Slashdot Mirror
Wireless at Firewire Speeds?
MeCoward writes "EETimes reporting on working group that hopes to leapfrog 802.11 to create wireless 1394 links.
Initially 100mbps but aiming for 400mbps." I don't expect to see this anytime soon, but it certainly makes things like wireless HDTV feasible. Sure would be cool. Of course Bluetooth is only now just catching on, so imagine how long it'll be before this becomes practical.
I don't expect to see this anytime soon, but it certainly makes things like wireless HDTV feasible.
Uh... maybe I'm just a dumbass or something, but wireless HDTV is already feasible. I watch it every day. It's called 8VSB.
However you encode it, broadcast HDTV is only 19.3 Mbps. It's feasible over dual-like 802.11a, or 802.11g.
Firewireless has been around a while. It even has DRM.
I don't expect to see this anytime soon...
Why would you? We've only been waiting several years already.
So much for being an 'early adopter'.
FireWire as an electrical interconnect is good. FireWire as a protocol sucks.
Down at the bottom, FireWire is a LAN. You send packets with a source address and a destination address. It's a TDMA LAN, more like token ring than Ethernet, with assigned time slots.
Video is sent as broadcast packets, on a rigid schedule, with no ACKs. That's quite straightforward.
The ugly part is the layer which implements load/store emulation for 32-bit data items in a 64-bit address space. This was designed by people who think in terms of "device registers". Control functions are exercised by stores and loads from "device registers". Typically, these "registers" have no physical existence at either end; one end has a CPU issuing commands and the other end receives commands and executes switch statements. Register definitions are supposed to be standardized; in practice, the standards are more ambiguous than they should be. This results in FireWire devices coming with unnecessary "drivers". A command/response protocol like SCSI would have been far better. With the current system, generic drivers are hard.
There's already Ethernet on top of FireWire, SCSI on top of FireWire, and raw IP on top of FireWire. This is too much layering of pure packet protocols.
MIDI + sampling over firewire has been around since 97, it's called mLAN and it was introduced by Yamaha. This would just be a phycial transport change from normal firewire cabling to UWB. For more info on mLAN see the mLAN alliance website Here
There are 4 boxes to use in the defense of liberty: soap, ballot, jury, ammo. Use in that order. Starting now.
If you were to look at all the RF sources going through the air at any one time, including radio/tv station towers and all of the wide-spectrum junk from that massive nuclear explosion that keeps us warm 93 million miles away, then you should already be paranoid.
Unlicensed transmission devices are already limited to 100mW ERP transmit power. Most modern cell phones are under 600mW maximum IIRC. We probably would have seen much worse already had this been a major problem. What about cordless phones? What about the CRTs, even the low-radiation kind? Those make me more nervous than a simple radio device because we are more frequently and directly exposed to their radiation than a transmitter on a device connected to electronic equipment.
A few years ago, the IEEE Spectrum had an article that addressed the problems of RF from sources like power lines. One of the most interesting conclusions: the radiation along the center axis through an earphone was actually a significant source of radiation to the brain. Does that mean we ban earphones?
Sure, we need to do studies, but I'm suspecting that we won't have to wear tin foil on our heads any time soon, if for no other reason than that we should've already been wearing them a long time ago.
Ever heard of 802.16? Seriously, the microwave folks have been doing point to point wireless to project mad bandwidth across serious distances for a LONG time.
http://tinyurl.com/4ny52
I am a member of the IEEE Standards Association, and I've spent the last month writing a paper on WLANs.
1)802.15.3 IS Ultra WideBand.
2)The FCC has basically crippled the original version of this tech.
3)Cellular providers & GPS want their freqs eliminated from this (UWB goes from 3-10 GHz)
4)The original spec only went to 100 Mbps, and there is no official working group trying to expand this.
5)The outermost range is 10 meters, while 802.11 can max out at 100 meters. Great leapfrog action!
6)Only 4 companies can currently produce UWB devices- 3 for imaging systems and 1 for some kind of "toilet device". (seriously! but I couldn't find any more enough about this toilet thing)
7)Thomson's 802.11a & HiperLan product has nothing to do with UWB, yet they quote 802.15.3 (see #1)
8)TOTAL HORSESHIT STORY
Happy day!
Vote monkeys into Congress. They are cheaper and more trustworthy.
You may have heard of this nifty new technology called "compression" that allows you to get more effective data transfer out of a link. 10:1 compression on video is pretty trivial these days, which means you'd only need 175Mbps for that resolution (although you'd probably drop it down to 24 bit color).
I read the internet for the articles.
DivX is *VERY* high compression.
Think for a sec.
Let's use 640x480 as a sample res, 16 bit color, 30 fps.
640*480*16*30 = 147,456,000. 147Mbit/sec. Without audio. Most DivX files are on the order of 0.5 - 1MB/sec. With Audio. That's 150:1 to 300:1 compression.
TODO: Something witty here...
Ok, first problem. Yes, iPod supports USB 2.0. It also supports FireWire 1394. When 1394b comes out, Apple will support that. Apple's goal is to sell you an iPod whether you want to use USB, FireWire, swizzle stick, or anything else.
Nobody but the engineers and management of Sony, Cannon, Panasonic, and JVC know if the digital video market will migrate from dv. I doubt it, but my opinion and i'm a nobody.
Most devices don't need the power of FireWire? Personally, every device I've worked with will gobble as much power they can get if they can get it externally.
OK, I could be wrong on this but didn't FireWire get accepted as the standard to connect digital TV? Haven't followed this for a bit. But the FCC was leaning towards FireWire despite Microsoft's and Intel's begging them to accept USB 2. Why? FireWire 2.
When you talk about FireWire vs. USB 2.0, remember that FireWire 2 (1394b or Gigabit 1394) is rolling out. Makes USB 2.0 look slow just as 1394 made USB 1.0 look slow as frozen syrup.
Doing a static analysis of a dynamic world always a bit troubling unless the time difference (seconds, minutes, days) is immaterial. When not (months, years, etc.), the linearization gets shot to hell and your analysis falls apart.
The FireWire vs. USB battle isn't over. In fact, it's just begun because USB couldn't compete with FireWire a year or so ago. Once FireWire 2 rolls out, then we'll see if Intel's gambit to compete with FireWire will work out for them.
Let us go to the stars, dream new dreams, and renew the embers of hope that have long since grown cold.
DivX/XviD/MPEG are all lossy codecs for natural looking video images. The math above just does not apply if you are thinking of a remote monitor situation.
You can already do a remote desktop with Remote Desktop Protocol on Windows XP at 802.11b speeds.
Er, that's precisely what it's for ! What else would you want to use it for ?
If you need a quick connection, then use 802.11x, together with the HUGE increase in component size and battery drain that it demands - the whole point of Bluetooth is that it's VERY small, and uses VERY little power (so you can put it in just about anything). If you want speed, you're looking at the wrong technology !