Now that we have had a good summary of benefits provided by a massive MIMO and beamforming, let's try to answer a question that we have always tried to answer. All these conceptual benefits look nifty, but what will be the practical benefits of massive MIMO and beamforming? To that end, my team at Qualcomm, execute certain additional simulations focused specifically on massive MIMO. Those simulations were conducted in real-life conditions and that would tell you that the results of those simulations would be applicable in real-life scenarios. The first scenario is that of a relatively flat suburban area and the second scenario is that of a dense urban area. Let's look at the suburban area first. For both of these scenarios, we have a before result and an after result. Before would be without massive MIMO or beamforming and the after would be with some degree of massive MIMO and associated beamforming. Our objective will be to quantify the performance improvement provided to the system by massive MIMO and beamforming and while we do so, we'll make sure to have a fair comparison between the before and after scenario. In that, both the scenarios will the feature the exact same amount of channel bandwidth and they will feature the exact same UE on the exact same phone because if you upgrade your phone, sure, you will get benefits. But what benefits the system gets just by making some improvement on the network without assuming that users will update, therefore. With those assumptions and our prerequisites in mind, let's look at a typical suburban scenario wherein you would obviously have macro networks and enterocyte distance in this simulation is more than a mile and that is what makes these cells macro, as we discussed during one of the earlier modules. This black line, plotting user throughput or the data speeds that individual users get, pertains to the before scenario wherein the genome B only has two antennas, as it shows here. Whereas this green line pertains to the after a scenario wherein all the other variables are exactly the same, just that the genome B now has 24 antennas instead of two and 24 is not truly massive MIMO, but it is well on its way to becoming massive MIMO. Even if we increase the number of antennas on the genome B from two to 24, and make no other changes to the system, we keep the bandwidth the same, other UE hardware and software are the same as well. What kind of user throughput benefits we start to receive? Well, these real-world simulations tell us that a typical user is expected to get more than 4x benefit in the after scenario, that is with beamforming and massive MIMO as compared to the before scenario which was non-being formed. Imagine how getting a 4x improvement in your throughput most feel like. If you have a home Internet that barely gives you let say 50 Mbps, when everybody else at your home is using the same Internet connection. Imagine what would happen if you suddenly got 4x the throughput instead of 50 Mbps each and every one of you got 200 Mbps at home. Just imagine how many of your problems that sudden throughput increase would solve, and that is a fundamental promise of massive MIMO and beamforming in that in typical scenarios, it would help your system increase user performance by a margin of 4x, which is significant. But you might ask, typical user is fine. What about the worst-case performance? Well, as it happens, even in the worst case, you happen to see about 4x improvement in their own performance as well. We see a common trend here that at least in this real-world scenario, based on how beamforming and massive MIMO were implemented, our results indicated that no matter our typical you will or a worst-case you will, located in challenging network conditions, both of them solve around 4x improvement in that individual speeds and that goes to show us that massive MIMO and beamforming don't just have conceptual benefits, but their benefits are also readily realizable in practice on the same UE, keeping the same channel bandwidth, just changing the number of antennas on the genome B. This is one simulation. The next simulation is in a dense urban area so that there are plenty of tall buildings, and even here, we tend to see a similar thing, in that even though the topology is different, our simulations, which are executed in the real life condition here, indicated that typically we stands to get about 4x improvement in its performance even in such a different scenario. That is the thing that we see here, that no matter the topology and no matter the user conditions if you implement massive MIMO and beamforming properly, it is not just a theoretical concept or [inaudible], but it has a potential to give you enormous benefits with respect to user throughput and network capacity, as well as network coverage in some of the most challenging real life scenarios as well.
