# How do the different forms of data connection compare? **URL:** \ **Category:** Factual Questions\ **Created:** [April 7, 2021, 7:32am UTC](https://boards.straightdope.com/t/how-do-the-different-forms-of-data-connection-compare/939527 "2021-04-07T07:32:24Z")\ **Posts on this page:** 7\ **Page:** 1
**Author:** ![MichaelEmouse](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/michaelemouse/32/101_2.png) [@MichaelEmouse](https://boards.straightdope.com/u/MichaelEmouse)\ **Post date:** [April 7, 2021, 7:32am UTC](https://boards.straightdope.com/t/how-do-the-different-forms-of-data-connection-compare/939527/1 "2021-04-07T07:32:24Z")
If we were to compare the highest bandwidth connections from a satellite, plane, land vehicle, ship, building or cable, what would it look like? Are there any others? I know this will vary a lot but are there some instructive data points? Is it mainly driven by the amount of electrical power available? ---
**Author:** ![Mangetout](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/mangetout/32/19_2.png) [@Mangetout](https://boards.straightdope.com/u/Mangetout)\ **Post date:** [April 7, 2021, 8:19am UTC](https://boards.straightdope.com/t/how-do-the-different-forms-of-data-connection-compare/939527/2 "2021-04-07T08:19:38Z")
> [@MichaelEmouse](#): > > Is it mainly driven by the amount of electrical power available? Electrical power is usually not a limiting factor - many high bandwidth communications methods (for example optical fibre) can consume quite low power in operation. The wavelength of the carrier is a big factor - for example: light waves are shorter than radio waves, for example, so they can in theory be modulated faster to carry more information on a single carrier, per unit time. Parallel capacity is also a factor; two physical optical fibres can carry twice as much data as one (probably minus some small overhead for managing two parallel connections), but if you want to do that with a wireless data connection (say, radio), you eventually run into the possibility of one connection/channel interfering with the other. ---
**Author:** ![MichaelEmouse](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/michaelemouse/32/101_2.png) [@MichaelEmouse](https://boards.straightdope.com/u/MichaelEmouse)\ **Post date:** [April 7, 2021, 6:08pm UTC](https://boards.straightdope.com/t/how-do-the-different-forms-of-data-connection-compare/939527/3 "2021-04-07T18:08:52Z")
Thanks. For a given amount of parallelization and the same wavelength, what else could impact bandwidth? ---
**Author:** ![DPRK](https://avatars.discourse-cdn.com/v4/letter/d/4491bb/32.png) [@DPRK](https://boards.straightdope.com/u/DPRK)\ **Post date:** [April 7, 2021, 6:23pm UTC](https://boards.straightdope.com/t/how-do-the-different-forms-of-data-connection-compare/939527/4 "2021-04-07T18:23:03Z")
Suppose your “parallelization” (aka transmission bandwidth) is _B_ Hertz and your signal-to-noise ratio is _SNR_. Then your transmission rate is theoretically _B_ log₂(1 + _SNR_). i.e., more noise = slower transmission channel. In the case of a small signal-to-noise ratio, increasing the signal power may indeed proportionally increase the channel capacity, but eventually you want more bandwidth, ---
**Author:** ![Mangetout](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/mangetout/32/19_2.png) [@Mangetout](https://boards.straightdope.com/u/Mangetout)\ **Post date:** [April 8, 2021, 10:52am UTC](https://boards.straightdope.com/t/how-do-the-different-forms-of-data-connection-compare/939527/5 "2021-04-08T10:52:21Z")
I suppose other slight considerations might include the communications protocol and the need for error correction, etc, since these carry overheads for their own control features ---
**Author:** ![Melbourne](https://avatars.discourse-cdn.com/v4/letter/m/b5e925/32.png) [@Melbourne](https://boards.straightdope.com/u/Melbourne)\ **Post date:** [April 9, 2021, 2:32am UTC](https://boards.straightdope.com/t/how-do-the-different-forms-of-data-connection-compare/939527/6 "2021-04-09T02:32:50Z")
Ethernet, ‘wired’ connection, is currently at 1Gbit/sec or 100Mb/s for most people. Wireless is up to 600Mb/s, but often slower than 100Mb/s. HFC ‘cable’ is currently at about 10Gbit/sec, up from 1Gbit/sec. Under investigation is 25Gbit/sec. The rise is from both using a wider frequency bandwidth and from better bandwidth use. Originally, of course, it was zero: HFC just carried CATV. Like HFC, and wireless, optical fibres have been increasing in capacity: TAT8 was the TransATlantic cable 8, the first TA fibre optic cable, 1988, 280 Mbit/s FASTER was Googles 2016 cable, 60 Tbit/s 1 TB 1000 GB 1000 0000 MB ---
**Author:** ![DPRK](https://avatars.discourse-cdn.com/v4/letter/d/4491bb/32.png) [@DPRK](https://boards.straightdope.com/u/DPRK)\ **Post date:** [April 10, 2021, 1:11pm UTC](https://boards.straightdope.com/t/how-do-the-different-forms-of-data-connection-compare/939527/7 "2021-04-10T13:11:07Z")
Here is another illustration: imagine a short pulse of light. For instance, assume the intensity as a function of time has a Gaussian profile. The instantaneous frequency will also vary, and the intensity in the frequency domain will be another Gaussian. The shorter in time you make the pulse, the wider the spectral intensity profile will be: Δ_t_Δω = 4 ln(2). So femtosecond pulses will have a large bandwidth.