Problem/Situation: Fast accurate data transmission is needed for mobile government workers.
Solution: Wireless data communication methods are evolving, but must be evaluated carefully.
Vendors: Rysavy & Associates, ARDIS, RAM Mobile Data, AT&T;/McCaw, Ameritech, Bell Atlantic, Metricom Inc., Datacomm Research Co., LCC Inc.
Contacts: Ira Brodsky, Datacomm Research Co., 708/256-1763; James Kobielus, LCC Inc., 703/351-6666; Peter Rysavy, Rysavy & Associates, 206/517-5654.
By Tod Newcombe
Untethered computing has always appealed to the road warriors of today's Information Age. Already enamored by the cellular phone, mobile workers - and those who speak for them - have touted the potential benefits of instant data communications from a car, park bench or airport lounge. Several companies have poured millions of dollars into building wireless networks, hoping users would pull the plugs on their laptop modems and start transmitting data over the airwaves instead.
So far, it hasn't happened. The performance of wireless communications is achingly slow when compared to data communications via wired modems and land lines. Because of proprietary protocols, developing applications can be complex and costly. As a result, the total number of wireless users remains small, and few major organizations have adopted the technology.
But just as multimedia seemed to go nowhere until CD-ROM became inexpensive, wireless data communications appears poised to move forward, thanks, in part, to some faster, possibly less expensive technologies. "It's getting close to becoming an irresistible package, like the cellular phone," said Peter Rysavy, a consultant with Rysavy & Associates. "People who work in this industry can see it all coming together and at some point it's going to reach critical mass where the whole thing just explodes."
For state and local government, the changes could open the door to some innovative applications. Conventional thinking about wireless data communications in the public sector has usually restricted its use to law enforcement. Police, it was thought, would benefit the most from instant access to remote databases they could search to identify suspects and vehicles in the field.
But a host of government agencies, which also put workers in the field, could benefit from wireless communications as well. Still other agencies might want to rethink how they use their office workers and use some of them more productively in the field, communicating with the agency via laptops and wireless networks.
However, before agencies send out swarms of roaming workers equipped with personal digital assistants (PDA) and wireless modems, they must address a long list of issues. These range from coverage and costs to data rate performance and reengineering applications.
THE CURRENT SCENE
To transmit data over a wireless network, organizations currently have a couple of choices. One is to use the commercial radio data services provided by ARDIS and RAM Mobile Data. Both firms are fully deployed and provide their subscribers with the ability to transmit data while roaming throughout most of the United States.
ARDIS and RAM use a technology called packet radio, which broadcasts data in bursts or packets at transmission rates that range from 4.8K to 8K per second. (Higher transmission rates are possible in certain locations.) Both companies use a network of base stations to route messages along the shortest path and have proprietary protocols.
Another, more popular approach, according to Rysavy, is to purchase a cellular modem, plug it into a laptop, and send and receive data over the existing cellular phone network. Called circuit-switched cellular, the technology is relatively simple to use, although it requires a connection between the laptop, modem and cellular phone. Also, cellular subscribers are restricted in the ability to roam. Typically, they need additional accounts and phone numbers if they move between two or more cities.
"Technology has improved dramatically in recent years for data communications over the cellular network," said Rysavy. "In fact," he added, "the number of people sending and receiving data over the existing cellular network is greater than all of the other wireless networks combined. That makes it the leading wireless data solution out there today."
The newest kid on the block is cellular digital packet data (CDPD), a digital communications service that operates at rates of 19.2K per second. CDPD sends data packets over the existing cellular networks, taking advantage of idle time between voice calls to transmit data.
While it has barely arrived, CDPD is causing quite a stir for several reasons. First, it's twice as fast as packet radio and faster than switched-circuit cellular. It is also built on open standards, including the TCP/IP protocol. "That makes CDPD marketable for those folks who want to get onto the Internet from the field," said James Kobielus, senior telecommunications analyst at LCC Inc., a wireless network design and engineering firm. Kobielus also pointed out that CDPD automatically encrypts its data transmissions, ensuring the confidentiality of wireless data communications.
Because some of the leading communications firms, such as AT&T;/McCaw, Ameritech and Bell Atlantic, are implementing and marketing CDPD, the wireless service is expected to grow rapidly in the next few years.
"Based on marketing, availability, standards and data rates, CDPD stands a good chance of becoming the MS-DOS of wireless data communications," said Kobielus. "It may not be the fanciest or the most sophisticated wireless data protocol in the long run, but it will be everywhere, available to everybody and fairly cheap."
Two other kinds of wireless services could be of special interest to state and local governments for different reasons. Specialized mobile radio (SMR) is a highly secure transmission technology that's often used as a regional service. For those reasons, it has become popular with police departments.
Another emerging service is called spread spectrum, which is described as highly secure and is considered fast for wireless applications. Right now, only one company, Metricom Inc., offers spread spectrum service, and it's available in just a few major metropolitan areas.
What makes spread spectrum of interest to governments is its architecture, which relies on a mesh network of numerous, low-powered, micro-radios located on street lights, utility poles and buildings. Ira Brodsky, president of Datacomm Research Co., a consulting firm for wireless technologies, believes the technology could become a source of revenue as well as serve as a practical, high-speed wireless data communications system. "It would be easy for local governments to set up their own spread spectrum network on top of street lights, and then they could lease it, or they could franchise the poles' rights of way to a provider and use some of the capacity for their own use."
Whichever service a government chooses, they have to be prepared for working within an entirely different environment. "The radio environment is a hostile one," said Rysavy. "The kind of data rates it offers, the problems with drop-outs, interference and fading result in much less predictable data transmission than you would get over a wired connection." He added that wireless service providers compensate for these problems using innovative ways of protecting the data, but the end result is unpredictable delays in data transmission.
Such unpredictability presents special challenges for wireless customers. For example, existing software applications, such as electronic mail or database access cannot be ported directly to a wireless network. They will need to be configured to compensate for the changes in performance and higher costs.
"What you need to do is step back and ask yourself: What do we really need to be doing here?" said Rysavy. "What's the minimum amount of information we need to send to keep costs down? What are all the things that can happen, such as running out of range?"
The performance issue regarding wireless is expected to remain for some time, according to the experts. Even spread spectrum service, which can transmit data at rates up to 77K per second, is far slower than today's landlines and local area network speeds. In fact, the hostile environment Rysavy mentioned can reduce the listed performance of a wireless service significantly.
"The main obstacle today is the performance of the wireless networks," said Brodsky. He compared today's wireless performance to that of a telex service. "People who have powerful notebook computers are not interested in sending and receiving data at 300 bits per second."
Another issue concerns the lack of standards in the wireless world. "Right now, there are too many of them," said Kobielus. Their impact has been felt on the equipment makers, who have to come up with modems and radio transceivers that can handle the proprietary protocols and interfaces. The result is few choices for the less popular wireless services and higher customer prices for hardware and software.
But Brodsky sees the standards problem as just the natural evolution of technology. "There's a major fallacy in pursuing standards early on," he said. "The first thing you want to do with this technology is not set a standard and lower the price, but to find out what people want. The best way to find that out is through real time experimentation in the marketplace."
WHERE TO BEGIN
With experimentation still happening and standards fluctuating, should state and local governments tap into the wireless world, and if so, how should they do it? Brodsky, for one, believes that wireless is the kind of technology that today's state and local governments need to incorporate into their operations. "If government is going to operate more like a business, then it needs to focus on opportunities," he said.
If they pass up wireless, Brodsky feels that state and local governments are ignoring what will become the next generation technology for service delivery. "Government should no longer take a wait-and-see approach to the technology. There are more options than ever."
Brodsky recommends that governments look for solutions, not the best technology. They should also put pressure on vendors to meet their expectations, rather than lower their own.
For specific advice, Rysavy recommends that agencies start by making sure they have wireless coverage where they want it. They need to look at the amount of data that would be transmitted over the wireless network and make sure they can justify it based on cost.
He cautions that an existing application will have to be reengineered and restructured so that it uses the available bandwidth efficiently. "The cost and throughput of wireless networks is not as good or as cheap as a wired connection," he said.
Going one step further, Rysavy said that using wireless networks will force government agencies to rethink how they work. "The application's mobility itself will allow people to work in fundamentally new ways. When you design a wireless application, you end up redefining how people do their jobs," he pointed out. As a result, work redesign must be taken into consideration during the planning stage.
To compensate for these and other changes, Rysavy recommends that agencies deploy a wireless application first in a pilot test and evaluate its effectiveness using a small number of users. "Allow your agency time to go up the learning curve before instituting wireless with a large number of users," he said.
For those who still doubt that wireless will prevail, they need to listen to Kobielus, who has worked in the wireless business for a number of years. "The demand for wireless is strong and getting stronger. CDPD will be available in 90 percent of the cellular market by 1996. As a result, supply will create its own demand as more and more services become available. When that happens it will be cheaper - for the equipment, service and customer."
PCS: THE WIRELESS GORILLA?
One wireless service sure to play a major role in mobile data communication is personal communications services (PCS). The big question is when.
Right now, PCS is not a service but a frequency allocation recently auctioned off by the Federal Communications Commission in a major reallocation of spectrum for the emerging personal wireless voice and data communications market.
PCS has two major categories: narrowband, which covers two-way paging, and broadband, which covers a much wider band of spectrum and will be used for a variety of wireless voice and data services.
PCS will bring lots of new capacity to today's cellular market. It could deliver as much as 10Mbits per second, making it nearly compatible for handling LAN traffic. PCS also will bring a half-dozen new competitors to the cellular market. They are expected to offer digital versus today's analog cellular service. Digital systems are considered more cost-effective than analog and offer much better integration of voice and data.
As a result of the auctions, PCS has taken its first step toward reality. "The thousand-pound gorilla in the wireless market is going to be PCS," said James Kobielus. "The new players will probably be offering voice and data services somewhere in the 1996-1997 timeframe."
Peter Rysavy and others think it will be longer before PCS is deployed. "Right now, people have only the licenses. They can't use the spectrum now, because there are people already using it." The PCS spectrum is currently used by a number of utilities and local governments, who must be compensated by the new license owners to relocate. "It's a complex situation," said Rysavy, "and it's going to take time to happen."
It's also going to cost the new players a lot of money. They have to pay for the licenses, the relocation of current spectrum users, and the infrastructure for the new services. "It's going to cost tens of billions of dollars to deploy," remarked Rysavy.
For those reasons, Ira Brodsky believes that the new PCS service providers will focus on providing voice, not data, services first. "They are going to be under tremendous pressure to make money quickly," he said. "The way to do that is to compete with the cellular voice services." Brodsky believes the wireless industry is at least five years away from providing wireless data applications via PCS.