I think the 5G private network sounds very interesting to me. It sounds a little bit like Wi-Fi. A 5G pirate network is it better than, let's say, Wi-Fi 6, or is it even fair to compare them? 5G and Wi-Fi are two different wireless communication technologies just for the benefit of our listeners. Wi-Fi 6 is the sixth generation of the Wi-Fi standard. Because it is nonetheless a Wi-Fi standard, it continues to follow some of the same principles as our traditional Wi-Fi, for most of which is the fact that it continues to operate in unlicensed frequency bands. For example, 2.4 gigahertz or five gigahertz. Because it operates in unlicensed bands, it continues to be plagued to various degrees, but continues to be plagued nonetheless by some of the same problems that current generations of Wi-Fi face. For example, if two neighbors deploy a Wi-Fi 6 networks in close vicinity of each other, they are going to interfere with each other at some or the other point in time. Another problem with Wi-Fi is that the scalability that it offers is not as good or robust as the scalability that 5G is going to offer in that if you imagine a graph, so to speak, wherein you plot the performance of the system against the number of devices, you will see that the performance of a Wi-Fi system degrades much more sharply as you add more and more devices as compared to the degradation that 5G will face as you add an equal number of devices. Scalability is one of the fundamental advantages that 5G has. It has been designed to be far more scalable and better scalable as compared to Wi-Fi 6 even. Another point in favor of 5G is what us engineers call QoS or quality of service. But in everyday terms, it is essentially about how well can you guarantee a certain throughput, certain latency, and certain reliability performance. The way 5G is designed, it can make quite lofty promises about throughput latency and reliability. But more importantly, it has been designed to perfectly keep those promises as well under a variety of operating conditions. Wi-Fi, on the other hand, because it operates in unlicensed band, it is prone to interference, low scalability. Wi-Fi cannot guarantee the quality of service to the extent that 5G or even 4G can guarantee you. Those are some of the factors for which I would still think that 5G has certain undeniable advantages over Wi-Fi 6 in general, but especially in an industrial IoT-like environment, wherein we are going to likely have tens of thousands of devices within the same industrial factory. If you recall the scalability point that I mentioned, if you plot that imaginary performance graph, your Wi-Fi performance is going to roll off or degrade much faster as compared to any degradation that your 5G private network will face. That is what makes 5G private network more advantageous as compared to Wi-Fi 6 even in an indoor private IoT-like deployment. Let's say if I worked in a factory and I decided to build this 5G private network, what do I need to do? Do I buy up the equipment, sign-up for spectrum or anything like that? Let's mentally recount what would a hypothetical private network entail, and that will take us closer to our answer. As you correctly pointed out, you will first need to ensure that you have a spectrum to operate it. Then you would need to buy up the RAN components, i.e. the towers on the base stations. Then you would need to arrange a core network that the RAN can communicate with, and lastly, you would need to ensure that your private network has access to certain services. Those services could be something like a database repository or an image recognition service, or simply, email and instant messaging service, whatever. But you would need your network to be connected to a service network as well. There are four components. A spectrum, RAN, core, and service. Now, the way private networks are being commercialized is that you can pretty much pick and choose what you want to do by yourself as the private factory owner and what factors you can outsource for a variety of reasons. Once again, as with any engineering problem, there is a spectrum of solutions, and different solutions achieve different degrees of trade-offs with respect to different factors. At the one end is where you do everything by yourself, and at the other end, you outsource every single thing. When you're trying to do everything by yourself, let's say you buy yourself a dedicated spectrum, build a dedicated network, a dedicated service as platform, you have the capability to precisely design and deploy the network exactly according to your needs, not a bit more, not a bit less. You can guarantee performance in terms of latency, throughput, coverage, reliability, availability, what have you. Your performance will be the best. However, because you are doing everything yourself, you are buying a dedicated spectrum, buying dedicated components, that is going to involve some significant CapEx and OpEx, as we engineers like to call it, that is capital expenditure and operational expenditure. You are going to have to make some significant investment in your end-to-end private network. At the other end of the spectrum, if you choose to outsource everything, that will probably be the most economical option, because you are, in most cases, just renting the setup as you need. But as you may have guessed already, your flexibility and performance may not necessarily be guaranteed. The performance that will be achieved by an end-to-end dedicated network will be, in most cases, superior as compared to the performance achieved by an outsourced setup, so to speak. Factory owners will look at the spectrum of reasons that might span the whole gamut between doing everything yourself and outsourcing everything, and they will arrive at a sweet spot depending upon their objectives and the resource budget that they have. This is, in general, how factory owners will think about when deploying their private 5G and IoT networks.
