Introduction to private 5G networks tech talk
Hema Kadia: Welcome to the 5G Tech Talk. Today’s topic is Private 5G Networks and testing. For this session, we will have a conversation with Adnan Khan, a senior manager for wireless and wireline market development at Anritsu.
What are the top 3 industry verticals investing in private 5G networks?
Hema Kadia: TeckNexus recently did a thorough analysis of the global private 5G enterprise use cases and spectrum winners of the mid-band and high-band frequencies.
We discovered that manufacturing is a key sector where we see a fair amount of traction, ongoing trials, and deployments from the enterprise perspective. Within the manufacturing vertical, automotive manufacturing, specialty chemicals manufacturing, and industrial appliance manufacturing are the key sectors that are investing in or deploying private 5G networks.
If we only look from the global private 5G enterprise use case perspective, manufacturing is again the top vertical, followed by transportation. Within transportation, the two key verticals are airports and seaports. The critical drivers for both these sectors are – Increased security, guaranteed quality of service, or guaranteed SLAs and how they can bring efficiency to their day-to-day operations.
Energy and Utilities Industries
If we look at the global spectrum winners – in Europe, APAC, or North America (including the recent CBRS spectrum), then the critical vertical after manufacturing is the energy and utility sector. Again, the key drivers for this sector are – security, quality of service similar to transportation, and real-time monitoring to support predictive maintenance.
What are the key 3GPP features enabling private 5G networks?
Hema Kadia: Adnan, from a 3GPP release perspective, what are the key features that would help us enable the private 5G networks?
Adnan Khan: To give background on the 3GPP release for 5G, we have primarily divided it into phase one and phase two. Phase one was release 15, which got frozen, i.e., all the core features that were supposed to be part of it. The primary use case was the enhanced mobile broadband that got frozen sometime in June 2019, and roughly a year from that this year, in July of 2020, we had release 16 that got frozen, and now the release 17 work is just going on. If you look at release 15, it was all about mobile broadband, and it was a matter of just getting high throughput out. In release 15, the focus was not on the optimization of the networks or other use cases that could use 5G as an underlying technology.
And that is beginning to happen in release 16. I.e., the focus is on how we can optimize the devices, optimize the networks, and the critical use cases that can use the underlying features and technologies built into 5G. As you rightly talked about earlier on, many security concerns are a bit better with 5G. There are many latency concerns in 4G, but 5G will subdue those concerns. And there is a lot of discussion concerning the fact that you can get a much higher throughput. You can also use features like network slicing, for instance, to dedicate a specific segment of your network across your core network and an access network to meet the industry’s demands. So as you see in release 15, 5G came on board; it’s going to get adopted in release 16; when the technology is adopted, the use cases will drive the technology to be different sort of use cases.
The important thing you can look at is the NR in the unlicensed band; for instance, that is something similar we had in 4G, whereby we had LTE, the license, and assisted access, but that was primarily on the downlink. But now we have NR, and you have much wider bandwidth so that you can have standalone without any sort of anchor in the NR unlicensed band. So that is mainly from the feature perspective. If you look from the technology perspective, there are many items in 5G that are not like 4G, e.g., if you look at the physical layer of 5G. It’s much more enhanced and introduces the link latency between the physical layer as you go onto the Mac layer, the RLC layer, and the PDP layer. And then there’s a new layer, SDAP, which is the service data application protocol. So that’s some items like that.
Numerology is pretty much being able to transfer things much quicker. You can have a flexible, basically subcarrier spacing. So in LTE, you had a fixed 15 kilohertz of carrier spacing. In the case of 5G, you have FR1 and FR2 that go all the way up to 240 kilohertz. So as you increase this subcarrier spacing, you have a lower short time to the interval so that you can send things quick and short bursts.
So that does reduce the battery life, and you can get much smaller latency all the way when we’re talking about sub-milliseconds. Those are some of the features and the reliability portion of it using different coding techniques like LDPC. LPDC (low-density parity coding) is one of the code encoding techniques used to provide ultra-reliability, whereby you don’t have to transmit a complete block. In the case of LTE, you send a packet. If the packet gets lost – you have to re-transmit the entire symbol. For LDPC, this is not the case. Only transfer the block, the fraction of the block that is lost.
What are the pros and cons of the three deployment modes of Private 5G Networks?
Hema Kadia: There are three key deployment modes for private 5G: the public 5G, where network slicing would be the critical capability to leverage for segmenting the various use cases. Of course, the other two beings: the private 5G and the combination of public 5G and private 5G. Would you spend a few minutes helping us understand each of the deployment’s critical capabilities and what are the related pros and cons?
Adnan Khan: Yes. The key thing we’re looking at when we have private 5G is to have guaranteed coverage whereby you could have some underserved areas and want to make sure you have adequate coverage. And again, network control is another one of them. So you can have different configurations. There may be a specific sort of configuration that may not be available in a public network, and you can apply that to the private network. So it could be some sort of a robotic factory, for instance, and that would have different configuration requirements. So you have more control over that.
And then you have other features like performance as well as security. So you can do better access management and better data connectivity. Again, from a deployment perspective, you have a public view of having 5G, and private 5G. And when I say public-private 5G, you are using the resources from the public network like an operator may have, but those resources will be dedicated to a specific purpose. You could have a network slice, for instance, taken from a public network and dedicate that slice. So you have some guaranteed quality of service, guaranteed coverage, sort of basically guaranteed network control. But you apply you’re using the public resources to meet the demand of a specific use case or a particular segment like a factory Industry 4.0. So that’s one deployment that you’re using for all public resources.
In other cases, you have some sort of an island deployment. An island deployment would be something that is very private, unlike a public, where private 5G will be configured to that location’s specific needs. And those particular needs can vary from place to place. You can have a private 5G that is a complete island by itself. And that island, meaning that it has its core network, has the access network, it has its spectrum. It could be using the mid-band spectrum you’re talking about or using an unlicensed spectrum as we spoke about the NR feature in the unlicensed band. But it’s completely isolated from any sort of macro network.
And there will be no kind of mobility. Everything will be intact. And the other form, you know, like a hybrid form whereby you do have this private network, but that private network has some sort of mobility with the macro network. So you could have some kind of roaming agreement between a private network and a macro network. An example is people walking into the factory. And as they move out, they could use the same sort of devices with the same kind of credentials done in the micro private networks going on and transitioning into the macro networks. Those are some of the things I can do when you have a complete inland sort of private network; it gives you more control, less risk of intrusion, and better security. And then as you go into some kind of a completely private network that is using a slice of the public network, in that case, you have all the built-in security, but you tend to have, I think it’s going to be a much lower cost because you don’t have to redeploy the network.
The network is already there. You just have to dedicate a specific slice of that network for a particular purpose. So it’s a lower cost to go with some sort of a network slicing way of a private network. Right. And then the hybrid, comes in. It’s going to be more expensive. However, it could be less costly than having a complete inland sort of private in this case. As I said, you can use a lot of the core networks. We don’t have a core network. You could use the credentials on the operator’s core network or, for instance, for access management. The way I would envision it, an entirely inland private network, it’s more cost-intensive. You have the hybrid deployment mode whereby you have connectivity with the macro network, and users can have a dedicated network slice for a specific purpose within a particular area. And that would be the most cost-effective way to start.
What are the testing challenges and considerations for private 5G networks?
Hema Kadia: Considering the complex deployments across all these three deployment modes and related frequencies, what are the key testing challenges or considerations, and how are Anritsu solutions addressing these testing challenges?
Adnan Khan: Sure, so the testing challenges have pretty much begun from the time we had 5G started right about two and a half years ago or so, and the key challenges that we see over here are both on the network side as well as on the device side. 5G introduces us to a millimeter wave, which is a bandwidth in excess of one gigahertz. When you’re in LTE, you’re looking at maximum bandwidth size, which would be a 20 megahertz channel. You can now have channels of one hundred megahertz, and then they can be connected and aggregated using the connectivity and multiple resources.
So it becomes complex, whereby the device and the network both have to support multiple ranges of frequencies, wide bandwidth, shorter bandwidth, different subcarrier spacing, and then different deployment modes. The key challenge is, how do you develop such devices? And once you have developed the device, you want to make sure that those devices are performing well. And need to ensure that the performance of the device or the networks meets your requirements and standards. First, the KPIs have to be correctly defined, and once those are defined, we need to have better testing methods. These are very cost-intensive as you expand into 5G, and you have to cater to millimeter waves, whereby you cannot see the antennas on the device. They’re pretty much embedded into the chip. How do you connect the device?
At the same time, you have to test LTE, as it is not going anywhere. You need to have interoperability with LTE as well as with 3G. So currently, there is no handover supported with 3G. It’s going to be if you break the call, you will make the call again. We at Anritsu want to ensure that we have a system in place. We have our network emulators in place that can cater to multiple radio access technologies and multiple carrier aggregation performances so we can aggregate more carriers with a similar common platform versus having multiple boxes to do that or having multiple boxes with different radio access technologies. We try to ensure that everything is built into one single box, whether you are trying to do some sort of RF performance testing or any kind of protocol testing. And one big part of it in 5G, as more use cases involve, is what we call application testing. I.e., the application is performing as per the use case that is determined by the user over there.
So again, in spectrum testing, make sure you have supplied different spectrums. We’re able to cater to different deployment modes – non-standalone, standalone, dual connectivity, carrier aggregation, and dual connectivity between NR and NR ads. Are we able to test latency at a seven-millisecond level as well as to OTA testing for millimeter wave, for instance? And be able to expand those millimeter wave testing not only between twenty-four to forty-three point five gigahertz but be able to go all the way up to higher frequencies as you’re going to release 16 and early 17, for instance. There are features that will require spectrum to be used over from two-point six gigahertz or seventy gigahertz. So we have tried to come up with common solutions. And it’s a lot of work to make sure that the solutions are robust and cater to different needs.
What solutions does Anritsu offer for private 5G networks?
Adnan Khan: The company has been in place for over one hundred twenty-five years in the test and measurement market. We play a role in different segments. We are not really at the tail end, but we try to make sure that we go through the various testing cycles.
And the reason behind that is as you perfect one cycle of testing, a lot of these cycles are sequential. And as the technology matures, the cycles tend to become more agnostic of each other. But in the beginning, if you can do that, for instance, in the chipset R&D phase, if you have worked with different chipset vendors, then there is a very high probability that the issues were found and eliminated.
You would have found all those issues, whether on the test equipment side or the chipset side. So that’s one area that we play into is the device chipset in the R&D phase. On the device chips, you could have features like 8CA and 10CA on FR2. They have been brought up now, & then it goes into the next phase, which is called acceptance testing, conformance testing, whereby you need to make specific devices or networks meet the requirements set by the specifications in 3GPP, so we participate in that.
And then the next phase is production. So now, as technology has become commercialized, there’s going to be more deployment of it. So we ensure that the devices that are getting manufactured are going through some series of testing to ensure that the performance and the protocol capabilities meet the end user’s needs. So we try to perform and align ourselves with the different phases of testing. And one of these products that we have right now that you see here is MT8000A. It is the platform that tends to take 5G testing to the next level. The same platform used for – the R&D phase would be the conformance system’s brains and used for production testing. So we know that by the time the testing has reached the R&D phase to the production phase, that specific hardware and the software residing on that hardware have been tested with its full capability and have been integrated with different chipset vendors. As you know, in 5G, there are a lot more chipset vendors that work during peak-time support.
What are the key differentiators of Anritsu Testing Solutions?
Hema Kadia: Hey, Adnan, do you also want to spend a minute highlighting some of the advantages/differentiators that Anritsu product solutions offer?
Adnan Khan: Yes. So some of the value propositions that we offer include simple hardware such as MT8000A. It supports release 15, it supports release 16, and we already have work looking at a high level on the features that will be in release 17. That’s one part. And then obviously, again, the hardware is common. It’s a different configuration for different functions, and it can support different frequency ranges, FR1 and FR2. MT8000A can support both frequency ranges used for private 5G, and support different deployment modes, i.e., standalone mode and NSA mode. The other important part is the interoperability with LTE, 3G, and the predecessor technologies. As I said, there could be an instance whereby you can have a private 5G network with connectivity with the access network on the public side. There could be interoperability, i.e., roaming issues that could be found between 5G and 4G.
Supporting LTE, 3G, and previous technologies without using a separate box is very important. The other differentiator is the capability to support different types of testing – application testing, protocol testing, and integration with varying vendors of the chipset, other OEMs – to ensure that we have achieved stability in the R&D phase and are able to provide repeatable and robust performance concerning the measurements we’re taking.