Optimizing In-Building Coverage

In-building coverage represents the next great frontier for cellular, PCS and E-SMR industries. Like all frontiers, in-building coverage is full of promises and problems. For service providers, the promise is an expanded subscriber base, and increased airtime and revenues. More than ever before, wireless phone use involves in-building calls, especially as wireless local loop (WLL) services are rolled out worldwide. The biggest problem concerns how wireless service providers will deliver on that promise. Wireless operators generally lack the design methodologies, staffing models and tool sets needed to guarantee universal signal coverage for most buildings in their markets

Spotty Coverage Unfortunately, wireless in-building coverage is almost universally spotty. This can't be taken for granted. Wireless service providers are letting millions in potential airtime revenues slip through their fingers because of poor signal coverage inside office buildings, convention centers, shopping malls, airports and other commercial establishments.

How much potential revenue are you losing? You won't know until you survey your market with in-building, RF-measurement tools and sniff out dead zones that need adjustments to outdoor base stations or installation of indoor repeaters or microcells. But surveying in-building RF coverage in wireless markets is a daunting challenge for several reasons.

First, wireless carriers need to rethink their network designs and frequency plans to serve in-building environments. The industry's standard RF-design methodologies have been geared to vehicular phone-usage scenarios but have not been well suited to the tricky indoor environment of cinder blocks, concrete and glass. Consider the diversity, multiplicity and density of materials through which RF signals propagate inside a building. Add the fact that only a few feet of open space often separate these materials. Consider the many sharply contrasting planes and angles off of which signals bounce. Finally, throw in the characteristics of an environment served by microcells.

When you weigh the effects of these variables, you realize that an RF signal easily can encounter 50 completely different morphological characteristics, change direction innumerable times and be handed off more than once before traveling 50 feet in a typical office building.

Another in-building challenge lies in the capability to provide universal coverage assurance. Wireless carriers have to survey and optimize many buildings in their markets, including each building's floors, corridors and rooms. Measurement and optimization would, in theory, have to be performed on an ongoing basis as wireless networks are reconfigured, in-building microcells are installed and moved, existing buildings are remodeled, new buildings are constructed, and weather conditions change.

No wireless service provider has the massive technical field force necessary to undertake an ongoing, in-building RF survey of this magnitude. Even if the trained technical staff members were available, operators would want them to do most of their RF measurement from their cars so they could cover more territory on their daily rounds and attend to other responsibilities such as installing and fixing cell-site equipment. In-building RF surveys are a much less efficient use of technical staff because they are more labor-intensive, time consuming and geographically constrained than drive-testing.

>From a logistical point of view, if you can't survey every room in every >building, you can at least cover the high-visibility properties frequented >by many business and residential cellular phone users. Wireless operators >and large customers (owners and managers of large office buildings and >shopping malls) continually need to flag the most egregious and/or >critical in-building dead zones and work out mutually advantageous >solutions (microcells, repeaters) to assure seamless coverage.

Drive-Testing Tools Wireless service providers tend to shy away from in-building measurement because generally they must acquire a whole new generation of tools specifically geared to indoor environments. Traditional drive-testing tools are unsuited for in-building measurements because of their excessive size, complexity, power requirements and because they rely on the global positioning system (GPS).

The biggest obstacle to in-building optimization has been that navigation and positioning techniques aren't as automatic and transparent as GPS or TravelPilot are. RF engineering depends on the ability to accurately measure radio frequencies and to identify precisely where measurements have been taken. With measurements attached to positional data, engineers can conduct a spatial analysis of all key RF parameters. This allows them to identify patterns such as coverage holes or areas of excessive handoff activity quickly.

When wireless-engineering-tool manufacturers first started to develop field-measurement equipment for vehicular environments, they were able to take advantage of two maturing technologies: GPS and geographic information systems (GISs). GPS sensors determine position based on signals received from GPS satellites deployed by the U.S. Department of Defense. Signal use is free and sensors, offered by numerous vendors, are inexpensive. GISs combine map display and manipulation tools with databases designed specifically for geographic data. When wireless system measurements, each tagged with a latitude/longitude coordinate, were dumped into a carefully tailored GIS, the software would display all key RF parameters in their proper geographic relationship.

When RF engineers first realized the need for indoor wireless measurements, they attacked the problem by trying to adapt proven vehicle-based solutions to pedestrian environments. The two primary stumbling blocks were portability and access to positional information (enabling spatial analysis). Vehicle-based field measurement tools can occupy the back of a van or a car trunk, but pedestrian tools must be light enough to carry. Vehicle-based tools receive positional information from GPS, but GPS signals do not penetrate buildings.

Past attempts at addressing these navigation or positioning issues have ranged from comical to promising:

* Shopping cart solution. The goals of field measurement vary depending on the phases of system construction. During site selection and construction, engineers use scanning receivers to assess system coverage. When service is available, engineers need data from phones in test mode (follow-phone data) and data from scanning receivers to evaluate and optimize system performance.

Engineers responsible for ensuring coverage in Giants Stadium in the Meadowlands of New Jersey simply transferred essential elements of vehicle-based measurement tools to a smaller vehicle. Shopping carts with car batteries and LCC's CelluMATE tools collected measurements as they were wheeled through the stadium.

Having solved the portability problem, the engineers addressed the issue of attaching positional data to measurements. Most field- measurement collection and analysis tools offer markers, non-measurement records posted to data files when an engineer detects a noteworthy condition such as noise or cross-talk on the phone. Markers enable engineers to locate and analyze follow-phone reports collected near the time of the condition.

AT&T Wireless engineers employed markers as a pseudo-GPS by maintaining a list of marker numbers and corresponding Giants Stadium locations. Spatial analysis of the resulting data was limited to identifying marker locations on paper maps and crudely interpolating measurements between data file markers. Later generations of this solution reportedly included a dead reckoning cart, which monitored wheel rotation to determine location.

* Ghostbuster solution. A different approach to indoor positioning relied on a laser range-finder with a fluxgate compass to automatically calculate distance and bearing of indoor data-collection routes. Safco's Walkabout portable coverage system displays a floor plan on which handovers are depicted as symbols on collection routes. Competitors jokingly refer to Walkabout portable coverage systems as Ghostbusters because engineers are equipped with range-finders, pen-based tablet PCs and backpacks containing measurement equipment.

* Position-less solution. Eliminating positioning for in-building measurements is the approach ZK Celltest's ZK-SAM Portable is taking. It displays the signal strength of user-selected channels or basic call- tracking parameters.

* Relative positioning solution. Relative positioning addresses the positioning problem by using a CAD drawing as the reference system in which the user location click (positional input) is interpreted. Availability of and access to CAD drawings, however, remains an issue. Both geographic and CAD-based coordinate systems are absolute reference systems with consistent, mathematically defined relationships (scales) to entities being modeled. At first glance, absolute reference systems seemed the obvious choice for field-measurement positioning and analysis. However, another way would be to attack the indoor-measurement positioning problem by using a relative reference system.

Relative positioning begins with the assumption that precise, scaled data is not important. Instead, truly relevant information has to do with the general location in the stairwell, outside the elevator, in the office of a measurement. This changes the rules for attaching positioning data to measurements. In a relative positioning system, you can sketch a rectangle on the computer and declare it a hallway. Then, by clicking on this rectangle, all measurements will be assigned to that location automatically. You become the source of intelligence for locating measurements.

This discussion assumes drive-testing and walk-testing as distinct activities that never overlap. This is a false assumption for three principal reasons. First, RF-signal coverage problems, because of poorly designed outdoor cells, sometimes cover an entire district or neighborhood and should be measured in a continuous path from outside to in-building. Second, some vehicle-inaccessible outdoor environments, such as open-air stadiums, can be surveyed only with portable tools. Third, wireless phone calls should hand off transparently when moving from outdoors to indoors, and measurement tools must follow in the precise footsteps of the end user.

Ideally, a field-measurement tool would use GPS for outdoor navigation and automatically switch over to relative positioning when a GPS signal is unavailable for an extended period of time. Users could be prompted to input or sketch a floor plan of the navigation-less terrain or simply to input obvious coordinates (such as room and floor numbers) as commentary fields. The tool also might continue to gather position-less data and default to the assumption that the user is walking in the same direction at the same pace as when GPS was available. The user might even be prompted, once GPS signals are available, to input latitude-longitude coordinates for key points on the perimeter of the position-less area.

This type of creative, flexible positioning technique will be required once the RF field measurement industry takes the bold leap to amphibious outdoor/indoor measurement tools. The wireless industry's goal must be to deliver service to every nook and cranny of every building in its service areas.