PANs give new dimension to wireless connectivity

By the time you read this, the much-discussed concept of the personal area network (PAN) should exist in fact. A wireless personal network is a cluster of devices that are close together physically-often literally at arm's length

May 1st, 2000

Innovative wireless firms bring the latest equipment literally within arm's reach.

Dan Sweeney

By the time you read this, the much-discussed concept of the personal area network (PAN) should exist in fact. A wireless personal network is a cluster of devices that are close together physically-often literally at arm's length from one another-and connected, often on an ad hoc basis, to serve the purposes of an individual user. Normally, the devices in a personal network would access one particular node in a larger network, if indeed the hub of the personal network is a part of anything larger. In other words, what represents an access point for a personal network would be an end point in the architecture of the local area network (LAN).

More than cord replacement

Initially, the personal network was envisioned by wireless data hardware manufacturers simply as a means of connecting peripherals to a workstation or portable computer, or, alternately, to dock a computing device with a mobile phone. The chief function of the wireless connections, as originally conceived, was simple cord replacement. That's still a big part of what personal networking is about, but manufacturers now realize that eliminating cable connections within a personal space-which may extend for several yards around a central point-allows for almost infinitely flexible ad hoc connections in a group of devices. Connections can be set up and torn down on the fly using a plug-and-play strategy, and individual devices can communicate on a peer-to-peer basis, with one device serving as a temporary hub for the purposes of network administration.

In such a network architecture, databases on personal devices may be instantly and automatically synchronized with central databases, a video display or audio system might be utilized for an impromptu presentation, or a visitor could use a short-range personal connection to access a wide area network. Sensors or monitoring devices might also be temporarily deployed over the network.

Such ubiquitous, even casual, deployment of wireless data connections would be based upon single-chip integral radio modems, which could be included in all computing devices for a slight price premium. Modem drivers would routinely be included in common software platforms, and network software would include distributed intelligence and advanced self-organizing capabilities.

Current and future standards

How far away are we from such widespread wireless connections? In fact, a wireless personal-networking standard has existed for some time, namely the IRData (infrared data) protocol embraced by all major computer manufacturers and embodied in a good percentage of PCs and peripherals made today. Millions of computing devices with IRData ports have been sold to date, and the standard is well- established on that basis, except, according to user surveys, neither consumers nor businesses generally employ the IRData connections. The standard has won general industry support and has been implemented on a wide scale, but it has had little impact on the end user.

The wireless PAN standards of the future will not specify infrared frequencies as a medium of transmission but will be based upon unlicensed spectrum just as are in-building wireless LANs today. A move toward RF will eliminate the distance and line-of-sight limitations of the older optical networks and make the personal network more robust. In addition, more advanced media access and transport protocols will be employed.

The almost certain industry standard for the short term is Bluetooth, developed in large part by Ericsson and Nokia and endorsed by more than 1,000 manufacturers. The standard, which specifies advanced network control features and goes well beyond the simple cord-replacement model, has been submitted to the Institute of Electrical and Electronics Engineers (IEEE-New York City) to be subsumed under that body's 802.15 PAN interoperability specification. Because it is the only submission so far, it is likely to win unqualified endorsement.

Bluetooth occupies the same 2.4-GHz spectrum as do wireless LANs and thus has the potential for interfering with them. The issue remains unresolved at this writing. Manufacturers appear to be assuming that Bluetooth will find a much larger market than have wireless LANs and that actual problems in the field will be negligible due to the relative scarcity of wireless LANs.

Philips and other companies have been working on an alternative wireless PAN standard based upon the high-speed, Asynchonrous Transfer Mode-based multimedia FireWire (IEEE 1394), which currently uses copper wire and optical fiber for the physical layer. Wireless FireWire presupposes throughput speeds in the tens of megabits per second, and while it has been demonstrated in the laboratory, it is not at the point where low-cost implementations are possible.

Also in the future lies a higher-speed (approximately 10-Mbit/sec) and longer-range version of Bluetooth. Standards work is already underway, but the revision does not enjoy the degree of industry support of the original-at least not yet. Such a standard could conceivably compete with the existing IEEE 802.11 specification for wireless LANs and might have difficulty gaining manufacturer acceptance on that account.

For the time being, PANs remain ill-defined and subject to competing visions on the part of their supporters. Bluetooth equipment will certainly become available in the near future. The applications that will grow up around it remain conjectural at this point.

Dan Sweeney is technical columnist for Wireless Integration, another PennWell publication.

More in Home