With the conclusion of the FCC auction of 3.8 GHz C-band radio-frequency (RF) spectrum, deployments by all of the mobile carriers will follow. According to Martin Zimmerman, director of antenna solutions at CommScope, two factors may affect whether carriers deploy the largest of multiple-input multiple-output (MIMO) antennas to serve central business districts — building penetration characteristics of such high-frequency signals and interference protection requirements for satellite communications (SATCOM) systems. Zimmerman spoke at the April AGL Virtual Summit session, “5G MIMO Antennas Are Taking Off, but Is There Room on the Tower?”
Session host Earl Lum of EJL Research asked Zimmerman about rooftop antenna sites, and which antenna technology and configuration will be the most helpful with extending coverage into skyscrapers and tall urban canyons. Although the more antennas included within a MIMO array, the faster the potential throughput, Zimmerman said that early indications from European deployments indicate that 3.5 GHz signals struggle to penetrate buildings. The frequency, though slightly lower than the U.S. assignment, is comparable.
“Multiple operators have started to back away from 64T and look more at 32T for the central business district,” Zimmerman said. The numbers 64 and 32 refer to the number of antennas in the array.
A particular problem in the United States, Zimmerman said, is that the 3.8 GHz C-band overlaps SATCOM.
“There are going to be some cellular sites that are too close to SATCOM antennas,” he said. “As a result, one of the solutions being looked at is a temporary filtering solution. It is still too expensive to build that into a 64T, because it is a niche application. But that problem is potentially going to knock out some sites from using 64T.”
Another problem Zimmerman cited with using 64T MIMO has to do with antennas concealed within flagpoles. “In general, there isn't enough room,” he said. “Those are mostly one-antenna-per-sector solutions. In those situations, carriers mostly use passive antennas, and all the radios are on the ground.”
Lum explained that most of the massive MIMO antenna products would be made to use with C-band frequencies, with time-division duplexing modulation in a single frequency band. He said that as upgrades extend massive MIMO use into the frequency-division duplex modulation world for LTE, the future would include multiband massive MIMO systems. He asked Zimmerman whether the weight and wind loads of those types of products would be a problem as they form a part of network upgrades in some urban sites and on towers.
CommScope is seeing a lot of interest in multiband products, Zimmerman said. “If you look across the United States, you see a big mix where some regions are mostly three or four pipes, and for right now at least, the C-band extension can just stand an active antenna unit (AAU) on an extra pipe for a given sector. Once customers get down to one or two antennas per sector, the standalone AAUs are just out of the picture. You have to use your slots for multiband products. We have come to market with a full portfolio of products for C-band. We are working with a radio OEM. We're about to deploy our first multiband product that has massive MMO built in.”
With deployments in the 3.45 GHz band coming, Zimmerman said, operators have to consider how to deploy two to three different massive MIMO radios. “You're not going to five poles or six poles,” he said. “So, standalone is where things are right now, and that's fine for right now. Two to three years from now, it's not going to work.”
Speaking of stealth poles, the flagpoles with 50-foot or 75-foot flags hanging from them, Zimmerman said that based on the anecdotal feedback from CommScope’s sales team, they represent at least 10 percent of deployments, nationwide. “It depends on what region you're in,” he said. “In some places, it's much higher.”
Zimmerman addressed the matter of wind loading and the American Society of Civil Engineers (ASCE) 7-22 Wind Loads Subcommittee’s definition of minimum design loads and associated criteria for the design of buildings and other structures. The calculation of wind loading affects the required strength of structures used to hold the antennas and radios. The more conservative the calculation, the stronger and possibly more expensive the structural requirement. In an effort to better reflect real-world conditions and reduce the strength required, CommScope has been providing data based on wind load tests, tunnel tests, for a couple years, Zimmerman said.
“That drops the number,” he said. “What we typically see, especially in certain jurisdictions, not necessarily in the United States, is people want to use the projected area. That is going to give them a bigger number. If they are designing their structures based on the bigger number, there is not going to be an issue. The challenge we have is getting people to accept our wind load values based on actual wind tunnel testing, which can be 50 percent lower, based on the fact that your wind load is specified on a very high airspeed, and most antennas at that speed are no longer in a laminar flow condition, they're in a turbulent flow. Once you get past what they call the drag crisis, the numbers just dropped like a rock. That is where you get the 50 percent reduction. If you want to use the back-of-the-envelope formula, you are going to get a bigger number. Therefore, there is not much risk. The only real risk is maybe you spend a little bit more on construction than might have been necessary.”
Lum raised the subject that active antennas consume a substantial amount of power and asked what would happen at sites where operators eventually place three massive MIMO antennas. Zimmerman confirmed that a 5G 3.5 GHz 65T massive MIMO antenna requires a maximum of 1500 watts of power.
“That's what you have to build for,” he said. “That drives your capex. Then there is your opex, which is based on your nominal typical power consumption. The bottom line is, power consumption is going up.”
Zimmerman mentioned two options:
“One is you pull out your 6 AWG power cables and replace them with 4 AWG,” he said.
“For another, CommScope has an option called PowerShift that conditions the power,” Zimmerman said. “It senses the drop and allows you to adjust the power level so you get just barely enough volts at the radio at the top of the tower. Customers are finding that it allows them to reuse their existing power lines in some cases. Therefore, you spend a little money on one thing, and you save money on ripping and replacing something else. Customers are seeing that as something that is helpful. However, the bottom line power consumption is going up, and it is not trivial.”
Regarding the placement of massive MIMO antennas relative to passive antennas on the same sector, Zimmerman said from an RF radiation perspective, the placement creates no problem. “The C-band is a higher frequency, so insertion loss is a concern,” he said.
However, if the C-band antenna is not an active antenna unit and instead is a passive antenna, Zimmerman said, it is important to place the C-band antenna closer to the radio, rather than farther away.
A problem Zimmerman cited involves the radios themselves, because the radios do not necessarily behave well when exposed to a lot of radiated energy. He said a CommScope solution developed in conjunction with a U.S. operator is a new structure that moves the radios closer to the tower.
“It's a sector frame, and it also spreads out the antennas a little bit,” Zimmerman said. “You may have the same four pipes, the same number radios and the same number of antennas, but you're moving things a little farther apart, and you're also moving the radios farther away from the antennas. One, that knocks down environmental PIM (passive intermodulation). Two, it reduces the load of the sector frame because you have less of a moment, with the radios being close to the actual tower leg.”
Lum made the observation that in many other countries where structures have only one or two mounts, antenna providers try to integrate active and passive antennas while staying within a 2.7-meter limit of the total length of the antenna solution. He asked Zimmerman whether the problem occurs in the United States and whether operators could be expected to deploy some of the active-passive antenna products.
“In Europe, operators have a much higher percentage of one or two pipe-per-sector configurations,” Zimmerman said. “Our current portfolio of active-passive solutions is really popular. In the United States, it is still under debate whether it is a niche product or a gotta-have. When we talk to customers around the country, if we are talking to somebody in a market with many of the two-antenna mounts, they are very interested in these designs. We know there are some markets actively planning-in our active-passive solutions at this point. In other markets, they will say, ‘Ninety percent — almost all our sites — are three or four pipes. It's not an issue.’”
Regarding a possible future market for single omni antennas for 5G, Zimmerman said CommScope continues to see short omnis being used for small cell applications, “but they are not like the omnis of 20 years ago, the big 20-foot stick antennas. They are very small antennas, 2 feet high, maximum, and typically with 16 to 30 ports, depending on whether it is a single-operator solution or multi-tenant solution.”
An audience member asked whether 5G speeds and latency goals can be achieved with 4 x 4 MIMO, or does it need to be higher? Zimmerman said that to achieve the necessary capacity requires a massive MIMO at minimum 8T 8R. However, with respect to latency, he said that massive MIMO will not give any advantage, and perhaps a 4T 4R could be better for the latency.
“With the internet of things (IoT), ultra-fast broadband or different 5G use cases, they might have different solutions,” Zimmerman said. “With IoT, that's where you're going to see people using 4T 4R or lower frequencies, because what really matters there is not capacity, it's coverage.”
Total Tech sponsors of the April AGL Virtual Summit included Raycap, Valmont Site Pro 1, Vertical Bridge and B+T Group. The Top Tech sponsor was Aurora Insight. Additional sponsors included NATE, Voltserver, WIA and Gap Wireless. The next AGL Virtual Summit is scheduled for Sept. 8 at 2 p.m. Eastern time. The Summit is free to attend; register here.
Don Bishop is executive editor and associate publisher.