In March of 2021, Tareq Amin, then CTO of Rakuten Mobile and now CEO of Aramco Digital, and I contemplated the question of how Rakuten Mobile could ramp itself onto a “brownfield” operator’s modernization roadmap and drive Open RAN adoption. Of course, this would have to be done while minimizing risk and cost to the operator and fostering interoperability and observability across an evolving RAN technology and vendor portfolio.
It is a question that continues to vex Open RAN vendors today.
I mentioned the necessity of middleware to position Open RAN into brownfield networks. Tareq agreed and cited his intent to interface RCP (Rakuten Communication Platform) with legacy RAN.
“Nice!” I exclaimed.
Toward the end of our discussion, Tareq pointed to a box in the RCP conceptual architecture diagram labeled the “Multi-Domain Orchestrator & Automation Platform.” In concept, it looked a lot like what we call SMO, or Service Management & Orchestration, but with a twist.
The incumbents got the SMO memo
SMO is emerging as an important integration tool and interoperability enabler that operators need to accelerate their infrastructure modernization and their move toward O-RAN compliant or ready networks. Originally, it was specified to foster the orchestration and management of services across multi-vendor ORAN environments, then the incumbents flipped the script.
Ericsson’s early toe-dipping into Open RAN started with the introduction of SMO-like platforms before the company officially belly flopped in with an industry-shocking splash at Fuyz 2023.
In December of 2021, Ericsson introduced its Ericsson Intelligent Automation Platform (EIAP) capable of integration with Ericsson’s purpose-built RAN as well as O-RAN-compliant RANs. Nokia followed three months later with their Intelligent RAN Operations solution. Not to be left out, Huawei introduced its IntelligentRAN Wireless Architecture built around their Mobile Intelligent Engine (MIE) at MWC 2022.
Since, several players have thrown their hats into the SMO/RIC ring, including Juniper Networks, Mavenir and eventually Rakuten Symphony, as excitement about rApps and xApps washed over the Open RAN community. However, that excitement and anticipation was met by a paucity and modest scale of Open RAN deployments.
To wait or not to wait for open interfaces
Late last year, Vodafone and DOCOMO published the findings of their joint study exploring the challenges of SMO deployment in Open RAN environments. The study suggested that open standards for SMO interfaces may take time to be adopted among RAN vendors.
The study also concluded that “greater standardization of the interfaces between existing parts of the network and the newly created Service Management and Orchestration platform” was necessary for intelligent automation. It seems like the incumbents had been on to something.
Today, a growing number of operators are taking an earnest look at blending Open RAN principles, architectures and vendors into their 5G modernization strategies and roadmaps. But should they wait for standards to get things rolling?
Integration and interoperability across a diverse and evolving portfolio of proprietary and O-RAN compliant/ready RAN assets and technologies are becoming recognized as the first order value of SMO among operators. They are also recognized as prerequisites to realizing cloud-native RAN and rApp/xApp promises of intent-driven autonomous networks and, of course, 5G monetization.
Consequently, the proposition of SMO is evolving from its ORAN-only genesis toward a multi-tech and multi-vendor regime addressing the interoperability and integration challenges of the “brownfield.”
Incumbents like Ericsson and Nokia who entered the SMO game touted native interfaces for their purpose-built systems and a promise to provide O-RAN compliant/ready adapters as the standard settled. They also offered SDKs for the development of interfaces for everything else.
Is this enough? There is a lot of legacy and diversity in the brownfield. System integration is costly to boot.
An integration first approach to SMO
On June 13, 2022, Qualcomm announced that it acquired Cellwize Wireless Technologies Inc. without much fanfare. It seemed a bit off script at the time for Qualcomm. Why would they buy a SMO platform company?
Durga Malladi, who was SVP & GM of cellular modems and infrastructure at Qualcomm at the time of the acquisition, described Cellwize as an accelerator for Open RAN adoption as well as a “best-in-class RAN automation” platform for the next generation of open, cloud-native mobile networks.
Since 2013, Cellwize cut its teeth on SON developing backbones and tools for integrating multiple vendor RANs under a common management and orchestration plane. As the O-RAN Alliance introduced the RIC, Edgewise became an early rApp developer demonstrating how a SMO could make open and purpose-built RANs programmable and manageable under a common orchestration framework.
The tools and integration assets that Cellwize developed over the years ultimately came together in the SMO platform that is the Qualcomm Edgewise Suite that we know today.
According to Ofir Zemer, VP of Product at Qualcomm, Edgewise incorporates what he calls the Data Mediation Layer (DML) that provides a vendor-agnostic abstraction that serves as an integration hub with various vendor EMS systems. To close the loop, the Provisioning Gateway takes actions from the Edgewise SMO and translates them into native instructions for proprietary vendor systems.
In a way, Edgewise is by its Cellwize SON and system integration legacy pioneering the truly “open” SMO (Open SMO) for orchestrating, automating and optimizing across a heterogenous and multi-vendor RAN portfolio. Edgewise does this by providing operators and system integrators with an expansive library of connectors for accelerating legacy and Open RAN system integration.
The race to open SMO
Interoperability and integration are requisite to achieving the observability operators needed to begin their journey toward intelligent automation. However, waiting for “the greater standardization of interfaces with existing part of the network” will not drive the adoption of SMO much less accelerate Open RAN adoption.
The sobering reality is that most if not all operators will not be flipping a switch and suddenly transition to a completely Open RAN network. Network modernization has been and will be long, complex and tough transformations for operators. Operators will also be looking for the tools and the capabilities to enable and accelerate ROI along the way.
Ironically, an Open SMO that is not siloed by O-RAN compliance and dogma can bring the promises and benefits of O-RAN SMO to operators by accelerating interoperability, observability and management across brownfield portfolios. It can also ease the adoption of Open RAN along while the Open RAN movement and operators committed to it wait for greater standardization of interfaces.
That’s always been the beauty of middleware.
