Some of the fundamental types of spectrum usage that we have seen earlier, such as licensed and dedicated or shared or unlicensed, continue to apply for 5G private networks as well because they are after all 5G networks. Pursuant to that, there are multiple options that private network owners can choose when deploying 5G private network for IOT purposes. At one end they can use licensed spectrum, but that spectrum may be owned by a different cellular network operator which may be in the public domain. So the industrial owner can either by the spectrum from that operator a small sliver of the spectrum or it can lease the spectrum long term or it can just ask for short term allocations of that licensed or dedicated spectrum. And because it is a dedicated spectrum as we saw earlier, all the benefits with respect to performance guarantee can be attained although at a slightly higher cost. There is a different type of licensed spectrum that is introduced specifically for IOT networks, and that is called the dedicated spectrum. In that in several countries, certain slivers of frequency spectrums have started to be dedicated for industrial purposes. And industrial factory owners by their own volition can participate in auction for those particular frequency channels rather than them being limited to traditional cellular network operators. So even this is licensed or dedicated spectrum, but with key differences in that it is in a slightly different range. For example, in Germany, it is in 3.7 gigahertz. That spectrum maybe in a slightly different range as compared to our traditional cellular networks. And secondly the factory owner doesn't have to lease or ask for a spectrum from a traditional cellular network operator anymore, a factory owner can also try to buy some of the dedicated spectrum. And furthermore that spectrum doesn't have to be owned by one single factory owner as we saw earlier, there is a shared spectrum model. In that multiple factory owners can jointly, buy a certain piece of spectrum and they can use that spectrum in either different geographical regions or in slightly different frequencies of the same band or at different times of the day, for example. And more complex arrangements can also be chalked out. Then at the other end of the rainbow, we had unlicensed spectrum which will continue to have. As you know, it is free so the industrial factory owner doesn't have to shell out any money to buy this spectrum. But because many other players may also be using the same unlicensed spectrum stands tourism, that performance cannot be guaranteed in such a private network deployment built on top of some unlicensed spectrum. That is okay if your fundamental use cases are mostly mMBB or IOT related, massive IOT related, I should say. However, if URLLC a time sensitive performance is going to be your requirement, but you still don't have the budget to buy licensed or dedicated spectrum, some of the further releases or generations of 5G are working on certain mechanisms that would allow two or more factory owners to use unlicensed spectrum in a synchronized shared manner. For example, they can chalk out an agreement either dynamic or a static agreement, so that both the networks know when exactly they are and are not allowed to use the unlicensed spectrum in their respective private networks. And although the underlying spectrum is unlicensed, some of the advanced mechanisms that are coming up in future releases of 5G will allow those players to use that unlicensed spectrum in a synchronized shared manner. And that is expected to allow operation of even time sensitive use cases such as URLLC. And in continuation of other theme of discussing some of the practical considerations of the theoretical concepts we have learned so far, let's try to put all these concepts together. And let's try to see how an industrial network might look like when it transitions from the current legacy technologies to an upcoming technology such as 5G. So on the left you have a Legacy Network, you have multiple pieces of industrial machinery, you have input and output devices, such as utility meters etcetera. And you have other myriad industrial devices, such as robots or handheld terminals and your other mechanical components. At the other side of the network, you have all the industrial controllers, you have databases, you've servers etcetera. And currently all this framework communicates with each other using legacy ethernet networks in most cases. As you know, ethernet gives you good enough performance, but the fundamental disadvantage of ethernet is that, because it is tethered to a wire ethernet clearly discourages mobility. Whereas in reality, in modern industries, much of the equipment is not stationary, it is mobile. For example, robotic equipment, automatic guided vehicles, augmented or virtual reality, headsets, handheld terminals, all these devices will greatly benefit from mobility. And that is what tells us we have to move away from this legacy paradigm of ethernet which is static and rigid into a modern paradigm such as 5G. And that is where similar considerations as we had earlier discussed during NSA versus SA deployments of 5G networks come into picture, although in a slightly different context. In that ideally you would not only want to replace this ethernet network which is rigid and inflexible with 5G wireless networks, but you will also want all these devices to have native connectivity to 5G. That is you would want them to have 5G antennas and then being able to exchange 5G wireless signals with the 5G wireless network. Although that is a utopian idealistic scenario we would want to get to in the near future. That cannot however, be accomplished immediately. Because upgrading all these devices would be a significant challenge not just that, it would require a lot of budget and careful planning on the part of the industrial owner. So is there a stepping stone in between? Yes, there indeed is. And that is shown in this figure which has not just a self sufficient 5G network including the rung and the core network over here, but on both the interfaces it is mediated by an adapter. That adapter is called 5G to ethernet adapter or y salsa. The basic job of this adapter is to convert 5G signals into ethernet in this direction and ethernet signals into 5G in this direction. And because this adapter works transparently in the middle, what will happen is that these legacy devices will continue to send and receive ethernet messages, but instead of communicating with an ethernet network they will be communicating with the adapter. On the other hand, 5G devices or 5G network will continue to send 5G signals but those signals will be received not by these individual devices but rather by these adapters. And those adapters will convert 5G signals into ethernet in this direction and in upward direction they will convert ethernet signals into 5G signals. So this is one example of how the ethernet can be replaced with 5G wireless network in a manner that is completely transparent to all the legacy devices. Because keeping my legacy devices still keep talking in ethernet protocol thinking that there is an ethernet device on the other end. But the adapter converts those ethernet messages into 5G signals, so to speak. And the 5G network on the other hand thinks that it is communicating with 5G devices rather than legacy devices. So this is one example the way in which 5G systems can replace the jungle of ethernet cables and switches and other components in a manner that is transparent to the legacy devices and machines. And that allows us to offer all the benefits of private 5G networks in an IIOT like paradigm to be leveraged with backward compatibility.
