Impractical, for many reasons:
- The only type of illumination medium presently known that could handle such data transmission would be LED. LED
bulbs are still in their infancy, with the only bulbs I can find locally being 40W or less. Street lights are typically high-power halogen or
mercury-vapor units of 500W or more.
- Car headlights are also extremely powerful bulbs, far beyond the ability of LED bulbs using present technology while still remaining economically
viable.
- Visible light is easily blocked by anything opaque, interrupting the flow of data. This was the major drawback of IR remotes back in the day...
they required line-of-sight communication.
- AC signals are noisy and subject to interference from any metallic, magnetic, or electrical medium surrounding them. This is the major difficulty
implementing data-over-power internet access. The same difficulty would apply to any light-based transmission source that was powered by AC power,
aggravated by the need to regulate power inputs to a device that is required to deliver kilowatts of light energy, instead of milliwatts.
- Vehicle electrical systems are also very noisy and subject to interference, primarily due to the close proximity of the circuitry to rapidly
spinning pieces of metal and high-voltage medium-frequency ignition pulses. That's why high-performance car stereos have such huge filters on the
power inputs and why controller circuits are low-speed processors and also highly filtered on the power inputs.
- "Li-fi" would require a two-way communication medium, meaning that signals would have to be pretty much omnidirectional, further exasperating the
problems with opacity.
- This also means every street light and car headlight would also require a reception mechanism.
- Light scattering would mean that amplification requirements would have to cover an immense range from negligent to astronomical.
- Tying data transmission to lighting requirements would mean that the ambient light level would be dependent on the binary stream emitted at any
particular time. Lighting would become very hard on the human eye, as it constantly tried to adjust to changes in the ambient light level.
There seems to be a move in society toward one item that does everything, a'la the iPod, iPad, SmartPhones, etc. This is a completely impractical
model. In the first place, one malfunction equates to the loss of a myriad of functions, and in the second place, upgrades require the replacement of
a range of functions that are not directly affected. In the old days, we called this approach "putting all your eggs in one basket", implying that
if that one basket were dropped, all the eggs would be broken. In financial circles, the concept of not doing such is called "diversification" and
is considered one of the best investment strategies available.
In any (realistic) server configuration, there are multiple APUs designed to operate should one fail; there are multiple servers, each one capable of
picking up the load when one fails; there are multiple hard drives arranged in RAID configurations, designed so that any one can fail without loss of
data. So why do people keep trying to tie everything together in other areas? A better, cheaper, simpler, and more reliable solution would be to mount
wifi antennas on every street light and on every PC or laptop. The transceiver mechanisms are already standardized through 802.11a/b/g/n, the upgrade
would be just that instead of a refit of the existing lighting system, a loss of one service would not necessarily affect the other, repairs and
maintenance would be simpler, and the problems of opacity and scattering would become moot subjects.
Jeff Goldblum said it best in
Jurassic Park: "Just because one can do a thing, it does not necessarily follow that one should do a thing."
TheRedneck
