As mentioned in a previous note, the Trace Center has joined the Infrared Data Association in order to have some representation of the disability community within the association. Two weeks ago, I attended the Infrared Data Association meeting in Chicago. Although I was unable to stay for the full session, I was able to participate in the consumer products or long-range infrared portion of the meeting. At that meeting, I presented work we are doing on the infrared protocol for connecting alternate interface devices to standard products via infrared. I presented the general case or need for such a protocol, followed by the work we have done so far in developing such a protocol. (A copy of our current protocol can be found on our web site, at http://www.trace.wisc.edu/world, in the Standards section.)
The presentation was well received by the group. From the discussions, it was clear that our impression that the protocol we were proposing was something that ride on top of IrDA rather than interfering with IrDA (requiring a change in IrDA) was correct. The IrDA protocol essentially can be thought of as an RS232 protocol. That is, just as RS232 specifies the cable, the pin-out, and the voltage levels for connecting serial ports, the IrDA specifies the frequencies, packet structure, and handshaking necessary to connect two devices via infrared. Once the serial cable (for RS232) or infrared link (for IrDA) is established, you then use other protocols for transferring data across the link.
In our case, we don't have a need to change the basic IrDA linking protocols. Rather, we are defining the protocol for controlling devices remotely over the IrDA link once it's established. In fact, the protocol we're talking about wouldn't necessarily have to travel over an infrared link; the protocol could be used to control devices over a radio link or a wire. Our focus is on the infrared link, however, since it's low cost and hands-free hook-up make it the most likely interface to exist or be incorporated into devices on a widespread basis.
It is also interesting to note that if an IrDA device is in the middle of communication and it detects any other infrared source, it shuts down for a half second, and then starts looking to see if the way is clear. If it is, it will pick up where it dropped off. Thus, it is possible for someone to use a standard infrared TV controller while an IrDA session is going on. While the person is using the infrared remote control, the IrDA device will simply back off when it is stepped, and reduce its transmissions after the person stops using the remote control. However, if the person used infrared headphones, which transmit infrared continuously, the IrDA device would back off and never come back on line. Every half second it would stick its head up, see that there was still infrared in the environment, and go back to sleep for another half second.
The IrDA 1.1 also has a limited range. As a result, it does not really meet the needs of individuals wanting to control appliances from across the room. For this reason, a second protocol, being called the Long Range or Consumer Products protocol, is being explored. It is not clear exactly what form this other protocol would take. Some forms suggested that it might operate in a burst mode, like current remote controls. It would work in the same room as IrDA 1.1 devices, as long as it was only used sporadically. When used, it would cause the IrDA 1.1 device to back off and wait for the burst mode device to stop transmitting.
One problem discussed at the meeting was the fact that some remote interface applications (such as people with joysticks trying to control a device) would require greater bandwidth than might be possible using some early IrDA protocol proposals. This is especially true trying to use data rates in the 115 kilobaud range. However, they are now talking about data rates up into the 4 megabit range over a fairly reasonable distance. This would allow for a sufficient data rate to handle such interfaces. It should be noted, however, that a joystick would need to have continuous data flowing, and thus the use of joysticks in a room would completely preclude the use of any IrDA 1.1 devices which could pick up the light from the remote devices. These devices would therefore not be compatible with IrDA 1.1 as it is currently implemented, unless some interspersing strategy were developed.
I hope that this update isn't too disjointed to comprehend. There is so much to think about, and in fact so much involved, that it's hard to know how much explanation to put in as I go along here. Instead, let me invite you to read through the update twice, and if there are sections that are still unclear after the second read-through, drop a question back on the listserv, and I will try to provide a more detailed discussion of those parts.
I will also blame part of this on the fact that it's the wee hours of the night - actually, it just turned morning, so I'm going to bed. I'll send this off later today, and await your comments and questions.
Gregg