Almost every operator sooner or later claims to Open RAN implementation. This is how the participants of the RCR forum answered the questions.
Many players in the mobile market (suppliers, operators, integrators) are delighted with the Open RAN technology. And rightly so.
What, in practice, will make Open RAN networks more meaningful than traditional networks – let’s call them Single RAN here?
The question is whether customers know, what can be gained thanks to Open RAN and whether there is any gain.
Or due to risk there is more to lose than gain and it is better to stay with Single RAN?
Most operators carefully observe and listen to signals from the market adapters such as Vodafone, Jio, or Rakuten.
How does the mobile network market see the future of Open RAN?
I came across a very interesting survey conducted by RCR Wireless News on its forum. RCR Wireless News survey was conducted during the Open RAN European Forum among 100 respondents.
Its authors asked only five simple questions, looking for answers to the potential interest of buyers of Open RAN products.
Among the five questions, there is a particularly important one, question 3.
Question 3: What is, or will be, the main driver for you to deploy Open RAN?
I decided in the New Year to share with you my thoughts on the question asked, allowing myself to comment. Comments based on experience from several Open RAN tenders over the last two years.
I had the opportunity to personally assess how this market looks today from the inside, tendering the vendors.
Let’s discuss these answers.
Cost Reduction. Open RAN implementation.
What amazes me is that most of the answers I believe are based on “heard” opinions and not on facts.
Why do I say that? This is a reflection drawn after analyzing a few tenders. I included also the opinions of CTOs from networks that have already started Open RAN implementation.
What is the cost of Open RAN implementation in the mobile network? Will it be cheaper or maybe even simpler?
Respondents, probably following the claims of CTO Rakuten Mobile, count on reducing the cost of implementing radio networks by 20%-30%. Compared to the Single RAN network. This thesis was put forward by Tareq Amin on the TIP forum in 2019. Right next to reducing the cost of network maintenance and the cost and time of deploying base stations.
It’s probably no secret now that it’s not a matter of buying a cheaper system. Tareq is right but in a long-term settlement. In a recent interview, this was also confirmed by Yago Tenorio, Network Architect of Vodafone.
I will add something about this.
What myths circulate about the lower cost of Open RAN implementation in the mobile network?
- Tenders for 4G and 5G radio infrastructure do not confirm this thesis either in terms of CAPEX investments or OPEX maintenance costs. There is no “golden shot”. There is no magical lower price and golden mean for a good morning. It is quite the opposite. Open RAN providers can surprise the customer with a set of fees. Fees related to the adaptation to the Single RAN network and, above all, the high costs of its Open RAN development (SW, HW roadmap).
- Many times, solutions that have not yet been developed to the level of Single RAN suppliers require the involvement of a lot of time by operators in learning and “arming” the system. As a rule, Open RAN providers do not yet offer operational procedures on how to use Open RAN solutions in practice (with exceptions, but I will not advertise the providers here, let them promote themselves). Offers often contain multiple components from different partners. I’ve seen vendor offers that included 2-3 different management systems with different FCAPS methodologies. Such a product imposes a much greater obligation on the shoulders of operators. The costs of training the operator’s staff are increasing. In addition, it is often not easy to find among the operator’s employees proficient in virtual, container technologies or programming and knowledge of DevOps environments. Modern solutions without a coherent operational management system (DevOps or AI Ops) must be mastered by enclosing them with systems introducing elements of automation. Today, not every operator’s Network Maintenance team has an efficient branch of employees for such tasks that go beyond the daily network maintenance tasks.
- In addition, some operators may wish to combine several not-so-well-integrated Open RAN providers. (O-RU radio from another supplier, O-DU/O-CU from another supplier). It delivers additional, really time-consuming, and expensive tasks in a long-term project.
Why Open RAN may ultimately be cheaper than Single RAN?
These are just a few facts that torpedo the opinion of lower Open RAN costs today.
Systems are more expensive, highly complex, and require investment in new functions and operator organization teams.
Those who have already passed the first attempt to build an Open RAN network know that it is about something else.
Dedicated algorithms will reduce the number of physical resources. The number of base stations, the number of bands served by the station (expensive band), or the transmitting and receiving power, thanks to clever and efficient algorithms.
When an operator buys a Single RAN base station today, it fully depends on the supplier’s idea of how the radio interfaces are to work, of course, based on 3GPP standards.
Returning to the promise of reducing costs by 20-30%, we return to this point.
Open RAN implementation in a similar area to Single RAN, but with well-optimized RRM algorithms based on the RIC system, can reduce the number of resources needed by 20-30%.
The conclusion is that these savings must be earned.
The question is whether the operators will be willing and able to build the appropriate competencies and resources to operate such a network.
One solution may be automation, which I will write about later. One of the leading suppliers of Open RAN employed over 1,000 people to prepare appropriate products in this area for two years.
It’s not easy, is it? It won’t be cheap at first either, but it will last.
What are the other advantages of optimizing radio algorithms besides managing radio resources?
The Open RAN forum has prepared documentation describing many other cases of optimization of various areas of radio network operation:
- Lower electricity consumption (10%-15% according to estimates)
- Better quality thanks to QoS (Quality of Service) algorithms operating, for example, directly in a district with a high concentration of residents and heavy traffic at rush hour.
- Optimization of cooperation between the RAN radio network and the CORE network by changing the RIC-MEC algorithms. For example: directing video streaming traffic to a closer, more efficient MEC (stream) server. This reduces transmission costs and improves QoE for the client.
- Etc. There is no limit to possibilities because everything that happens in the radio network is based on some solutions. Many areas can be corrected or modified. Creativity in this area is probably one of the greatest advantages of the openness of Open RAN and 5G 3GPP systems because the same philosophy also guides the new 3GPP standards.
With an Open RAN provider, you can multiply the number of changes you can influence.
Remember, however, that this will not come by itself and it will not come for free. This requires investing in a new generation of radio network planners and algorithm programmers to support them.
This is the cost of changing the organization and the way operators operate, but it is a cost that will pay off.
Can Automation Simplify Open RAN Deployment?
Automatically solving repetitive problems on the network, as close to the customer as possible, is a job for “robots”.
Creating processes that will regularly examine the “nuances” of network operation, depending on environmental variables such as weather, subscriber traffic, and local phenomena (accidents, etc.) will allow you to control the quality of the network.
This, of course, means an increase in the number of tasks performed concerning the existing staff in the network maintenance departments.
From practice, the operator will not spend time on improving the network at the level of one base station daily. The number of employees for problems that cannot be seen with the naked eye is not enough.
Open RAN has the advantage that it allows you to build quality-conscious processes deep within the resources of the radio network. RIC systems have supervision over all levels of air interface resources. Through Non-Real Time, Near Real-Time, and O-DU loops, it can analyze and modify radio network activities in milliseconds or longer. I guess I have to finally write an article about RIC, right?
In my opinion, this is a huge advantage that will translate into lower network costs in the future.
You have to roll up your sleeves and learn it all. Do you agree with me?
The operator could build a team to deal with it today, it would be a great start for “smooth” adaptations to the Open RAN technology on a massive scale.
We now know what an operator should look for in the Open RAN world in terms of cost savings.
Gaining practical influence and working in favor, which is within the reach of every player in this market.
Could Single RAN providers modify RRM algorithms and what are the risks behind the self-deployed algorithms?
Now go back to the original question and try to get the Single RAN provider to change the RRM algorithms for your needs.
Is there a chance to get such a product and how much will it cost you?
I’ll give you a hint. No chance and not worth it.
As I wrote, anyone can create new algorithms.
However, I must now add a very important caveat.
While anyone can influence the changes in the algorithm, doing it well and optimally is no longer a trivial matter. Experienced scientists and planners work on good RIC algorithms.
This causes me to fear that the market for RIC algorithms will narrow down to a few vendors in a few years.
Wish I was wrong.
Even Microsoft realized how big a difference it makes and how lucrative this market is when it invested in a team, building RIC solutions in its organization.
I will now move on to the next answer to the third question.
“What is or will be driving the desire to implement Open RAN?” provoked the second largest response in the group:
Vendor Diversity – Variety of Suppliers.
We can guess that the respondents wanted to change the influence and management of the current Single RAN providers.
Question 4 was, Do you foresee Open RAN being deployed for brownfield, greenfield, or both?
This is a strategic and cost aspect.
When we have one supplier, and this one starts to raise prices over time or, even worse, not provide the network or products, the operator naturally looks for “someone better”. The second or third.
This is how the market works, although the issue of supplier switching (SWAP) in a city like London or Munich is a huge project, it is often done this way.
Looking at tenders for large commercial networks, I noticed that Open RAN providers fall into the same pattern of operators thinking and acting.
They become potential network providers who can provide similar technology at a lower price as part of replacing the existing supplier or adding a new band, for example, 5G.
And here begin much more serious challenges for today’s devastating Open RAN providers.
First, Open RAN providers do not have the technology to provide a similar technological base as Single RAN providers. Open RAN networks are developing in the area of 5G and 4G. Few Open RAN providers can provide 2G and 3G networks (these are rather marginal solutions, but still exist on the market).
It is therefore impossible to replace the existing “recalcitrant” supplier with the Open RAN supplier. We need to change our networking strategy. For example, turn off 2G and 3G networks by refarming fully to 4G and 5G technologies. I wrote about the consequences of such action in the article about VoLTE Roaming.
Vendor Management in the Open RAN implementation process.
Secondly, the issue of Open RAN interworking between providers is far from perfect, and even from the status of “good”. It relates to strategic and technical issues.
Defining the Open RAN standard at the split 7.2 level causes a huge number of problems. Especially for the more complex Massive MIMO or mmWave technologies.
Of course, you can choose a provider that performed most of the testing and offers working interfaces. But, in this situation, we choose the lesser evil, i.e. the Open RAN provider from the Single RAN stable.
In addition, a huge obstacle is an inability to combine Single RAN and Open RAN providers in the “pepper and salt” model. I won’t get into this topic here, but I will just mention that the X2 interface invented by 3GPP to connect base stations from different vendors is not compatible even with Single RAN vendors.
A separate article is under preparation on this topic. A real firecracker.
Has anyone ever tried to persuade two suppliers to change their interface (even compliant with 3GPP standards, etc.)? These know that if there is no will, there will be no will and solution.
Without an interface (read cooperation) nothing will work. I always say that:
90% of the operation of telecommunication networks is based on interfaces.
I know that even the largest operators are not able to persuade suppliers to submit.
If it were otherwise, they would not have formed an alliance for Open RAN (Vodafone, Orange, T-Mobile, Telefonica).
Open RAN can be a new set of similar suppliers.
When we decide to choose an Open RAN provider, we load ourselves into another SET of providers. As I described in the first chapter, it is a matter of technology immaturity and the operation of the Open RAN cooperative. Society by definition assumes the cooperation of many smaller partners.
Bizarrely, operators try entering Open RAN as an escape from Single RAN providers. As a result, the operator must take on his shoulders a large portion of coordination and project management.
The operator must start acting as an integrator or even partially a radio solution provider.
Does this make it possible to think of Open RAN as an antidote to Single RAN providers?
In my opinion, when we look only from this perspective, the operator would fall “from the rain into the gutter”.
What do you think?
Now let’s discuss the third answer to question 3.
The respondents referred to the possibilities of RIC (Ran Intelligence Controller) solutions and the choice of xApp application providers.
The theory is that the RIC platform will allow any xApp application to be installed and deployed on the network.
In theory, we will get immense flexibility in expanding the network.
I will refer again to practice, this time practice in working with teams of IT programmers.
Once an operator purchases the Open RAN platform with the RIC solution, it will include development and support.
The RIC provider will have the operator in hand (similar to the selection of radio stations and the interworking).
Why is this seen as a problem when we have already bought a great RIC platform?
As long as everything continues to be delivered from the same supplier, it will be fine. They will keep operational processes and control the purchasing point of view.
When we want to implement a “foreign” xApp application on the RIC platform, we are faced with the onboarding process.
No matter if it will be a commercial application/algorithm or modeled by the operator’s team.
What does the xApps onboarding process involve?
What ensures compatibility at the level of Open RAN standards?
For example, a xApp provider (3rd party) may want to use methods (even Open RAN standard) than the RIC provider. And what will the RIC platform provider say then?
There is the first scratch.
Someone has to do something, modify, add code, do tests and finally take responsibility for quality and efficiency.
Both providers will have to operate in a similar DevOps environment, i.e. application development and maintenance processes. It can be done because this is how the entire IT industry works. Unfortunately, in many cases, the capabilities of telecommunications providers and operators are still quite modest in this area.
I participated in many hours of talks between the suppliers of the RIC and MEC platforms. They set a whole list of requirements and reservations already at the platform selection stage. Before they even sold the RIC.
Do you know what I’m writing about? Do you know it from practice?
Managing the xApps in the Open RAN implementation process.
The process of cooperation between the xApps provider and RIC will require efficient coordination. This can be performed by the operator or trusted integrator, which entails costs.
This is important to consider on the operator’s side.
Such action requires building new competencies or developing existing competencies in application development. An efficient team of business analysts and application integrators.
The operator must enter the booths of the integrator and the developer (supplier) at the same time.
Do you know mobile operators who have similar competencies developed in practice?
I know few and I know that the Core Competence of operators is the provision of mobile services and maintenance. Not the development of algorithms and platforms.
A big thesis to think about.
Of course, flexibility cannot be viewed only from a negative perspective.
Many suppliers of IT products are on the market, such as Zabbix. They can integrate with everyone and everything, even the cash register in the supermarket. Zabbix opted for such a strategy. They have full control and supervision over their product.
Zabbix built competencies and created optimal processes as well as employed and integrated a team of people.
To achieve this flexibility, operators must follow a similar path alone or with a partner/integrator.
These conclusions also apply to the fourth category of answers to the third question.
Easy/Seamless Deployment. Open RAN implementation.
Such a state cannot be achieved without meeting all the above points.
I would rather call the Open RAN implementation a goal to achieve.
There are necessary changes in processes and approaches to technology strategy than functionality to buy.
There is no golden mean without hard and strenuous work and changing the way the mobile operator. Operators shall follow the examples of the IT world.
The Easy & seamless Open RAN implementation is the goal of a longer process, not a product.
Is there any shortcut?
I can give you a shortcut to achieve this goal.
It is necessary to focus on a supplier who understood all this perfectly. Better to select those already heaving practical methodology and solutions implemented today.
You just need to bet on a good partner.
However, this strategy requires efficient management in the field of Vendor Management and purchasing.
However, is it not again going into the Single RAN vendor “strategy?
Thanks for reading and I encourage you to comment.
Shortly I will write about building private networks.