DAS: The Ultimate Small Cell

Small cells are getting a lot of attention as a solution to coverage and capacity challenges in buildings and urban areas. But long before small cells (femtocells, picocells, metrocells, and the like — we’ll call them picocells for the sake of simplicity) were invented, DAS was providing pinpoint mobile coverage and capacity to buildings and outdoor areas. In many ways, DAS was the original small cell. In this article, we’ll compare pros and cons of DAS and picocells and see how they can work together to provide optimum mobile services.

DAS Pros and Cons

A DAS works by carrying a signal over fiber or CATV cable from a base station and head-end, through managed hubs, and out to remote antennas. A DAS can scale from smaller buildings to very large venues such as airports and stadiums, urban canyons, and suburban areas. Fiber-fed DAS can stretch 15 miles or more.

In addition, a DAS can carry multiple frequency bands and can support multiple carriers with a single set of electronics, antennas and fiber or cable connections. Advanced DAS products can support up to 8 different frequency bands with a single set of head-end and expansion hub equipment.

Also, a DAS provides consistent, uniform coverage and capacity from every antenna point, creating a
blanket of coverage that eliminates cell phone handoffs. (See Figure 1.)

Figure 1. A DAS creates a uniform blanket of coverage with centralized backhaul.

Finally, a DAS has a single, centralized backhaul connection at the head-end; and a well-designed DAS is future-proof in that it is able to accept new frequencies and services without upgrades to the remote antennas.

The drawback to a DAS is that it can be expensive, particularly when it is supporting a limited number of frequencies or operators. A DAS is an infrastructure investment that, depending on the size of the venue, can cost hundreds of thousands of dollars.

Another potential drawback to DAS: for larger venues it can take a significant amount of time to deploy. The cable infrastructure required to support a DAS (assuming new cable must be installed) can be extensive and installation timeframes are often lengthy. In addition, the mobile operator(s) must be involved to supply the initial RF signal (either through a roof-mounted donor antenna or through a base station deployed on site), and in the instance where a base station is required, preparing the necessary space can take weeks.

Finally, DAS deployment can be disruptive as workers remove ceiling panels, install conduit, or perform other tasks related to the installation.

Picocell Pros and Cons

In contrast, picocells are small, self-contained radio units that mount on a wall or ceiling and provide mobile services to the immediate area. They are comparatively inexpensive and easy to deploy. Given a mobile operator’s permission to use them, picocells can be deployed in a day or two in a small- to mid-size venue.

Moreover, picocells can use Ethernet backhaul through the enterprise network, rather than requiring the dedicated backhaul of a DAS.

Finally, picocells are best at filling in coverage gaps in specific areas of a building, similar to a Wi-Fi "hotspot" application.

Still, there are drawbacks in terms of how picocells provide service.

While DAS was designed to scale, the picocell was designed to deliver coverage and capacity over a relatively small area, similar to a Wi-Fi access point (AP). In a wireless LAN, if architects want to cover a larger area, they can deploy multiple APs to facilitate this. With picocells, the idea is that carriers or enterprises can deploy multiple picocells in the same way to cover large buildings or outdoor areas.

But the mobile network is not a private, unlicensed wireless LAN, and using multiple picocells to provide coverage over a wide area raises certain issues. In small buildings or residences, picocells make sense because the coverage is matched to their capabilities, but deploying multiple picocells in a larger venue poses problems.

One main issue is figuring out how to reduce interference with adjacent cell towers and other picocells. When such interference occurs, network performance is inevitably reduced. Making those multiple picocells work well together will take a lot of RF engineering, and it won’t be easy. Even in an open industrial space, the delicate interaction of multiple picocells can be thrown off when a piece of equipment is moved and thereby changes the RF characteristics of the area.

Avoiding interference with the outside macro network is another problem. To ensure sufficient coverage along the inside building perimeter, picocells are placed close to the outer walls. However, due to these cells’ proximity to the building exterior, a certain amount of interference between those cells and the macro network is almost unavoidable. To minimize picocell interference, some mobile operators are considering deploying picocells on a different (perhaps dedicated) RF frequency from macro cellular networks, but this chews up precious spectrum at a time when spectrum is scarce and acquisition of additional spectrum assets is costly.

A third challenge with picocells is that today they are single-frequency, single-service devices. To deploy more than one frequency or mobile operator’s services, it will be necessary to deploy multiple picocells in each location where they are distributed. Since each picocell will require its own backhaul and power, this can be a complicated and expensive arrangement. Manufacturers will soon release picocells that support multiple frequencies, but planners will likely still need to deploy a separate picocell for every mobile operator to be supported. (See Figure 2.)

Figure 2. Distributed picocells require individualized backhaul.

In a multi-picocell environment, handoffs can be problematic. In such an environment, the user’s handset has to hand off the connection from one cell to the next as the user moves through the building. This drains handset battery life. This occurs with far greater frequency than in the macro network due to the picocells’ small cell size (coverage areas typically 5,000 square feet or less). In buildings covered by a DAS, in contrast, there are no handoffs as the entire area is essentially one large cell.

Signal dominance is tricky as well. To function properly, an in-building system must establish signal dominance in order to minimize the potential for hand-offs between the indoor and outdoor signal sources (this is particularly critical in high rises). Such hunting can adversely impact the surrounding macro performance, as well as reduce device battery life and create a poor user experience. But it is difficult to establish a dominant signal source with picocells due to their very low output power. With a DAS, however, it is easy to deliver enough power through the antennas to create a dominant signal source and minimize hunting.

Peak traffic engineering is the final challenge. This refers to the need for high user density locations (conference rooms, cafeterias, and other communal areas) to be over-provisioned with picocells in order to provide enough capacity for peak usage times. During low usage this investment in extra picocells (and capacity) is effectively wasted. Over-provisioning issues are more effectively managed with a DAS because all of the system’s capacity is available to every antenna within the coverage area. As such, there is no need to account for movement of people/devices throughout the day. And with a DAS, if capacity needs increase in the future, additional radios or base stations can be easily added at the DAS head-end to increase capacity throughout the coverage area rather than deploying multiple picocells in distributed locations.

Blending DAS and Picocells

One way to overcome the limitations of using picocells alone while providing strong and consistent mobile service in the enterprise is to combine these devices with a DAS: picocells provide the capacity at the DAS head-end, and the DAS distributes it throughout the building. DAS that can accept baseband signals from picocells can provide a solution in such a scenario.

Six advantages of this approach are highlighted below:

Advantage #1. DAS is multi-frequency. A DAS can distribute multiple cellular frequencies to serve more than one mobile operator, so just one set of remote antennas is required, rather than multiple picocells in each location.

Advantage #2. There is no interference. Since the DAS simulcasts radio channels throughout the building, it is one large cell. This eliminates multi-cell interference along with the need to hand off from one cell to the next as the user moves about.

Advantage #3. There is one dominant signal. One signal source means one dominant signal. The DAS simply provides a uniformly strong signal throughout the interior of a building so user devices don’t hunt between signal sources.

Advantage #4. There is no need to over-provision. All antennas in the DAS have access to all of the feeder cell’s capacity, so there is no need to add new picocells for higher capacity requirements in certain areas. If additional capacity is needed throughout the building, additional picocells or radios can be added in a central location at the DAS head-end.

Advantage #5. Operating expenses are lower. A DAS needs little maintenance once deployed. With multiple small cells, the cells will require continuous adjustment to function in an optimal manner. In addition, using a picocell as the RF source for a DAS eliminates having to use a much more expensive, full-sized base station. Full-sized base stations require a lot of space, power, and cooling to operate, and using them with a DAS is overkill because much of their output power must be attenuated.

Advantage #6. Backhaul costs are lower. A group of picocells centrally located feeding a DAS head-end can be combined to use a single backhaul connection. This contrasts favorably the need for separate backhaul connections.

There is much positive talk about picocells in small cell networks. And for the right application, they are an effective solution. That said, DAS was the original small cell and has many advantages over distributed networks of picocells. The challenges outlined above make distributed picocells a less attractive solution for coverage over a large area. A far better tactic is to combine the benefits of these devices and DAS to get lower costs, easier deployment, better quality of service, multi-carrier coverage, and a future-proof infrastructure for wireless services.

John Spindler is Director, Product Management for TE Connectivity’s Wireless Business Unit. During his more than 20 years of industry experience, Spindler has held a variety of product management positions with companies such as Nortel Networks, GTE and InteCom. In these positions, he had responsibility for the areas of networking, network management, computer telephony integration and wireless technologies. He can be reached via email: [email protected]. For more information, please visit www.te.com/das.