WiFi, Light Bulbs, And The FCC

JFMulder writes "According to Cringely, 802.11 WiFi wireless networking is going to get in lot of troubles when Fushion Lightning starts marketting low-power light blubs which causes interferences with Wifi signals. Read about it at I, Cringely. Supposedly the new kind of light bulb is a real electricity saver and can wreck havoc to wireless networks in a half a mile radius. So what would you prefer? Wireless networks or low cost light bulbs all around the country to save more and more on electricity?" Update: 06/13 03:52 GMT by M : Cringely confused the FHSS-or-DSSS 802.11 standard with the DSSS-only 802.11b standard, but the general warning about the potential for interference is certainly troubling.

Comments

For crying out loud...

[The_Deacon]

This was posted -- what, two or three weeks ago? Come on guys!

The old story even had a poster who mentioned that he'd used the lighting technology Cringley mentioned, and it's nowhere NEAR primetime, so it won't be causing probs for several years, if ever.

Not a problem, an opportunity!

[NewtonsLaw]

Well the answer appears very simple to me.

If these lightbulbs are emitting RF in the 2.4GHz spectrum then when will some smart-assed entrepreneur come up with the dual-function lightbulb/WiFi node?

Half the guts is already there -- the transmitter.

If every household and business had these bulbs, think of the massive 802.11 network we could build!

Each bulb could become a node in a new, better, "brighter" Internet.

Okay so I'm kidding!

Of course if that doesn't work -- why can't they just use some sheilding on these bulbs? A very thin (transparent) metal-film conductive coating (of the type they use on LCDs) should do the trick quite nicely and at minimal cost.

Uncontrolled frequencies are doomed anyway

[jukal]

802.11xx are doomed anyway, as long as the frequencies can be used without regulation. Yes, wlan shrinks cells automatically and delivers less when there is more users and traffic, and yes, you can "just" add network elements. However, when it really becomes a success story, it is doomed. WLAN with it's uncontrolled frequencies just will not work in very tightly populated areas if a significant percent of people begins to use them. And we don't even need any assisted interference to achieve the congestion.

Too much power?

[forand]

Anyone know exactly how much power these lights are supposed to give off? If these are supposed to save power better than current technologies(e.g. florecent) they need to put out 12W. But the claim is that it interfers with 2.4GHz so how much power is going out in that band if the whole thing is only using 12W? It seems unreasonable that 12W falling off at 1/r^2(okay I assume a sphereical bulb) would have enough power to interfer with WiFi .5miles away. So does anyone know the power output(or usage) of these lights and exactly what intensity a WiFi will pick up?

Re:Too much power?

[wowbagger]

It takes very little signal power to disrupt radio communications.

In radio, you measure signal power in dBm (decibels referenced to 1 mW). A typical narrowband FM receiver, like a cell phone, has a sensitivity of about -120 dBm - in other words, if the phone is getting -120 dBm at the antenna port it will just barely work. -120 dBm is one 10E-12 of 1 milliwatt - or one thousandth of a picowatt.

For a wireless LAN card, I think they usually want to see about -76 dBm for a good link - that's 10E-7.6 of a milliwatt being received, or about a twentieth of a nanowatt.

Let's say the lamp is a 5 watt lamp (and from what I understand most of these lights are more like 100 watts excitation energy - these aren't your floor lamp!) Let's say the lamp leaks .01% of its power - that would be half a milliwatt. Assume the leak is an isotropic radiator (radiates equally in all directions). So you have .5 mW (= -6 dBm) into a 0 dBi radiating antenna. Your wireless card is about +30 dBm into a +3dBi antenna - so you would have -6 dBm interference vs. +33 dBm ERP for the wireless transmitter, assuming they were both at the same place. That wouldn't be a problem.

Now, if the lamp is anything like the microwave pumped lamps I used to work with, you are talking about 500 watts or more of excitation energy - that's +20 dB. If the lamp leaks 1% of its signal (still small enough to have no real effect on the light output) that would be another +20 dB of leakage. You now have taken the lamp from -6 dBm ERP to +34 dBm - and you are now just as "loud" as the network card.

Also, these lamps are pumped by magnetron tubes, same as your microwave oven. These aren't nice, single-frequency sources - they spatter over a fair chunk of the band. So you cannot modulate them and use them as network nodes, and you cannot easily skip over the frequencies they use - they don't just impair one channel of a frequency hopping system, the impair many channels.

However, I have to wonder how the efficency of these lamps compares to the new LEDs on the market - it may be a moot point.

What nonsense.

[AlphaOne]

I find it incredibly hard to believe that one of these low power lights can radiate so much field strength in the 2.4 GHz band that it will knock out wireless networks for a half mile.

These lights are governed by the same standards as WiFi networks with regard to field strength. Namely, FCC Part 15.

These light fixtures would likely be considered "incidental radiators" by FCC Part 15.

An incidental radiator must use "good engineering" practices and must not cause harmful interference to radio services.

It seems to me that wiping out a half mile of wireless networks is harmful.

Just for sake of argument, let's bump these lights up a notch to "unintentional radiators," which means they generate radio energy internally for whatever use but do not by design radiate it into space. In this category, they are limited to 500 microvolts per meter of radiated field strength as measured at 3 meters distance.

This is exactly the same field strength limitation placed on intentional radiators in the 2.4 GHz band.

This means that these lights may only produce as much radio energy as a WiFi base station/client card with a unity gain antenna.

The FCC has also classically ruled against unintentional radiators which cause interference with intentional radiators due to their excessive field strength, regardless of whether they meet the requirements of Part 15 or not.

The FCC normally requests that unintentional radiator manufacturers show good faith by being far below the legal limits permitted in Part 15.

I'm not even going to go into the fact that WiFi is a spread-spectrum system and is very immune to traditional forms of interference. Unless these are spread-spectrum, intentionally radiating low power lights, I don't think we've got much to worry about.

Also, whomever thinks the FCC just doesn't care what goes on in the unlicensed portions of the spectrum is wrong. They certainly don't chase down every Part 15 violation, but they do randomly sample finished products from a variety of manufacturers to determine their compliance.

The manufacturer gets into trouble if these things don't meet Part 15 requirements, so these lights will simply never get off the ground if they interfere as much as it has been said they do.