Global Sources
EE Times-Asia
Stay in touch with EE Times Asia
EE Times-Asia > Amplifiers/Converters

Wireless future: At what cost?

Posted: 01 Jul 2000 ?? ?Print Version ?Bookmark and Share

Keywords:bluetooth? ericsson? mixed signal? homerf? rf?

Stephan Ohr considers the future of Bluetooth in light of competing wireless standards.

I recently attended "Bluetooth Geneva," a conference sponsored by UK-based IIR Limited. Bluetooth, a proposal by cellphone makers Nokia and Ericsson along with computer makers IBM, Intel and Toshiba, is now endorsed by over 1,600 manufacturers. That this technology is hot is evidenced by the questions from the Taiwanese engineers at the executive panel discussion, "Interfacing for the Millennium," sponsored by Global Sources during the IIC-Taipei conference in May. Bluetooth was intended to link portable PCs and cellphones with a short distance wireless link. In principle, it will enable you to download an address book from your PC to your cellphone. Though it frequency hops at 2.45GHz, the Bluetooth transceiver was conceived as dumb, low power (10mW), and cheap.

But electronics manufacturers began to speculate on what they could do with a really cheap wireless transceiver, and turned Bluetooth into something that belongs in a sci-fi film. Technology fantasists have speculated that you can use Bluetooth to automatically create a wireless connection to the Internet for your laptop in convention centers, hotel lobbies, and airport gate areas.

This is exactly what IEEE802.11 wireless LAN technology proposes to do. It allows PCs to connect with each other (or with hubs and Internet gateways) on a peer-to-peer basis. Like Bluetooth, it frequency hops at 2.45GHz?although with different hop rates and packet sizes. Unlike Bluetooth, it offers a wider range (to 100m) and a faster data rate (up to 11Mbps).

Thus, there will be competition between Bluetooth and 802.11 technologies for developer and user loyalties. The HomeRF specification, a proposed standard for consumer-based wireless networks, should be mentioned here, because its success will also depend on accessibility to consumers?particularly ease-of-use and low cost.

If Bluetooth is to become the standard, it must be scaled upward from the point-to-point "cable replacement" suggested in version 1.0 of the specification. The Bluetooth Geneva conference was helpful in reminding potential implementers that point-to-multipoint stack negotiation was already built into the v1.0 spec. This means that a Bluetooth node can distinguish between eight users, or even more when nodes are cascaded (a topology called "scatternets"). Similarly, increasing the transmitter power to 100mW could increase the range of Bluetooth transmissions from 10m to 100m.

But all at what true cost?

The costs of doing this will be considerably more than the US$5 intended by the 1,600-member Bluetooth special interest group (SIG). Ericsson's Bluetooth modules are currently multichip devices on low-temperature co-fired ceramic (LTCC) that currently cost about US$28 (and remember, Ericsson is one of the original developers of the spec). "No matter how well you integrate the antenna drivers, RF transceiver, link-level interface and microcontroller ICs, a major contributor will be assembly and packaging costs," says Stefan L?f, Ericsson's senior product manager. RF transceiver architectures currently depend on dozens of "tank circuits". They are resonant elements made up of inductors and capacitors (20 to 30 of them) and these will contribute to packaging and assembly costs. Thus, Bluetooth radio costs are likely to average US$27.2 in 2000, and US$10 in 2003, forecasts L?f.

Semiconductor integrators like Silicon Systems Limited (SSL) and Cambridge Silicon Radio were promoting single-chip radios that they could not yet demonstrate at a conference. National Semiconductor and Samsung Electronics are now demonstrating or discussing multichip sets with separate transceivers and baseband processors.

K.W. Lee, vice president and general manager of corporate research at Samsung, suggested such a chipset would be US$15 in 2000, with a total implementation cost of US$26. Lee cited data provided by TDK, stating "the chipset will drop to US$5 by 2003, and cost several dollars to solder down. So the actual total cost of a Bluetooth implementation won't drop below US$5 until 2005."

With Microsoft, Lucent Technologies, Motorola and 3Com now part of the Bluetooth promoters group, the technology will get even more attention. But the truth is, 802.11 systems have an advantage in terms of development and manufacturing, as well as the benefit of higher data rates. The current Bluetooth spec uses an encoded 1MHz signal, in a data rate of 721Kbps. The 802.11b spec goes to 11Mbps, with modem-like fallback rates. There are even proposals for extending this up 54Mbps.

A panel discussion I attended at ISSCC in San Francisco in February on competing wireless technologies seemed to feel that costs of 802.11 were coming down?and would be soon in range of the Bluetooth target. The panel?which included Brent Myers of Intersil, Ran-Hong Yan of Lucent Technologies Bell Labs, and Modest Oprysko of IBM?seemed to feel semiconductor volumes would soon bring the cost of an 11Mbps 802.11 transceiver into the US$100 range. Lucent's Yan was the most aggressive, suggesting that cost per node of a wireless LAN will drop below US$50?and as low as US$5?within the next few years. At the panel discussion during IIC-Taipei, Carl Andren of Intersil commented on the future of the standard, telling the Taiwan audience that IEEE 802.15 is looking into 20Mbps enhancements for personal area networks (PANs).

Certainly, there is a lot of overlap between these technologies in what they propose to do. In fact, a Lucent executive confided that his company was working on a compatibility scheme for Bluetooth and 802.11 systems. Samsung's Lee saw great potential for Bluetooth devices in home appliances and consumer electronics. Their overlap with 802.11 and HomeRF solutions would depend on bandwidth and distance considerations. HomeRF and 802.11 transceivers would dominate, for example, where distances between connected devices exceeded 50m, he speculated. Bluetooth would dominate for shorter distances.

Of the competing wireless options, HomeRF seems the most risky. Siemens, which has almost single-handedly promoted the use of DECT cordless phones in the USA, recently added their name to the supporters of HomeRF, hoping perhaps that this would expand the market for DECT products. National Semiconductor, which has been manufacturing DECT chips for many years, similarly is interested in the cordless phone part in the HomeRF, though they have not (to my knowledge) formally committed to supporting the SWAP specification.

Besides being called a "technology searching for a market," the HomeRF SWAP specification tries to do too many things. It tries to integrate an IEEE802.11 wireless networking technology for data, with a DECT (which now stands for Digital Enhanced Cordless Telephone; no longer "European") transceiver and software for voice, and a handshake that will automatically recognize either one.

I cannot imagine how you can get the above solution down to the US$200 per node originally projected for HomeRF at its inception in 1998. But it seems like a trap, since the complete system would have to have a consumer end price of something like US$160 (or maybe down to US$69 a node). And finally, how this then competes with Bluetooth, which is claiming US$5 per node, is beyond me.

Americans frequently talk about "boat anchors." These may be seemingly bright technologies one year (like steel-cased PCs), which may become useful for little more than paper weights one or two generations later. Wireless is not immune to this potential. Look what happened to the Iridium project. This was the network of low-flying satellites that would allow you to make cellphone calls from anywhere on earth (i.e., the North Pole or the Amazon rainforest). But at US$675 for each brick-like phone, there were few takers. After wasting US$10 billion in development costs, the Iridium organization went bankrupt, and announced they would start bringing down the satellites, letting them burn up in the earth's atmosphere. Some wealthy friends jested they could use the descending satellites as ultra-dramatic fireworks displays for their birthday parties. The lesson is to be wary of too much wireless connectivity propaganda.

Article Comments - Wireless future: At what cost?
*? You can enter [0] more charecters.
*Verify code:


Visit Asia Webinars to learn about the latest in technology and get practical design tips.

Back to Top