The Heavy Lifting of O-RAN Standardization

When it comes to potential debate of whether a standards body provides enough utility to positively shape an entire industry, a cooperative, patient, and innovative group of contributors is often a convincing recipe for success.

Leadership in driving standards work is no easy task.  The O-RAN Alliance, with its vision of transforming Radio Access Networks (RAN) toward more open, intelligent, virtualized and interoperable architectures, is perhaps the most transformative effort the world has seen in new standards development. For Mavenir, active leadership on the ground floor of an emerging technology has been both rewarding and massive in total effort.

The Stakes are High

The O-RAN Workgroup 3 (WG3) is focused on defining an architecture that is centered on the Near-Real-Time RAN Intelligent Controller (RIC), which enables near-real-time control and optimization of RAN elements and resources via fine-grained data collection and actions over the defines E2 interface.

The RIC enables Communication Service Providers (CSPs) to take advantage of an open platform with industry-leading Artificial Intelligence (AI) and Machine Learning (ML) techniques. The RIC is resource-efficient, carrier-grade, and cloud-native while supporting use cases for existing 4G and future 5G networks.

An open, optimized, and automated network featuring the RIC is necessary to break legacy vendor lock-in while promoting innovation for new use cases.  In existing networks, the use of newly developed RIC algorithms can yield increases in over-the-air performance for network operators. Traffic steering, available today, can significantly improve QoE in 4G and 5G networks to enhance the quality of experience for end subscribers.

Recently the O-RAN WG3 had approved the E2SM-RC (E2 service model for RAN “Control”) specification version 1.0 by a consensus-based voice, which is a critical milestone for the E2 v2.0 package release. E2SM-RC v1.0 was a huge effort, with the spec running to 250+ pages, generated over a 4-month span, and all starting from scratch. Mavenir has taken a lead role in the WG3, ensuring the adoption and acceptance of the standard that enables the near-RT RIC to control the RAN.  This work ensures the continued policy and strategy of the O-RAN Alliance’s open networking with open interfaces across the vendor ecosystem.

The Big Lift

In its initial form the E2SM-RC specification had five key ingredients:

  • The Insert Service: Defines how can the RAN “request” the RIC to control a given RAN functionality on a per-UE level/bearer-level?
  • The Control Service: Defines how the RIC can “control” the RAN for a given RAN functionality on a per-UE level/bearer-level?
  • The Report Service: Defines how the RAN can “report” state and context information to the RIC on a per-UE level/bearer-level (in addition to cell-level and node-level)?
  • The Policy Service: Defines how the RIC can “prescriptively control” the RAN for a given RAN functionality on a per-UE level/bearer-level?
  • The Event Trigger: Describes how to define the “events” in the RAN and how the RIC may subscribe to such events in order to execute the above services.

All these services are meant to control the following functionalities in the first release:

  • Radio Bearer Control
  • Radio Resource Allocation Control
  • Connected Mode Mobility Control
  • Radio Access Control
  • Carrier Aggregation Control
  • Idle Mode Mobility Control (predominantly for traffic steering and QoS/QoE use-cases)

The E2SM-RC drafting team was comprised of four companies due to the massive effort of the standard. The E2SM-RC drafting specification (including the parts drafted by other companies) was primarily based on Mavenir’s “parameter and message encoding contribution” that was adopted in the specification. This allows the RIC to control any parameter in the RAN in any level of hierarchical nested 3GPP structures using a “generic encoding scheme,” instead of having to enumerate individual message structures containing individual parameters and categorizing them all as M or O or C-, as done in 3GPP.

Mavenir’s proposal categorized every RAN parameter as either an ELEMENT (a singular RAN parameter) or a STRUCTURE (further containing a sequence of RAN parameters) or a LIST (containing a list of itemized structures) and the message between the RAN and the RIC can encode any parameter based on this categorization, while using a simple structure (again, defined by Mavenir). Mavenir also identified which RAN parameters would need to be controlled by the RIC for the above functions.

The Future is Bright

This contribution was crucial for finalization of a full-blown E2SM-RC specification that controls all the above functionalities.  Without it, O-RAN WG3 would not have been able to meet all the defined requirements for Traffic Steering and QoS/QoE use-cases. Mavenir’s contribution was not limited to these 2 use-cases alone. Any O-RAN use-case (including slicing, V2X, DU scheduler, MIMO beamforming, etc.) could employ the same technique.

One of the leadership challenges in any standards development is bringing together a polarized group of members to a unifying vote on a proposal.  Mavenir helped WG3 steer clear of this obstacle, as its contributions were strongly supported by service providers, such as Verizon Wireless and Telecom Italia, as well as vendors.

As O-RAN standards development continues well into the future, the effort currently maintains significant tail winds to make concrete, disruptive, and positive impacts for CSPs, the vendor ecosystem, and consumers at large – it’s a WIN-WIN-WIN!

Read more about Mavenir’s RIC leadership.

Dark Mode