February 16, 2017 —
The cost and difficulty of providing backhaul to small cells can make the difference whether carriers would construct the access points. In the early days of cellular network construction, carriers were reluctant to share antenna space on their towers with competing carriers. Similar reluctance to share small cell installations with other carriers either provides a competitive advantage or hampers network densification, depending on one’s point of view.
Randall Schwartz, a senior analyst and consultant with Wireless 20/20, led a session about small cell connectivity at the Tower & Small Cell Summit in September 2016. He said there’s no doubt about the importance of small cells and the importance for operators to densify their networks. One possibility he cited has macro networks ultimately disappearing because they could cause so much interference to small cells that they would become a hindrance.
Ron Mudry, president of UnitiFiber, spoke as one of the session panelists. UnitiFiber, a unit of Communications Sales & Leasing, acquired Mudry’s former company, Tower Cloud. UnitiFiber also acquired PEG Bandwidth. Mudry said Uniti Fiber provides fiber backhaul for macro sites. More recently, the company started providing service to small cells, initially with only fronthaul fiber connections for the antennas. The company now offers turnkey installations for the carriers.
Regarding most of the small cells Uniti Fiber serves or constructs, Mudry said they boost network capacity, instead of filling coverage gaps or extending coverage to new areas. “We’re seeing small cell construction for capacity increases in all of our markets,” he said. “We operate in 19 states, and we cover a large Tier 1 market. Everyone talks about dense urban areas that need small cells, and those areas are seeing a lot of activity.”
Mudry said Uniti Fiber also is deploying large numbers of small cells in Tier 2 and Tier 3 markets in some rural areas with populations of 150,000 to 200,000 people, and on military bases. “What surprised me the most is that densification is taking place across all of the markets, although maybe not to the same degree,” Mudry said. Meanwhile, Mudry said carriers speak to the need to reduce the cost of small cell deployment, directly relating the number of small cells they would deploy to the cost.
To link radio equipment control devices with the radio equipment, Mudry said carriers use the Common Public Radio Interface specification over dark fiber — and that fiber connections can be expensive to build. He said carriers face a multifaceted problem involving the network design, small cell design and what local governments will allow.
“The engineers and the wireless carriers perform an RF analysis to determine where they want to put the small cell, and then a further analysis to design what small cell configuration they want,” Mudry said. “These first two steps are technical. The third stop involves site acquisition to obtain permitting and zoning for the site. What’s possible to get permitted and zoned can be out of step with what they’re looking to achieve for the network. It kind of boils down to having a match between what you can get done in the jurisdiction versus what they’re trying to do technically with the network. We see a lot of cycles wasted where a design will come out, either a fiber design or an RF design — both have issues when they’re looking at a small cell, centralized radio access network (C-RAN) design — and the technical solution that the customer wants just isn’t feasible to get it permitted and zoned.”
In a specific C-RAN design that Uniti Fiber prepared, the hub location was placed where it was difficult to extend fiber. “The cost of bringing the connection was really the driver that broke that business case,” Mudry said. “In other scenarios, they want to put in a cabinet or a certain pole height, but the particular location selected doesn’t allow it because of the zoning. The zoning can change by street. The zoning one street over can be different from the current street.”
Time to Deploy
Mudry said carriers consider more than the cost of small cell installation in choosing their deployments, they also consider the time it takes to deploy. “It’s not always the time from when you have the contract and you’re saying, ‘Go, let’s deploy it,’ but it’s actually the planning time,” he said. “A lot of unproductive time is being spent for lack of understanding what can be done in a particular jurisdiction. Many jurisdictions are in the process of evaluating small cells and making the rules for them, so it’s a difficult thing to get ahead of.”
A lot of fiber has been built to provide Ethernet backhaul to macro towers in most of the United States, although Mudry said the fiber is not always where it’s needed. Also, connecting small cells using the CPRI specification requires a high fiber count, for the most part, to have enough to dedicate a pair of fibers to each antenna.
“Once your fiber run goes past that pole that you’re dropping those first two fibers at, you’ve stranded or wasted the rest of that infrastructure in the ring,” Mudry said. “What we’ve been doing is looking at our fiber design to change the way we can reuse fiber around the ring. The traditional fiber company doesn’t like to cut off its fiber rings because it needs the diversity for reliability and so forth. But in a small-cell scenario, you’re using it on a point-to-point dark fiber connection. We try to put in more frequent handhole access so that we don’t have to build a long fiber route to get to the handhole where we’re going to splice the small cell in, and we can actually get the splice closer to the location of the antenna. We’re using multiple conduit so that we can dedicate conduit to the small cell network and use it as an access on-ramp to the larger backbone.”
Extending fiber to some of the required small cell locations calls for some new greenfield build, Mudry said. He said being able to blend the greenfield capital cost with other uses of the fiber network versus having it only dedicated to the small cell would change the price point for providing backhaul and fronthaul to the antenna.
Mudry said that when Uniti Fiber talks with carriers about various projects, they express reluctance to become involved in neutral-host small cell installations because they want something different from their competition. They don’t want to have the same coverage, or they don’t want to enable their competitors to come along later and join that same system, taking advantage of the first carrier’s work.
In some cases, carriers install a neutral-host system if it’s related to building rights, contractual rights or another signed agreement, and if the first carrier is going to be the operator for all the carriers in that venue. Mudry said he believes using a shared infrastructure is important, at least on the fiber-connected systems, even if carriers use different locations for the antennas or if they use different poles. “If it’s just a pole right down the street or next to the other one, they can still share a lot of the cost of the fiber system serving multiple carriers, even though they have individually negotiated transactions,” he said.
Mudry said small cell economics would motivate carriers to share infrastructure, especially as long as fiber is a main component. With sharing comes standardization, which Mudry said also brings down the cost and speeds deployment.
Wireless as a Necessity
Although hotels seem to view providing wireless communications for guests as a necessity, Mudry said he sees different perspectives from various cities and other jurisdictions. “Some see regulating wireless as a revenue opportunity, and they want to charge for the right of way or something to install small cells,” he said. “Other cities take the view that if they don’t have advance wireless communications available, they won’t be able to attract businesses or retain businesses. As with venues such as hotels, cities and the other jurisdictions that, with more enlightened leadership, are realizing how important wireless is to their communities and how they benefit from it. So, they’re going to set standards that favor infrastructure sharing. They’re not going to want the streets dug up two, three or four times to deploy small cells for everyone. They won’t want each carrier to have antennas on individual poles. They’re going to want small cell construction to be restricted.”
Where to Start
Mudry said those deploying small cells need to improve the information RF planners have for designing networks at the outset. The information comes from local markets. “You have to understand what can get done in a certain city and how to go about doing it, rather than finding out while you’re in the middle of the project after you already have the design in place,” he said. “When that happens, you have to start over and reconsider the design. Learning what the city will allow or what can be done in the particular jurisdiction has to be your starting point when you design your solution.”
DALLAS — May 25, 2016 — One of the sessions at the Wireless Infrastructure Show dealt with emerging technologies and their implications for wireless infrastructure. Ron Mudry, president of Tower Cloud, led the session. He described how backhaul is evolving to meet the needs of emerging technologies, saying there is adequate backhaul in most markets and even in rural areas.
“What we’ve seen lately is Verizon moving from lit service to dark fiber,” Mudry said. “Many dark fiber builds are going on around the country. It’s a big initiative that is putting a lot of infrastructure in the ground. The previous fiber build is nearly 20 years old, dating to the dot-com era.”
Mudry said wireless carriers have been densifying their networks because of capacity constraints. “They’ve added a lot of macro towers,” he said. “That’s provided a lot of the growth for backhaul providers and others in the industry. Now we’re seeing that shift a little to bringing capacity with small cells and mini-macros and centralized radio access network (C-RAN) technology.”
Dr. Rikin Thakker, a research assistant professor at the University of Maryland, said that cellular network operators have enough RF spectrum to serve their networks, for now. He said that operators say they need more spectrum because of a forecast rise in data demand. But research indicates other substitutes for spectrum.
“Macrosites are not going away, even though we are talking about the Internet of Things, 5G cellular technology and small cells,” Thakker said. “Macrosites will play an important role, and that could be a good substitute. Increases in efficiency with technology decrease the burden on spectrum. Wi-Fi offloading has kept the demand on licensed spectrum lower. Just increasing macrosites by 5 percent could lower the need for licensed spectrum by 98 megahertz.”
Aaron Blazer, a senior partner at Atlantic ACM said the network operators’ end-user revenue comes under pressure as competition increases. The result trickles down into infrastructure. “Operators pay attention to operating expense and the ability to deploy capital on infrastructure,” he said. “When spectrum is tapped, you look for the most efficient way to boost the network. Deploying more macrosites is a business model that carriers understand. The economics of backhaul and macrosites are well understood.”
Blazer said that when macrosites aren’t enough, non-macro densification emerges in the form of small cells and outdoor distributed antenna system (DAS) networks. He said another alternative is C-RAN technology, where operators use remote radio heads with a centralized baseband unit to make more efficient use of spectrum. He explained that a heavy fiber component changes the cost structure, especially a dark fiber component, and sometimes fiber is not available.
“After that, we see operators looking to Wi-Fi and other offloading strategies to support the network,” Blazer said. “But Wi-Fi comes third because it is not always seen as a carrier-grade technology.”
Rich Grimes, the chief operating officer of the DAS and Small Cell Group at InSite Wireless, said the carrier market for in-building DAS is finite. According to Grimes, from a carrier perspective, venue revenue-sharing is questionable. He said there is higher scrutiny for lower-capacity venues, and more cost-effective solutions will be used.
“In the forecast for DAS, capital spending for this year is pegged at about $4.8 billion and rising about 28 percent per year to more than $16 billion in 2020,” Grimes said. “A focus we’re all seeing is on reduced cost for in-building wireless systems. Also, fiber will become increasingly available to commercial buildings, and third parties in the enterprise will take a greater role in deploying DAS with the carriers’ focus being more on the capex for the LTE-Advanced roll out and small cell preparation for 5G.”
Kishore Raja, director of strategic programs at Boingo Wireless, categorized emerging technologies in three domains.
“Number one is the process of natural evolution within the licensed spectrum,” he said. “You have macro towers, and you have DAS, which augments existing towers. You have small cells, which augment by adding capacity and coverage. Number two is emerging technologies on unlicensed spectrum, such as seamless Wi-Fi access to networks. Number three is emerging technologies in the area that bridges licensed and unlicensed spectrum, such as LTE-U[unlicensed], LAA [License Assisted Access], LWA [LTE – Wi-Fi Link Aggregation] and muLTEfire. MuLTEfire provides LTE-like performance with Wi-Fi-like simplicity.”
Robert Long, director of sales at Crown Castle International, said that regardless of the path it takes, the need for more infrastructure will continue. “By 2018, 4G data use is expected to increase by a factor of 10,” he said. “Cell phone data use will increase by a multiple of six. Add the Internet of Things, smart cities and autonomous vehicles. Providing a solution that’s sharable, whether it’s fiber, towers or small cells, if it’s sharable, it’s much more economical for the service providers.”
From the pages of AGL Magazine
Part Two — For Tower Cloud, the migration from Ethernet to dark fiber readies wireless carriers for 5G and supports continued use of macro towers as the foundation for mobile networks.
March 15, 2016 — Tower Cloud is a backhaul specialist that started when T1 1.544-Mbps dedicated digital data transmission lines from telephone companies were the main transport used for cell tower backhaul. Once fourth-generation (4G) broadband mobile telecommunications initiatives started, Tower Cloud grew and increasingly built fiber connections to towers, providing Ethernet backhaul. The company serves Georgia, Alabama, Florida and South Carolina, including large cities such as Atlanta, medium-sized cities and rural areas. Tower Cloud built its business during the transition from T1 circuits to lit Ethernet.
Ron Mudry, the company’s founder, president and CEO, said Tower Cloud is increasingly serving small cells and other businesses to load its fiber network. His experience includes starting Progress Telecom in 1998 to provide fiber and backhaul services. He led negotiations to sell the company to Level 3 Communications.
In its early days, Tower Cloud used microwave for as much as 20 percent of its connections and still uses microwave in some rural areas. Today, Mudry sees macro tower backhaul moving more toward the use of fiber, other than in extremely rural areas. Mudry spoke at the Tower & Small Cell Summit session “Macro Tower Backhaul Solutions” led by Jennifer P. Clark, vice president of research at 451 Research, a New York-based market research and consulting company.
The past five years has seen backhaul rotating existing T1 sites onto Ethernet service, a task that Mudry said the wireless carriers have done effectively. The current cycle of macrosite densification has some carriers adding more sites than others, along with migrating existing sites from lit Ethernet to dark fiber, which is fiber leased by individuals or companies to use in establishing connections among their own locations. Mudry said the shift will change the way macro backhaul networks are operated and how the use of small cells will grow. “Macrocell site backhaul is not separate from the small cell backhaul,” he said. “They’re going to be integrated, and what happens in small cell is going to affect what happens in the macro side.”
Another aspect Mudry identified as important is the forthcoming renewal cycle for fiber leases. What began as long-time T1 contracts and continued as Ethernet contracts usually have five-to-seven-year terms, and they’re beginning to expire. Meanwhile, carriers are focused on what to do with fiber network providers such as Tower Cloud to expand macro backhaul and cell site densification, and whether to renew with their current backhaul providers.
Mudry said the relative quality of backhaul service will have much more effect during carriers’ reevaluation of their backhaul providers. The selection of backhaul providers during initial deployments emphasizes coverage and cost. Mudry said some of the initial providers aren’t performing up to the standards that the carriers want. “You’re going to see some migration of the existing base because of quality issues,” he said. “If you can perform well, you can take market share from your competitor.”
Macrocell site backhaul remains highly important for fiber providers, and Mudry said the embedded business is growing about 8 percent per year. “It doesn’t sound like a large percentage growth, but it’s on a huge base,” he said. “It’s a big revenue opportunity, and we’re trying to get as much of that as we can.”
Mudry said carriers are spending more money on macrocell site backhaul than small cell backhaul. “It’s not the huge growth opportunity that it was when it shifted from T1 to Ethernet,” he said. “Nevertheless, I see new towers coming in. Density is a big growth opportunity.”
The shift from Ethernet to dark fiber will enable the building of more fiber infrastructure, and Mudry said that in this growth phase, the fiber network can be designed with backhaul and wireless networks in mind. He said that much of the initial wave of fiber construction involved networks built for enterprise services, carrier hotels and data centers. The fiber didn’t extend to the right places for cell site backhaul. “A migration from lit service to dark service by one or more of the large carriers is going to change the fiber landscape in many of these markets, and wireless service is going to be a big part of that,” he said.
Bandwidth use per tower basis is growing. Mudry said it’s near an average of 100 MBps per tower today, and some towers at aggregation points use 300 Mbps or 500 Mbps. “Carriers are now talking about 1 GBps per site,” he said. “They’re not spending all this money on dark fiber because they believe bandwidth growth is flat. They believe it will skyrocket. Having their mobile network sites talking to each other more on the edge means a faster network with less latency and more direct connection.”
For Mudry, the macro world is still the heart of the network wireless carriers. How that part of the business is managed is going to enable the migration to small cell backhaul and whatever happens with fifth-generation (5G) cellular when it comes.
Microwave for Backhaul
Although Tower Cloud used microwave for backhaul to a greater extent in the beginning, Mudry said that, in the long run, many of those microwave legs should have been fiber to begin with. Using microwave comes with more expense because of paying tower rent, usually on both ends. Microwave is not quite as scalable as fiber and isn’t as easy to tap into. Because of the savings on operating expense, it’s affordable to invest more capital in fiber to reach a particular site.
The capacity growth of microwave is there on paper, but in practice, Tower Cloud hasn’t seen it scale as rapidly as necessary. Although it’s possible to handle the growth of one carrier with microwave, it becomes difficult to add a second or a third carrier or to add another tower down the line, which requires aggregating the data traffic through the intermediate tower. Then, the initial length becomes constrained. For this reason, Mudry said bringing fiber deeper into an area makes sense in the long run. “The carriers don’t want to be replacing these backhaul lengths every five years, and we don’t, either,” he said.
In Tower Cloud’s experience, if a cell site doesn’t meet the economics for fiber, the carriers aren’t willing to pay more. “In metro areas where data traffic is growing, fiber is going to be the choice,” he said. “But if you can’t get it done with fiber, the carriers are likely to use their own microwave or just stay with the local exchange carrier until they can get a better solution.”
Mudry said that microwave will have a place in backhaul in lower-capacity growth areas where it’s there for coverage purposes. Where the capacity is growing, bandwidth is growing, and backhaul providers will need fiber for the capacity. “It’s easier to add sites into fiber than to tap into a point-to point-microwave length. Fiber is more flexible for future growth.”
Nevertheless, once a microwave length is in place, its owner doesn’t usually take it down as long as its meeting requirements. “Unless it needs to be overbuilt for capacity reasons, why fix it if it’s not broken?” he asked.
Mudry said that after the evolution to Ethernet backhaul, carriers mostly were looking for availability — is the service up or is it down? Now, there’s more focus on latency (the interval between initiating a transmission and receiving or detecting the result) and other performance factors. Measuring them is essential to revealing service quality. He mentioned the carriers’ wanting a live portal, which would offer real-time viewing of the fiber network’s performance statistics via Internet access.
“All the carriers would love to have a live portal,” Mudry said. “I’d love to have a live portal. I can go to my network operating center and see it, of course. I don’t think it’s a table stake.” [A table stake refers to a minimum level of investment or technology.] “Many carriers have been talking about portals since I have been in this business for 20 years, and most service providers never get there with them because they’re just not a table stake. The carriers know whether the fiber network is performing or not without a portal. You do have to provide the monthly reports and have meetings with the carriers. But the most important thing to the carrier is what you do when there’s a problem.”
Mudry framed the situation with questions: “How do you handle the outage? How do you report about it after an incident has occurred? It doesn’t have to be a fiber cut or an outage. It can be a packet loss problem in a particular area, a node getting confused or something like that.” In those situations, Mudry said carriers want fiber providers to reveal more detail and demonstrate responsibility.
“Carriers want to see that you’re not pointing a finger the other way and you’re doing something about it,” he said. “That’s the kind of reporting that differentiates between the high-quality performers and the ones that are so-so in the minds of the carriers.”
Effect of 5G
Wireless carriers continue to move the network intelligence (its processing capability) closer to the edge of the network. In Mudry’s view, 5G is going to be about centralized radio access networks C-RAN and other steps that enable a high degree of local connectivity. An example would be interaction between small cells and macrocells, using connections with low latency. It will increase the need for fast connections at the network edge, which will provide motivation for making fiber enhancement to the network.
In fiber networks, latency affects transporting the data traffic from the cell site to the mobile network’s switch, but Mudry said latency isn’t a problem when the switch is in the same market as the network tower. But intercity fiber routes can be too long to be useful. For example, if a macro tower served by a network switch is 200 miles away, it can cause latency problems. “Even today, the carriers’ one-way latency standards are 5 milliseconds to 8 milliseconds,” Mudry said. “When they talk about 5G, I hear talk of a need for 1 millisecond of latency.”
Light moves very fast, but it still takes a certain amount of time to cover a given span, and Mudry said that connections add to the time. Every connection box through which the fiber passes adds latency. “The answer is to take the intelligence further out in the network so the data doesn’t have to travel all the way back to the switch,” he said. “That’s going to increase the need not only for small cell aggregation hubs at the macros that are going to be talking to small cells, but you’re going to see more edge data centers and little places out in the network where they’re aggregating things and feeding video, and doing a lot of things that the wireline side of the business is doing today to improve the customer experience for people over the top.”
Over the top refers to the delivery of audio, video and other media over the Internet without the involvement of a multiple-system operator such as a cable TV service provider. “It’s obviously coming as a big service, and that’s going to come to the wireless networks equally and it’s going to cause more need for intelligence, collocation, and fiber connectivity out of the edge that is very fast from a low latency perspective,” Mudry said.
In conclusion, Mudry said macrocell site backhaul is a growing business. “There’s a big and better base there,” he said. “The bandwidth is growing. The migration to dark fiber is growing. The migration to dark fiber is probably the biggest question of what’s going to happen over the next couple of years on the macro side. Will all the carriers go that way? How far will the carriers take it that are going to dark fiber? Is it just going to be the big markets or will they go further? The further macro backhaul goes with dark fiber, the more it’s going to provide the foundation for everything else the carriers are trying to do with small cells and other things in 5G. That’s probably the most exciting thing that we see happening in the macro space right now.”