CDMA Technology
Members Sign-In
Testing Time

By Fred Dawson

Defining, let alone measuring, all the technical parameters essential to success in CDMA wireless operations is an ongoing challenge that’s about to get tougher.

There’s never time to stand still in the communications industry. Just as suppliers are delivering compact, powerful devices that resolve most of the requirements for testing today’s IS-95 systems, the Telecommunications Industry Association’s wideband spread spectrum subcommittee (TR- 45.5) is about to approve a set of specifications for IS-95B that will require new test parameters focused on data service performance. Moreover, operations engineers and vendors alike say that because CDMA is still relatively new as a PCS interface, management should expect new contingencies to emerge that will require new measuring techniques and procedures even before they adapt to IS-95B or, coming later, a converged third generation CDMA standard.

"Where TDMA and GSM have had years of field experience contributing to refinements in system testing, CDMA is still maturing," says Charles Neal, a project engineer with Expert Wireless Solutions, Inc., a Fort Lee, N.J.-based supplier of design and operations support services for mobile operators in all categories. "We have the equipment we need to test the parameters we now know need testing with CDMA, but there’s a good chance we’ll find new parameters as we gain more experience."

Routine
At this stage, CDMA is a bit more hands-on with respect to ongoing performance testing requirements compared to other technologies, often requiring more time to investigate the exact causes behind a particular problem, Neal says. "Commitment to testing varies among operators, but, no matter how centralized and automated you become, there’s no substitute for taking measurements from the field on a routine basis."

"Testing is a continuous process," agrees Alex Gogic, technical director for AirTouch Communications, Inc., the West Coast cellular carrier that is partnered with Bell Atlantic and U S West in PrimeCo. "Measuring signal to interference ratio at all base stations has proved to be a major, important task, but we’ve found vendors have responded with equipment that makes this much easier to do than it once was."

Identifying parameters for testing remains a work in progress at Wavetek Wandel Golterman, which intends to bring out a line of gear for testing soft handoffs this summer. "There are a lot of things that aren’t certain yet, although we’re comfortable with our ability to test for most of the key parameters at this point," says Allen Shuff, vice president of marketing at Wavetek.
"It’s reasonable to expect there will be some ‘gotcha’s’ we haven’t accounted for in our measurement parameters so far," agrees Neal Buren, a product manager at Hewlett-Packard Co. "Performance that looks good now may have some weak spots that don’t show up until there’s more pressure on system capacity from increased [market] penetration."

For example, Buren notes, current placement of microcells to accommodate growth might prove to be less than ideal once network usage intensifies. "You should be able to figure out how to solve the problem using existing tools, but there may be new things to look for that could make the job easier," he says.

Seasonal
Another unknown is the full impact of seasonal changes. "No one has yet quantified the effects of foliage changes in various localities through all four seasons," Buren notes.

Further complicating matters, "no single company has the full range of equipment you need," he says. "There are a lot of different parameters being measured by different types of equipment and more being added all the time because of feature enhancements — all generating data that has to be compiled into useful information."

Earlier this year HP took another step toward simplifying performance measurement procedures with the introduction of an integrated drive test system combining receiver-based RF coverage measurements and phone-based service quality analysis into two software-compatible data streams that can be manipulated many ways in diagnosing problems. Such information-gathering tools combined with back-office data-processing systems can generate highly specific information as to the location of the trouble spot as well as the nature of the problem, allowing field workers to quickly tackle problems as they find them.

Receiver drive test gear or "pilot scanners", first introduced by HP over a year ago, have spread rapidly into the field, with products now offered by several vendors. Previously, drive tests relied on phone-based measurements, which made it impossible to know whether the mobile was reading the random test code known as PN (pseudo noise) from the proximate base station or the time-shifted PN of a neighboring base station.

With scanners that operate off the local geo-positioning system (GPS), which is the keeper of absolute time for base station offsets, technicians on drive tests now know instantly which base station their mobile is communicating with. They can also get a sense of the "ping pong" effect that occurs at the fringes of cell coverage areas when mobiles hop back and forth among neighboring cells, eating up more system capacity than they should and occasionally creating problems with the connection.

Scanners

Germantown, Md.-based Dynamic Telecommunications, Inc. wants to go one step better on HP and other vendors offering GPS-connected scanners by bringing product to market that accomplishes the same thing without having to make radio contact with the GPS, says DTI CEO Paul Kline. The company, which is preparing to introduce a CDMA pilot scanner "the size of a CD player" by the fourth quarter, hopes to bring a version to market later that will use chips fast enough to keep track of all 512 base station offsets at once as the scanner moves through coverage. This will allow the identification of offset sources without connecting to the absolute time set by the GPS.

"You lose contact with the GPS a lot in city environments," Kline said. "We’re working on a patent for our idea."

As thoroughly covered as the raw data is at this point, the information may still not be sufficient to tell the operator what he needs to know about performance vis a vis his competition, which leaves a lot of guess work to the process of deciding how much to spend on improving performance. The need for reliable comparative information is so great that Chicago-based Safco Technologies, Inc. has created a new business within its TEC Cellular unit to offer comparative performance information to PCS and cellular operators of all stripes.

Safco, which offers engineering and back-office processing support as well as hardware solutions for wireless testing, has come up with a means of supplying electronically generated mean opinion scores (EMOS) for mobile call performance across all air interfaces in a market, says Martin Singer, Safco CEO. "We’ve completed an extensive survey of CDMA cellular and PCS and analog cellular systems in the Chicago area and intend to proceed with more surveys at a rate of two cities a month on a global basis," Singer says.

In the Chicago survey, the test not only picked up very significant differences in voice quality between analog and CDMA systems; it also picked up slight differences between the cellular and PCS CDMA systems. A difference between the average frame erasure rate of the RF carriers over links between the network and the mobiles in the two digital systems probably accounted for the spread in EMOS results, Singer says.

Such fine points can make a difference in sales performance but may never be picked up in routine testing, Singer notes. "We want to be the Nielsen (U.S. TV audience measuring company) of wireless systems," he says. "This will be a completely untainted, objective warehouse of data that people can access to get information that otherwise wouldn’t be available."

Looming

As these developments suggest, current testing systems have gone a long way toward meeting current needs. But looming as large unknowns are the test parameters associated with new iterations of CDMA, including the forthcoming IS-95B, which was to be reported out of the TR 45.5 sub-committee by sometime in July. Amid widespread press attention to third generation CDMA, still at least a couple of years from commercial deployment, IS-95B has emerged as a largely software-based means of adding data delivery capability at rates of up to 96 kilobits per second to existing infrastructures that should be commercially available by early next year.

"There will be some data-specific parameters people will want to measure that I’d imagine will require new measuring devices," says AirTouch’s Gogic, who serves as a consultant to the CDMA Development Group. "There will be a need to test for effective data throughput, and the high data rate has an impact on the RF interface and on power control that will have to be measured."

Moreover, Gogic adds, there will be smaller step sizes in power control adjustments with IS-95B than is the case with IS-95. With up to eight 1.25 MHz CDMA channels in use on a data connection, power has to be viewed across multiple channels, which tends to lower the marginal adjustment requirements, and, in the case of data, users are likely to be in a low-mobility mode where power sensitivity is not as high as it is under full mobile conditions. Rather than requiring adjustments of 1 dB per increment of power change, the new standard will call for 1/2 dB as the default step adjustment margin, with 1/4 dB as an option when data is being transmitted from a base station sector, Gogic says.

IS-95B will also change the threshold parameters that are used in determining whether a mobile should go into soft handoff, he adds. Under IS-95B, the soft handoff threshold, the noise level that forces the mobile to find another base station connection within the same frequency channel, will be assigned dynamically depending on the strength of the mobile’s immediate connection.

The procedures for testing under IS-95B, contained within the standard initiative known as IS-98, won’t be established until several months after the release of the initial specs, notes David Whipple, a member of the TR 45.5 committee who is also CDMA R&D project manager for HP. "It will be awhile before we know what parameters to build (test equipment) to," he adds.

Merit

Whipple is not sure whether there will be enough demand to merit investment of HP resources in a new generation of test equipment for IS-95B. "I’m not sure people have figured out what the business case is for data over IS-95B," he says.

While the CDG likes to talk about the 96 kbps potential of IS-95B, that number assumes the operator has installed gear supporting 13 kbps vocoders, which translates to a data rate of about 12 kbps per 1.25 MHz channel. "Virtually all U.S. operators are using 13 kbps vocoders, which means that you get at least 96 kbps of data throughput using all eight channels," Gogic says.

But, as Whipple notes, operators will want to reserve some of those channels for voice calls, which will diminish the data rate, as will contention among multiple data users within the same sector. Moreover, he adds, there is good reason to believe operators won’t be able to drive data at the 13 kbps level, because, so far, only the 8 kbps vocoder level, which is available on all systems equipped for 13 kbps, has been shown to work well in delivering data.

Operators are divided over whether to pursue IS-95B or to wait for the third generation. While AirTouch, for example, has been a major driver behind IS-95B and is planning to deploy it as quickly as possible, Sprint PCS is focusing on getting an early version of the third generation standard up and running, says spokesman Tom Murphy.

But, whichever way they go, CDMA operators will confront significant shifts in measurement parameters in the years ahead if they want to take advantage of the technology’s power to deliver data. With third generation moves to a five MHz channelization scheme, the changes will be more acute as time goes by.