CommunicationsAT&TChina MobileChina UnicomCohere TechnologiesDeutsche TelekomFacebookGoogleIntelNTTParallel WirelessRadisysSercomm	AT&TChina MobileDeutsche TelekomEricssonNTT DOCOMONokiaOrangeRadisysVerizonHCLViaviSamsungThe O-RAN Alliance for specifications has 27 operators as members and 221 companies as contributors, of which ONF is one of the contributors.
2	Activities and source code contributions	No details available	381 contributors3.95 K Commits3.09 M lines of code1.96 K total issues reportedWeekly meetings every Wednesday
3	Key feature details	SD-RAN v1.0 was first released to members on January 25, 2021SD-RAN software is based on ONOS and Aether ONF projectsNear-RT RIC is based on µONOSProvides app-SDK for creating xApps that are portable across different near-RT RIC platformsSD-RAN has shared its app-SDK code with ORAN SC; hence the xApp SDK in ORAN SC is the same.SD-RAN offers RAN-in-a-Box (RiaB) deployment option.The entire distribution, including xApp, µONOS-RIC components, CU/DU, UE and OMEC, can be instantiated within a VM/Server RiaB with just a few simple commands.Provides E2 Agent that exposes the KPM-SM IE for xAppsLikely to expand on the E2 interface to allow for scheduler control and network slicing and contribute this expansion back to O-RAN for inclusion in the specificationsO-DU, O-CU-UP and OCU-CP based on open source OpenAirInterface5G enhanced with ORAN-compliant interfaces and protocols.Aether5 used as the Virtualization Layer, VIM and Infrastructure Management Framework.Leverages both COTS and white box P4-programmable switches and plans to implement O-CU-UP using P4Additionally, RIC includes Topology Manager, Configuration Manager, RAN Control Manager, and a Distributed Storage Manager.	The latest release was the Cherry on December 15, 2020The next official release, i.e., D release is expected in mid-June 2021.Along with other components, the following xApps are planned:Traffic steeringAnomaly detectionLoad predictionHello World demo xApps in C++, Go, PythonBouncer xApp for RIC performance benchmarkingSecurity xApp – Signaling Storm ProtectionxApp portability SDK (ONF)Active development on the following components:O-CUO-DU HighO-DU LowSMO, Non-RT RIC aligned with ONAPRICAPPO-RU emulatorNear-RT RICOAM O1 InterfaceE2 SimulatorINF – HA
4	Compliance with ORAN standards	ORAN specification compliant for A1, O1, O2, F1, E1, E2 included as part of release roadmaps	ORAN specification compliant for A1, O1, O2, F1, E1, E2 included as part of release roadmaps
5	License type	ONF member-only software license, i.e., also two types:Limited license to all membersAdditional license to partner-level members	Apache 2.0 license
6	Membership cost	Dependent on the company’s turnover.Limited license to all members: 50,000 USD annuallyPartner-level members: 500,000 USD annually	NA
7	Flexibility	Does not permit any member, regardless of member level, to redistribute all or any portion of the works, except as part of a contribution back to the applicable project.	Highly flexible. Complete code access available online. Code can be modified and freely distributed.
8	Roadmap	Handover KPI and PCI xAppsUpgrade RAN SIM for RC-PRE/KPM/ other SMsSDK EnhancementsTopology managerO1 support	ORAN SC D-Release roadmap.TS xApp, QP driver xApp, QP xAppKPIMON xAppE2 Simulator enhancementsOpen Test FrameworkO-RU failure and recoverySupport for CLA is proposedAlong with various enhancements as well as test and stabilization of non-RT RIC, near-RT RIC, ODU, OCU, SMO

Source: Compiled by Capgemini Engineering

Conclusion and recommendations

There is a set of guiding principles for vendors to make informed decisions when selecting an open-source O-RAN implementation. Based on the details discussed in this blog, here are the key issues to consider.

In my opinion, the SD-RAN near-RT RIC platform release 2.0 is production-grade software for quick and easy trial by companies developing intelligent xApps. It provides a rich xApp SDK and an E2 simulator capable of simulating a number of RAN scenarios readily available for vendors planning to develop and test xApps with the RIC platform. In this use case, a vendor only needs to be an ONF member (not a partner member) and pay annual membership fees of the order of 50,000 USD.

In my discussions with vendors who have recently joined as members, it was evident that most of them have no plan to offer a complete RIC solution and have no RAN offerings. They are going with SD-RAN to focus primarily on xApp development using the SD-RAN and no plans to offer RIC solutions.

If vendors plan to offer RIC solutions, then O-RAN SC, backed by the Linux Foundation, is a better choice. One characteristic common to most vendors is that they come with the baggage of different 4G/5G RAN offerings. A simulator is not an essential requirement for them. Instead, they are looking for an open, robust platform that provides guiding principles for RIC design and not a ready set of tools and simulators just to develop, run, and test their apps. Also, the hefty Partner, license subscription, price tag for SD-RAN of 500K USD a year is another reason why many vendors may choose O-RAN SC.

Another interesting observation I made in my discussions with a few xApp vendors is that they are still looking for commercial RAN vendors to verify their xApps in a real end-to-end RAN lab. This essentially indicates simulators can also be used to help in testing xApp to a certain extent. As part of these discussions, it was evident that all are waiting for real customers before planning to become an SD-RAN partner and henceforth a RIC reseller as otherwise, it’s quite an expensive affair given the hefty annual price tag.

To conclude, SD-RAN is a better option for vendors planning only to offer xApps. In contrast, NEP and vendors with RAN offerings would prefer the ORAN-SC as the RIC platform for offering complete RIC solutions along with xApps and rApps.

---

¹ Sunay, Oguz Das, Saurav “SD-RAN Techinar”, Jan. 9., 2021, SD-RAN

² Skorupski, Martin “OAM Architecture,” Jul. 14, 2020, Atlassian Confluence Open Source Project

³ Thome, Joe Modlin, Cory Das, Saurav “SD-RAN v1.1. Techinar,” Apr. 28, 2021, Open Networking Foundation

---

Author

CHOPRA Subhash  Solution Leader 5G ORAN, Network AI and Industry 4.0Capgemini Engineering  Subhash Chopra is part of the Capgemini Engineering Innovation Team, focusing on 5G ORAN RIC and Industry 4.0 solutions. His responsibilities include incubating new solutions, developing proof-of-concepts, developing a roadmap for productization of solutions, service offering definitions and technology consulting. Subhash has 23 years of professional experience in the product engineering services and telecommunications industry with a focus on technology and innovation. He has deep experience in solution architecture development, system integration and technical consulting for complex telecommunication networks and related services. At Capgemini Engineering, Subhash has incubated several cutting-edge automation platforms that created immense value for our customers. His most recent project is Ratio, an innovative 5G ORAN RIC framework, another incubated network AI framework. NetAnticipate5G is a highly scalable, intent-based, artificial-intelligence (AI) driven network automation platform for realizing zero-human touch 5G network operation. The framework will be unveiled in MWC ’21 in Barcelona and has already been evaluated by telecommunications customers. NetAnticipate5G has earned awards, including AIconics ’18 and NetTA ’18. Subhash’s current area of interest includes advanced machine learning and AI for next-generation networks like 5G, SDN, NFV, SD-WAN and IoT, to realize zero-touch network automation. Subhash is a speaker at international conferences and has presented technical papers in various forums. He was a keynote speaker for “Advancing Network Operations and Reliability with the Integration of AI” at the Telco AI Summit in 2019. He is currently based in New Delhi, India.