Point of View: Interview with Nati Gabizon, EVP Head of R&D at Parallel Wireless

“Welcome to our new interview series where we sit down with industry experts and leaders to discuss the latest developments and trends in their respective fields. This time around we spoke with Nati Gabizon, a veteran of wireless technologies and EVP Head of R&D at Parallel Wireless. Nati shared his thoughts about the future of mobile networks, the adoption of 5G and what it will take to get us there.”

Where do you see the telecommunications industry going in the next 5 years?

One of the primary long term goals, the holy grail of the market, if you will, is 5G adoption and the transition to smarter networks, networks that are controlled by AI to eventually provide a better user experience with lower power consumption. For example, traffic steering to improve the user’s quality of service and temporarily powering down layers when they are not needed.

A key challenge of 5G is power consumption; when 5G reaches its full potential, the energy needs and costs will be significant. This addition, on top of the actual capital and operational costs of 5G, may torpedo a business plan and cause operators to delay their 5G implementation. In parallel, in most Western countries the regulatory pressures are for telecoms to reduce their carbon footprint. While the technology is making higher demands for power, other forces are driving lower power consumption and in the middle is a user with a phone who is hoping for a better user experience.

This means that the decisions need to be smarter, everything in the network will be controlled by better decision-making systems, which will predict user behavior and modify the conditions to secure smooth service.

I also expect to see a change in the way a network is deployed; the network is still being deployed according to “old school” rules. Simply adding autonomous sites is not scalable; each site is capable of handling a smaller number of users, so the cost of each additional piece of equipment is more expensive per subscriber. This clearly doesn’t make sense in dense, urban areas and it’s even true for less dense areas.

We are going to see the infrastructure improve, fibers will be more readily available, as will high bandwidth, wireless solutions to drive front-haul and mid-hau l interfaces. The “old school rules” will no longer be viable. There is just too much to gain from aggregating the resources and creating smart models based on clusters of dozens of sites. This is why we have focused on innovations such as our RAN Centralization and our RIC (RAN Intelligent Controller) and others. There aren’t many vendors that have all these components in place to support mobile operators in the mid-to-long term.

I am happy to say that from my perspective PW is well positioned for the future; I believe we made the right decisions along the way and now it is just a matter of executing and making the small adaptations that the market will require.

Is there anything that vendors can provide to assist operators as they transition from the “old school” to the new approach?

The reality is, that, as a whole, most operators are risk-averse, so they are not eager to adopt new approaches. Now, if you look more closely, they are not homogenous. There are those, typically on the technology side, who are eager to work with us to push forward and try out the next generation of technology and services. On the other hand, there are operators who are more operation-oriented, whose main concern is for the network to “hum” and to reduce the number of customer complaints. They aren’t willing to consider any changes as long as the new technologies haven’t been fully proven, which by definition places them further down on the innovation curve.

We are able to help operators by presenting them with a longer-term vision into the future, so they have a better idea of what the endgame will be. That way, even those who are less aggressive with their upgrades find value in trialing with us. This helps them understand the big picture and what future steps they may need to take. I believe this is a healthier approach than simply waiting until we are ready to present them with a finished product.

At the other end of the spectrum, we also have customers who are more proactive and more willing to test out new technologies; we basically consider them partners. Their feedback is invaluable when it comes to hammering out the details and figuring out best practices. They benefit from this since they are able to have a greater influence on the eventual solution, making sure it addresses their specific circumstances.

Overall, even taking off my Parallel Wireless hat for a moment, I recommend operators to join the discussion early, regardless of their deployment timeline. This puts them in a position to be part of the process to shape future solutions and they will get the best product to suit their needs.

Take us into the mobile NOC. What is causing the biggest headaches for mobile network CTOs these days?

The biggest headache is power consumption. Cells are getting more power hungry, energy costs are rising, sustainability concerns are getting louder and the situation is only getting worse. Energy efficiency definitely seems to be at the top of the list of concerns.

Second place can be split into two, depending on the operator. There are some operators who need to first optimize other costs while meeting the demands of the future (capacity, latency, reliability). This is another vice that CTOs get squeezed into and it is of no consolation that the best solution to address this, a true 5G network, requires a huge upfront investment and a prolonged adoption phase.

Other operators take a more visionary approach and are looking ahead as they start to build the network infrastructure to support smart vehicles, smart homes, city-wearables, smart lights, etc. leading to many more devices and much higher capacity with different use cases (which will require, for example, network slicing). This will require more capacity and support for differentiated QoS.

What is the tipping point between selectively deploying 5G and deploying 5G across the network?

The cost of rolling out 5G is significant, and operators are naturally hesitant to jump right in. Some of them are bound by regulatory directives in some countries that require a certain level of 5G rollouts, or support for specific features, by a specific target date. In some cases, regulation is driven by politics, a desire to demonstrate innovation. In other cases, it might be a by-product of other requirements, such as the European requirement to support mBIoT. Countries are pushing for infrastructure to be smarter; smart water meters, smart electricity monitors, etc.

Overall, I do not think that network capacity is the main driver, in most cases (except for mass events like a concert) capacity is not a problem, QoS is. The need to ensure QoS is increasing since more smart devices means different levels of QoS, even when there is sufficient capacity.

What does 5G SA support that 5G NSA cannot?

True 5G networks will be based on 5G SA core; the main reason to do NSA is an easier evolution on the existing networks. Choosing to deploy 5G NSA is understandable because of the short-term ease and 5G NSA is a reasonable half-step, but it will not be possible to support all the capabilities that 5G technology offers.

The classic use case for 5G SA calls for the full throughput, network slicing and ultra-reliable low-latency. Different devices that are consuming different services will be able to get the QoS that they require.

What is your view of RAN Centralization?

I have no doubt that this is the future, you can even read this between the lines of the 6G standards that are being discussed. Everybody knows that centralized RAN is more efficient, easier to manage, you can share resources and be smarter. When components are distributed you have a latency problem on the decision-making that you don’t have when they are centralized.

It may take time to adapt to it. The industry is 40 years old and in many ways is reluctant to make big changes. Let’s face it, there are points that need to be considered carefully. Redundancy is a good example. When sites fail, the impact is local so there isn’t an urgent need for a redundant backup. When designing a cluster architecture, it is important to plan for operational and geographic redundancy for each cluster, to avoid disruptions.

The good news is that the benefits of RAN Centralization justify the extra planning, and that PW architecture addresses most of these concerns. Efficiency is a key for future mobile networks and RAN centralization is a key component for that.