Mavenir & Northeastern University Partner to Offer Open RAN Simulator
Why did Mavenir partner with Northeastern University?
Mavenir was excited to work with Northeastern University largely due to their wireless IoT initiative called Colosseum, which has focused on the Open RAN compliant network simulator. Mavenir is engaging with other universities as well globally. Many of these universities are working on similar applications and concepts as Mavenir and expanding the opportunities for collaboration remains a top priority for Mavenir.
What does the partnership with Northeastern University mean for developers?
Partnerships like this are beneficial for third party developers who want to build applications and test them against Open RAN compliant or non-compliant networks. Mavenir is helping to empower universities and third-party developers to accelerate the development of Open RAN applications, feeding the ecosystem for more innovation.
Can you help explain the Open-Source Open RAN Simulation Platform?
With closed systems, all the intelligence remains within the vendor solution, which means relying on that single vendor to provide that feature. The idea is that you can use the ecosystem, have open innovation, give more control to the operator, do RAN automation, proactive management and then differentiate services.
There’s some confusion in the industry as to whether the Open RAN Intelligent Controller (RIC) is a different version of SON. The answer is no. There are two variants of the RIC:
- Non real time (Non-RT) RIC—which can be related to an evolved version of SON (self-optimized network that has existed for a long time in radio access networks but is enhanced using AI/ML mechanisms).
- Near real time (Near-RT) RIC—which is where the intelligence to tune the RAN comes in on the order of tens of milliseconds. There is no existing parallel to SON. The near-real time RIC can manage fine-tuned network control (such as controlling control features on a per user level). It permits customized per UE level optimizations within a cell, such as maximize UE1 spectral efficiency, minimize UE2 energy consumption, minimize UE3 call drops etc. This is all controlled by using third party applications, or applications that can be built by Mavenir. RIC provides a platform for healthy competition from different innovators to support RAN optimization.
The way the ecosystem is currently built, there are players who are working on just building applications (application players) and then there are RIC platform players, who are building platforms on top of which RIC applications can run. All of them work in the ecosystem with the Open RAN players.
Mavenir plays a part in all of this by providing the RAN software, the RIC platform and building applications, whilst also working with the ecosystem where people can innovate and build their own applications.
What does the Open RAN simulation platform do? What are its benefits?
One of the biggest issues with maximizing the efficiency of applications today is a lack of testing for 3rd party RAN optimization software. If someone wants to integrate into a traditional RAN vendor baseband software, it is difficult to do so, because traditional RAN vendor systems are only pitching non-real time RIC, which has a different interface, and only exercises the slower, cell level aggregate optimizations. Rakuten Mobile, with their recent Open Innovation Lab announcement, seems to be doing some parts of it, but there is not a clear view as to exactly which parts they are exposing.
Mavenir has focused on the near real-time RIC because that’s where the intelligence really comes in and where the integration with the RAN happens. Within the near-real time RIC, it is possible to exercise radio resource management level control to adjust resources per connection based on the operator’s desired objective functions for that user application. There is a tool called NS-3, which existed for a long time, and is a functional end-to-end network simulator. What Mavenir did was to build on top of the NS-3 open network simulator to develop an ns-O-RAN simulator, which effectively simulates the E2 interface as well.
Now, universities, or any third party, that are developing applications can test them against a real-world simulation before testing with a vendor. Internally, Mavenir tests with the ns-O-RAN simulator first, and then tests with the actual NF software, the CU/DU. This approach accelerates defect identification, issue resolution and effectively allows the ecosystem to develop more RIC applications faster.