What's Inside Counts

Look around your office and identify what affects coverage. You'd be surprised by what can help and hurt propagation.

The FCC always has been a place of uncertainty, and its move to the Portals building takes that to a new level: It's tough, if not impossible, to find a signal inside much of the building, including the commissioners' 8th-floor offices.

The FCC offices are just one of the more high-profile buildings where coverage ranges from dicey to nil. That's a shame because giving subscribers more places to talk means that they'll talk more. Some providers have begun highlighting their indoor coverage. Bell Mobility's Web site, for example, lists all the Ontario and Quebec buildings where it has coverage.

Delivering on that promise can take a lot of legwork. One way to estimate coverage inside a particular building is to look at the signal strengths immediately outside it and assume that the signal will be attenuated by 10dB to 15dB once inside. But the actual loss varies significantly thanks to variables such as building materials and even geography.

"Our engineers in the San Francisco market use about 15dB," said Steve Regitz, GTE Wireless manager, network-technology development. "The structures in San Francisco are a little different due to the earthquake zones. People on the east coast might use closer to 10dB."

Adventures in Cubeland
Although few engineers would choose cubicles for their own offices, they probably would prefer to see them in more buildings. That's because a wide-open space allows signals to propagate better than real walls do. Part of the problem is that modern office buildings use metal studs rather than wood inside walls.

"If you have cubes that are all 4 feet to 6 feet tall, that's a much easier environment for signals to get through than if everything is walled off," Regitz said. "Offices on the outside and cubes in the middle is going to cause some more problems because first you have to propagate into the building and then through a wall or two to get to the cube area."

Even office equipment can affect propagation.

"Too many metallic filing cabinets will greatly reduce the signal," said Zoran Kehler, Ericsson portfolio manager.

At higher frequencies, some wall coverings' absorption and reflection characteristics also become factors.

"I've read some Ph.D. theses that suggest you can use different types of paint characteristics to optimize the coverage: shape a wall, paint it a certain way and direct your radiation where you want it," said Malcom Oliphant, IFR strategic-marketing manager.

That strategy can be applied to elevator shafts, which can block RF or act as a helpful waveguide. Some innovative designs use elevator shafts as a cost-effective way to improve coverage.

"One experiment I saw in Frankfurt about two years ago (was) an enormous skyscraper with several elevator shafts," Oliphant said. "These guys figured a way to get pretty much universal in-building coverage strictly through the elevator shafts: They put the base stations at the top of the shafts and radiated down the shafts and out some carefully constructed portholes behind the walls. You couldn't actually see them, but they were transparent as far as the radiation was concerned. This drastically reduced the cost of the system because fewer base stations (were needed)."

Where Am I?
With that many variables, assessing in-building coverage can be complicated. Some providers do cursory testing by walking around the building with handsets in test mode and looking at signal strengths, while others do more in-depth testing on par with drive testing.

One obvious difference between indoor and outdoor testing is the gear itself: With drive testing, the main limitation is how much gear you can pack into the van. But indoors, there's a limit to what you can cart around — literally. For in-depth tests, it's not uncommon to load gear into shopping carts, projector tables and even baby carriages. Indoor test gear continues to shrink, and vendors such as Agilent and Safco now offer ergonomically friendly equipment that can be worn like a peanut vendor's tray or carried in a shoulder bag.

Another difference is identifying location. GPS signals rarely are available indoors, although recent advances suggest that enough signals can be recovered to determine location. Until that can be done in real time, one option is to use a dead-reckoning wheel, which attaches to the cart holding the gear and uses distance traveled to determine location.

Another option is to digitize the building's floor plan and feed it into the growing number of test tools that accommodate digital maps. As you walk between "landmarks," such as the end of a hallway, you tap the location on the screen, and the tool 0notes that all measurements were collected along that path.

Most buildings are private property, so floor plans aren't as readily available as maps.

"Most often we use floor plans provided by the customer to map coverage," said Jay Noceto, Cellular One - San Francisco RF-engineering director.

Another option is to get the building's floor plan from the local planning office. Whatever the source, a plan in electronic format beats writing performance parameters on a printed plan.

"Being easily able to have soft copies of those floor plans would be great," said GTE Wireless' Regitz. "Being able to have the software automatically save the signal-strength data that the test gear is receiving on top of that electronic plan would be handy because the process would be somewhat automated."

In high-rise buildings, identifying floors is helpful because coverage and interference can vary significantly from floor to floor. That's because on upper floors, the phone can "see" more sites than when it's on ground level.

"You'll see interference on the 20th floor that you wouldn't see on the ground because there's no clutter on the 20th floor to knock down the interference," Regitz said. "The floor definitely is a critical thing to know."

Some providers are including in-building coverage as part of site selection.

"We've seen them do site selection based on the data collected on various floors," said Ericsson's Kehler. "They want to know if they put their site at a certain position, what kind of coverage it provides in areas where they have high-rise buildings. In Brazil, for example, operators put requirements to test fifth floor, 10th floor, 15th floor as well as the coverage on the ground level."

Poor speech quality also can indicate poor coverage.

"With some digital technologies, you may notice that as you go inside the building, speech quality becomes worse because maybe the system wasn't designed to put good coverage inside," Kehler said. "Then your error correction kicks in, and that translates into you noticing degradation in speech quality."

Even switch statistics can help identify buildings worth testing.

"If you have access failures, that would signal that maybe the cell serving the inside of the building doesn't have enough capacity," Kehler said. "Dropped calls might mean that you have erratic coverage. If, for example, you're moving through your office, and the system tries to hand off to another cell, that would mean that your system doesn't provide good coverage in that area."