So, we've got one more of the IoT Verticals that we want to spend a little bit of time with, and that's the connected car, it also includes autonomous driving. We sometimes overload these a little bit our minds. They're actually going to be some different functionalities in them but again, if we think about the transportation issues associated with the connected vehicle, we'll see that we've got some specifics the functionalities that drive us to look at a 5G type of a network. So, for the connected car, the implication associate with the service delivery in there, looking at the cylindrical icon here, is in-car infotainment systems. So, you think about streaming video for the passengers in the back. The massive sensors for the car monitoring itself. This ties in both not only to the modern car itself, that has an amazing amount of compute resources and sensors in it that are used for monitoring and controlling that device, but also tying that back into the functionality associated with autonomous driving. Even think about simple things of what you can do with backup control now, and tying those information and into the backup camera, and being able to change the velocity of vehicles you're backing up as those devices may in fact be detecting that you got an obstacle, that you might not be paying quite enough attention to, that you want to avoid or actually apply the brakes as you're approaching that device, and the connectivity that's associated with that, but also the sensors being able to relay information about the vehicle's health and safety features, back out to a service monitoring point if that's necessary. The end-user you tethered into that car system itself. So, not only if attainment in the car itself but you may also have Wi-Fi hotspot type of connectivity to use your endpoints that are in that car and then using the macro connectivity of the device itself to consolidate those elements. Then the big vector, obviously here is the autonomous driving aspects of the vehicle itself. While we wouldn't be looking at necessarily the 5G network for every aspect of it, the significant aspects of it, that will come into play there and that's going to possibly vary depending on the vendor itself, it's going to be building and integrating those platforms. So, earlier on, we mentioned very briefly the dataflow information requirements that the autonomous car or the autonomous vehicle would bring to bear, and we mentioned the fact that to do all cycled uses that platform is less than 24 by seven. By one report, we can take a look at the graph on the screen here and see the bumpiness if you will, obviously between the hours of 12:00 AM and 5:30 AM in most regions the utilization of those vehicles is relatively low, and consequently any of the data that flows from that vehicle into the network is going to be low. Then we've got what we would expect to see the bumpiness associated with a morning commute drive time, it gets a little bit abatement during the mid morning hours, it goes back up again as people are running errands around lunchtime, again some abatement during the middle of the day itself, and we've got an evening drive time and most of those geographies, and then a continual drop-off in there. Unlike the transportation network, the physical network which has to accommodate that flow, we may in fact look at a communications network, where we can turn services on and off or ramp them up and down if you will, to consume our resources or allocate them to other bits of logical functionality during those quiet times, between 1:00 AM, let's say at 4:30 AM or whatever it happens to be, maybe we quench from big data, operations with our platforms on these aggregated systems in that type of functionality. Back to that bit rate, a variety of sources depends on who you ask, as to how much information is actually going to be sourced to the network, if you look at 3GPP, standards body responsible for 5G, they're giving one of the lower estimates, if you look at Hitachi for example, industrial firm that has a lot of experience with a relabeling centers on very large devices, they're couples or magnitude larger from their estimate. So, there's still some debate going on as to how much information the connected car is going to drive on to the network, but it's certainly a lot more than we have today. When we look at the functionality of those use points, now in that connected car, the services, they're actually different characteristics. So, they're are going to have different quality of service, QoS, in there, and there are also some of those elements that are going to be heavy in the control plane, that is, there the signaling aspects, and others are going to be heavy in what we call the user plane or the data plane. If you think about streaming a video to a connected car or streaming video from a connected car, that's going to be heavy user plain type functionality, if you think about information from sensors, that's a control plane, an example of a control plane, bit of functionality. Also, as we talked about one of the things that's going to happen inside of that network, the calm service provider network, is going to allocate those resources based on that demand, so that we don't overbuild and allocate user plane functionality where necessary during those busy times and allocate the control plane functionality similarly in the area. On the control plane area, this is difficult table to go through, so you have to study it a little bit but, highlighted here is an example of the 4G versus 5G. So, average LTE capabilities versus what we would estimate would be need by a connected car. As we see, there's a significant differences in some cases between 400 and 700 percent difference in that IoT control plane usability from that car standpoint. So, a connected car using 5G, for example for control plane functionality using the Hitachi example, for example, we place a burden on the LTE 4G network is not capable of meeting. So, if you were to do that in this type environment, you clearly need a 5G capability in order to provide that.
