by Mark Miller, Product Manager, L-com
Distributed Antenna Systems (DAS) are usually visible only to people like readers of this magazine who notice such things, but they play a major role in maintaining reliable voice and data communications, even in scenarios where tens of thousands of people are densely congregated. And now they’re in an increasing rapid pace of change, as the systems in huge venues for which they’re best known for have been deployed, smaller yet equally lucrative smaller systems remain, and 5G is emerging fast.
There is no better example of how much DAS can achieve than the Super Bowl, where statistics of the game aren’t the only metrics examined in detail. In the last few years, wireless carriers, equipment manufacturers, and system integrators have been competing to top the amount of traffic handled by the previous year’s event. This year at Super Bowl LIII at Mercedes-Benz Stadium in Atlanta, 24.05 TB of data was transferred using the stadium’s network, beating last year’s record-breaking 16.31 TB in Minneapolis by 47%.
There were 48,845 users of the 70,081 in attendance and at one point in the game more than 36,000 people were simultaneously using the network. To accommodate this data deluge both in the stadium required 4,000 miles of optic fiber and 1,800 wireless access points-(1,000 in the seating area (Figure 1) and 800 in the concourses), along with upgrades by wireless carriers throughout Atlanta as well.
Verizon spent $97 million, including more than 350 miles of new fiber, 30 new cell sites, 300 small cells and upgrades to 150 existing base stations and a DAS at Hartsfield Jackson Airport, and deployed antennas behind the seats in the stadium. AT&T spent $43 million in upgrades including 1,500 antennas and amplifiers in the stadium, a new DAS in the city, and additional public safety (FirstNet) capability, while T-Mobile upgraded cell sites and deployed a cloud-based network in the tailgating areas along with 300 cells in the stadium. For its part, Sprint added the equivalent of seven complete base stations in the stadium, equivalent to what a small town would require, hundreds of small cells throughout Atlanta, and massive MIMO in many areas.
Although DAS has received most of its attention in the last 15 years or so, the concept itself dates back more than three decades, well before there was a need to provide indoor wireless coverage and even before the “wireless revolution” began. It was first described in 1987 by three Bell Labs researchers who published a paper in IEEE Transactions on Communications. They realized that a single antenna could not provide wireless coverage throughout an entire building because of signal attenuation, varying propagation conditions, and multipath delay spread (the time difference between when the first and last multipath signals reach a destination). They proposed two solutions based on dividing up a target coverage area into cells.
In the first, a building would be divided into many small cells, each served by an antenna in its center with adjacent cells operating on different frequencies. In the second, the building would be covered with fewer but larger cells, each served by a distributed antenna system or radiating (“leaky”) cable meandering throughout the building.
The latter approach had the inherent benefit of eliminating signal handoffs from cell to cell, but measurements showed that either approach would dramatically reduce attenuation and multipath delay spread. The result, they wrote, would “make possible the implementation of sophisticated broad-band TDMA-type systems that are flexible, robust, and virtually building-independent.”
It didn’t take long before the first analog system was built, by Decibel Products (ultimately becoming part of Commscope), in 1989 that provided distribution of signals inside railway stations, tunnels, and other RF-resistant areas. Since then, the major DAS system developers such as Commscope, Dali Wireless, Corning, and SOLiD, have developed a wide array of passive, hybrid, and active architectures designed to meet the needs of virtually any outdoor or indoor environment.
Not surprisingly, passive systems, the first DAS architecture, are the most widely deployed throughout the world. They use coaxial cable, splitters, couplers, and other passive components to distribute the signal, and are still a mainstay of the industry (Figure 2). They’re also the least expensive to deploy as there are no active components past where the signal source is amplified and use the least amount of hardware. In addition, their antennas, such as the Model HG72708XWPPR-NF from L-com (Figure 3) are comparatively inexpensive. The cross-polarized indoor panel antenna covers 698 to 960 MHz and 1710 to 2700 MHz.
Their basic simplicity is complicated by several factors, including the possible need to run coax through plenums, which requires cables that meet more stringent building and electrical standards. In addition, the attenuation of coaxial cable increases with distance, reducing its use to smaller systems. As the wireless industry moves to higher frequencies, their use of coax will make them less appealing in the future.
Many of the problems of the passive system are mitigated using an active approach (Figure 4), in which analog signals at the input are converted to the digital domain using analog-to-digital converters, which allows the use of optical fiber or Ethernet cable rather than coax for distribution. Fiber offers the most benefits as loss is negligible, the potential for interference is dramatically reduced, and cable length is unlimited. It is also much easier to combine Wi-Fi, public safety, and other services in an active system. However, in addition to the ADC at the input, a digital-to-analog converter (DAC) is required to convert the signal back to the analog domain for transmission by the system’s remote radio units (RRUs) that replaced the simple antenna nodes used in a passive system.
As its name suggests, hybrid DAS systems use both coax and fiber for signal distribution, and the RRUs are separate from the antennas. They are less expensive than active systems because fewer RRUs are required to cover a given area as a single RRU can connect to antennas via coax throughout the coverage area. The digital system backbone is all fiber, so the length advantages of a digital system are retained, although this type of system is more complex than either active or passive approaches because both fiber and coax are used and like the active type, ADCs and DACs are required.
Finally, the most recent and advanced type of DAS architecture is the “all-digital” type, which has the overall potential benefit of simplifying the system in several ways but primarily because the signals for almost the entire system remain in the digital domain. This eliminates most of the conversion steps required by conventional DAS systems.
That is, other types of DAS architectures require that the digital input signal from the carrier’s base station be converted to analog form after which the DAS head-end distributes them to the RRUs, where they are reconverted back to analog form for transmission by the antenna over the air. To avoid using coax, the head-end converts the signals from electrical to lightwave in a process called RF over Fiber. Once they arrive at the RRU, they are reconverted to analog form, amplified, and broadcast.
Needless to say, this is a rather inefficient process as many steps are involved that could be avoided if the digital signal could remain in the digital domain without resorting to multiple conversions before being transmitted. In a digital DAS, the signal is taken directly from the carrier’s fiber backhaul and captured by a baseband unit (BBU), eliminating the base station entirely with the only conversion being at the remote nodes. The BBU mainly performs as a high-speed digital switch or router.
The all-digital DAS also begins to look a lot more like a modern cellular network in which signals remain digital from the fiber to the remote radio head at the tower top or to one of the many variants of small cells, from metro (Figure 5) to femto, depending on the requirements of the coverage environment. The digital DAS systems of the future will look more like this than their predecessors and will simply blend into the heterogeneous network environment that can consist of any service, from cellular to Wi-Fi, and typically both, along with accommodations for public safety that although not yet mandated, are certain to become essential in the future.
Oddly perhaps, one of the biggest hindrances to widespread deployment of the digital DAS is a protocol called the Common Public Radio Interface (CPRI) protocol created by Nokia, Ericcson, Huawei, NEC, and Alcatel Lucent (now part of Nokia). As its name might imply, CPRI is designed to be “common” (i.e., open), but it’s not, or at least not yet, although its mission is, according to its creators, a publicly-available specification for the key internal interface of base stations between radio equipment control and the radio equipment itself.
As the initial application of CPRI was for use with the fronthaul portion of cellular macro cells, its goal was to replace the copper or coaxial connection between the RRU on the tower and base station so that a variety of functions could be performed remotely, rather than at the base of the tower (or worse yet, on top of it). One of the advantages of CPRI is that it allows RF measurements (like PIM) and others to be made without a tower climb from virtually anywhere. By pushing some system complexity to the higher layers of the system, connecting a BBU and RRH can be performed with minimal configuration steps.
The problem with CPRI has been that manufacturers have developed their own proprietary versions of the protocol, which makes it extremely difficult for DAS vendors to use BBUs from all companies that make them. CPRI is also not a standard but rather a protocol defining the critical conditions for connectivity, control communications and interfacing transport between BBUs and RRUs. So, while equipment manufacturers adhere to its basic tenets, they encode their own unique versions of CPRI dedicated to their own hardware.
Nevertheless, CPRI is widely used by wireless carriers for their own purposes; while DAS builders have been frustrated by this “open, but not really” environment, as a true neutral host DAS system will inevitably use varying equipment from the wireless carriers. In short, CPRI is a work in progress and there is a movement toward creating different solutions that carry less baggage and provide better performance as well, since CPRI may have challenges meeting the requirements of 5G.
The Market Moves to Lower Tiers
Most of the very large venues like stadiums, entertainment facilities, and many other facilities of more than 500,00 ft.2 represent about 30% of the in-building wireless market. They either already are or will soon be enabled with sophisticated DAS installations, which leaves the so-called second-tier opportunities of which there are more, but somewhat smaller. These Tier 2 venues, which range in size from 100,000 to 500,000 ft.2, consist of a very diverse array of structures as well as operational ownership models.
Tier 1 venues are very expensive to deploy, but the cost to wireless carriers has been mitigated to some degree by other factors, such as creating a competitive (and visible) advantage, investing in an asset like a stadium in which benefits will accrue over many years, and the need to provide coverage where all other major carriers also do. Although small venues are still expensive, there has been less interest in providing coverage unless it can be monetized in some way, either directly or indirectly by, for example, offering coverage in a mall or apartment building that can be marketed as a benefit.
In these cases, carriers are far less interested from a funding perspective, which has produced a different ownership model in which cost, return on investment, scalability, and “future-proof” technologies are becoming essential. With this in mind, it’s not difficult to build a case for enterprises to create their own DAS networks in collaboration with partners that may include wireless carriers but to a lesser extent. While the carriers don’t own these neutral-host systems, the carriers retain considerable control over their operation.
Carriers also have the benefit in second-tier cases of expanding their coverage footprint by combining DAS and small cells, potentially reducing the number of costly macro-cell base stations that are far more difficult to deploy. This will only become more appealing to carriers with the rollout of 5G, in which higher-frequency operation translates directly into the need for more infrastructure, primarily delivered by small cells and DAS.
The Future of DAS
The DAS of the future will become less a stand-alone solution to in-building coverage than a complement to cellular networks and a means for providing coverage in places that would ordinarily require huge numbers of small cells. 5G will ultimately deliver on its long list of promises and it will get there faster if DAS is part of the solution.
For example, the increasing presence of IoT implementations, all of which require a means of communicating with the Internet, is generally achieved by cellular networks or Low-Power Wireless Area Networks (LPWANs) that take the data produced by sensors at the edge and send it to cloud-based data servers. Maintaining low latency, required for many applications, can only be achieved over very short distances, so wireless carriers’ small cells complemented by DAS systems will satisfy both in equal measure.
In the long term, a major challenge for DAS is accommodating the demanding specifications of 5G, especially the millimeter-wave frequencies it will soon employ to meet capacity, coverage, and throughput requirements. These new bands must be integrated with DAS systems, which while a challenge for new deployments, are likely to be well beyond what earlier passive systems can handle, especially where coaxial cable is used. One thing is certain though: DAS systems won’t become obsolete with the emergence of 5G, but rather even as important as serving Tier 1 applications.