RF chip giant Qorvo explains how 5G will get deployed.
Sumit Tomar, general manager of the Wireless Infrastructure Products Group at RF chip giant Qorvo, sat down with Semiconductor Engineering to discuss the development of next-generation 5G wireless networks and other topics. In 2014, RF Micro Devices and TriQuint merged to form Qorvo. What follows are excerpts of that conversation.
SE: 5G, the follow-on to the current wireless standard known as 4G or long-term evolution (LTE), will enable data transmission rates of more than 10Gbps, or 100 times the throughput of LTE. Is 5G really happening or not?
Tomar: Absolutely, it’s happening. With major operators, we’re actually doing trials.
SE: Why do we need 5G?
Tomar: If you look at 4G, it’s connecting people. 5G is about connecting people and things. We anticipate that there will be 50 billion connected devices. And with the way we download videos and YouTube, and upload videos from Facebook and other sites, mobile broadband is exponentially growing. If we don’t get to higher bandwidths, the data that we get on the phone will be more like 2G. Nobody wants slow data on their phones.
SE: What’s the issue with 4G?
Tomar: Regarding the frequencies used for 4G, which are below 2.7 GHz, there is a very limited spectrum. And whatever spectrum was available is already taken. So we see major operators looking at frequencies higher than 3 GHz. The way the market is now split is that below 5 GHz, people will start deploying what they call LTE Pro with massive MIMO. It’s sort of 5G, but it uses a LTE modulation scheme below 5 GHz.
SE: 5G involves several frequencies, right?
Tomar: We also see frequencies being talked about from 10 GHz to 90 GHz for 5G applications worldwide. The first base station field trials will move forward with frequencies from 28 GHz to 32 GHz and from 39 GHz to 44 GHz. That’s where we see the traction. But people are also talking about frequencies from 24 GHz to 28 GHz, 10 GHz to 15 GHz, and 60 GHz to 98 GHz. There are all kinds of millimeter wave frequencies, anywhere from 6 GHz to 100 GHz.
SE: What are the issues?
Tomar: As you go higher in frequencies, the cell radius reduces because of absorption. The coverage is inversely proportional to the square of the frequency. So the higher you go in frequency, the lower you would go in coverage. So we anticipate when these base stations from 28 GHz to 32 GHz and from 39 GHz to 44 GHz become a reality, the cell radius will be more like 200 meters. So that means you will need to provide massive capacity in very localized areas. You need WiFi-like things.
SE: What else?
Tomar: We need coverage when we are driving on the road and everywhere else. So for that, we still need 4G. So our view is that both 4G and 5G will co-exist. 4G will be predominately used for wider area coverage. And 5G will be predominately used for localized high-capacity data. For example, you go downtown or a shopping area, where the population density is very high. That’s where 5G will get deployed.
SE: Isn’t 4G good enough?
Tomar: It is for providing basic phone functions. Maybe you can download some e-mails. The problem comes when you start using this as a TV. For example, if you are live streaming a game that you are watching at a stadium, there are 100,000 people. Then, other people start streaming video. That’s when the problems start. The problem cannot be solved with 4G. 4G doesn’t have enough bandwidth. Capacity is directly proportional to the bandwidth. The more bandwidth you have, the more data you can have. And that’s where the need for 5G comes.
SE: The transition from 4G to 5G will be complicated, right?
Tomar: It will be complicated. But 5G is needed. Today, there are 7 billion people on the planet. About 4 billion people are still not connected to the Internet. Within the next three or four years, many of those will be connected to the Internet. Today, we are already reaching capacity at 4G with less than 10 billion connected devices. Think of the scenario if 50 billion devices get connected to the Internet. What would happen to the overall network?
SE: There are other problems with 4G, right?
Tomar: Today, the network latency is slow. Take a look at autonomous driving. The network latency is over 10 milliseconds today. If you try and apply the brake automatically, and you are running your car over a cellular network, the chances of survival are small. You’ve got to cut the latency down to less than 1 millisecond. What 5G does is not only provide a massive amount of connected devices, but it also cuts the latency down to less than 1 millisecond. So the self-driving car can become a reality. And it also addresses the frustration that consumers have when they download video on their phones.
SE: How do you see the transition towards 5G happening?
Tomar: There are two phases of 5G. One is from now until say 2020. And then there is 2020 and beyond. Before 2020, I believe that we will see massive MIMO, where as many as 64 to 256 transmitters will be used. They might be deployed at frequencies below 5 GHz. Then, beyond 2020, we will see the deployment of 28 GHz to 32 GHz and 39 GHz and 44 GHz.
SE: What is massive MIMO?
Tomar: MIMO is multiple-input and multiple-output. The word massive is used because the channels could grow as much as 128 times. In massive MIMO, what people do is they use a number of transmit and receive channels. If you look at it today, the number of transmit and receive channels in a base station is usually 2 transmit channels and 4 receive channels. To increase the capacity, the number of channels will go from 2 on the transmit side and 4 for receive to as many as 64. It could go as high as 256.
SE: What does that mean for the base station?
Tomar: If you look at a macro cell base station today, the power level in those base stations is in the range of somewhere between 30 Watts to 80 Watts average power. Regarding these massive MIMO systems, there are 64 transmitters instead of 4. But still, we see the power dropping anywhere from 2 Watts to 10 Watts.
SE: To accomplish this, the industry is talking about moving from LDMOS processes to gallium-nitride (GaN) in the base station, right?
Tomar: The base station in 4G has been historically dominated by LDMOS. LDMOS has device physics limitations. That’s where GaN enters the picture. The need for higher peak power, wider bandwidth and higher frequencies are driving the adoption of GaN inside the base station.
SE: Is that happening now?
Tomar: The level of integration is moving towards GaN. We have a product called the 2705. That’s the smallest form-factor power amplifier for 5 Watt applications. Specifically, it’s designed for massive MIMO base stations. To me, these are 5G. If you look at the data you can get using those base stations, even at 2.5 GHz to 2.7 GHz, it’s almost close to 3-gigabits-per-second. It’s growing from 100mbps to 3-gigabits-per-second using the existing LTE-A technology. It’s 5G-type performance level. And then, in frequencies higher than 6 GHz, we see GaN technology playing an important role. From a GaN perspective, we are ready for 5G today.
SE: GaN has been used in the defense/aerospace sector, right?
Tomar: Qorvo has been investing heavily in GaN. We’ve been working on it for more than 15 years. We are working closely with our defense customers to enable these phased-array radar systems. You use these in defense communications. If you look at a 5G base station architecture, it looks like a phased-array radar system. All of the systems and process capabilities that we’ve designed for phased-array radar systems fit well in a 5G base station.
SE: A phased-array device consists of an array of antennas with individual radiation elements. Basically, it can steer a beam in multiple directions using beamforming techniques, right?
Tomar: In the base station, the beam is consistently moving and focusing on the area where it needs the most capacity.
SE: So GaN technology is moving to 5G?
Tomar: We are sampling devices to our 5G customers that are doing base stations. Based on those components, which predominately come from defense communications, we have successfully completed trials with more than 22 operators worldwide.
SE: So when will 5G happen?
Tomar: I don’t see a major 5G deployment before 2020. That’s for 5G in higher frequencies like 28 GHz and 32 GHz. But below 5 GHz, the massive MIMO systems will start deploying next year. You will start to see a significant improvement on the data you get on your phone starting next year. Our customers are deploying massive MIMO systems, which are 5G architectures but at lower frequencies. They use existing LTE modulation. The reason why customers are using LTE modulation is because 5G modulation doesn’t exist. The standards bodies have not defined those yet.
SE: Will the 5G base station be based exclusively on GaN?
Tomar: There is never one technology. There are always multiple technologies. Look at 4G today. In a single sub-system solution, we have SAW filters, BAW filters and two different flavors of gallium-arsenide. We also have SOI technology. In a handset today in a 4G module, there are as many as 5 or 6 process technologies.
SE: What about handsets?
Tomar: If you look at the 4G handset today, it has gallium-arsenide and SOI. In 5G, it will be GaN and SOI. Maybe GaAs. GaAs and SOI will still exist. GaN will also be added, particularly at higher frequencies.
SE: For defense applications, GaN has already moved into handheld devices, right?
Tomar: In handheld devices, you need more battery life and peak power. GaN is already getting adopted in handheld applications.
SE: How do you bring down the cost for mmWave technology from the defense sector to the commercial world?
Tomar: It’s already happening. The way we see it is the technology starts from the defense world, whether that’s cellular communication or GPS. Then, it finds its way into the commercial market. It’s the same thing happening here. We bring the technology from the defense side and then bring down the cost structure from our mobility business.
Waiting For 5G Technology
New wireless standard will significantly speed up communication, but dealing with mmWave technology isn’t going to be simple.
Inside The 5G Smartphone
Video and overall demand for higher bandwidth and throughput are creating new opportunities in the smartphone market, sparking new technologies, approaches and materials.
Mobile Edge Computing For The IoE
The network edge will become one of the most critical metrics of the IoE and 5G, but it will require new technologies and solutions.
Plotting The Next Semiconductor Road Map
Industry leaders examine enablers, implications, and perspectives for a changing technology ecosystem.