Download transfer rates and LAN connectivity

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts
  • Dave2002
    Full Member
    • Dec 2010
    • 17842

    Download transfer rates and LAN connectivity

    I've been slightly surprised on a few occasions recently to discover that local connectivity changes don't always make a very significant difference to some activities, such as downloading. For an example consider the time it takes to download an HD video from the iTunes store.

    With wireless data rates now fairly easily up in the 30 Mbps and above region in domestic situations, there is now much less advantage in trying a hard wired connection using ethernet directly to the router. The limit to transfers is most likely to be be the rate at which the network connection from the ISP operates, which will depend on factors outside one's domestic boundaries. This is obvious with not much thought, but the location of likely bottlenecks seems to be a little fluid, as technological advances are made and new equipment deployed.

    I'm currently downloading a file with 73 minutes left to run - and this does seem the likely timescale. I can switch the wired connection off, or the wireless connection, and providing the link doesn't crash, the remaining time does not change.

    However if I move into another room the wireless performance does drop off considerably. Initially I was given estimates of up to 20 hours for the download, which reduced as I moved closer to the router. This wouldn't happen with wire, though I don't want to trail 20 metres or more of wire around - even though I have the wire somewhere.

    I also think I could get at least a 5 times improvement in download time if I switched my connection to our other ISP - BT - but I have been concerned about their charging practices, so for the present I'm electing to do the download more slowly, but on an unlimited download account with another ISP.

    This observation about transfer times does not apply to data sources within the local LAN, where wired connections may sometimes give a very substantial improvement over wireless for activities such as file transfer. For example, installing files or applications on a Mac using remote disc sharing can be much faster using a wired ethernet connection, reducing times to just a few minutes instead of hours.

    Some users may find that wired connections are much better for them because the communications links are not shared with other users, or at least not significantly so. Households where every user connects wirelessly, and many users are regularly consuming media on demand may suffer from poor wireless connectivity as each user's share of the available bandwidth reduces as more users try to access the same wireless access point.

    Another poster recently noted the bottle neck in some powerline adapters, which are notionally capable of 500 mpbs, but use 10/100 Mbps ethernet to transfer data either from the source or to the final destination, so transfers will be limited to 100 Mbps or lower.. At the current time the major bottlenecks are still with the link to one's ISP, but in a few years this hopefully will no longer be the case.
  • richardfinegold
    Full Member
    • Sep 2012
    • 7300

    #2
    Originally posted by Dave2002 View Post
    I've been slightly surprised on a few occasions recently to discover that local connectivity changes don't always make a very significant difference to some activities, such as downloading. For an example consider the time it takes to download an HD video from the iTunes store.

    With wireless data rates now fairly easily up in the 30 Mbps and above region in domestic situations, there is now much less advantage in trying a hard wired connection using ethernet directly to the router. The limit to transfers is most likely to be be the rate at which the network connection from the ISP operates, which will depend on factors outside one's domestic boundaries. This is obvious with not much thought, but the location of likely bottlenecks seems to be a little fluid, as technological advances are made and new equipment deployed.

    I'm currently downloading a file with 73 minutes left to run - and this does seem the likely timescale. I can switch the wired connection off, or the wireless connection, and providing the link doesn't crash, the remaining time does not change.

    However if I move into another room the wireless performance does drop off considerably. Initially I was given estimates of up to 20 hours for the download, which reduced as I moved closer to the router. This wouldn't happen with wire, though I don't want to trail 20 metres or more of wire around - even though I have the wire somewhere.

    I also think I could get at least a 5 times improvement in download time if I switched my connection to our other ISP - BT - but I have been concerned about their charging practices, so for the present I'm electing to do the download more slowly, but on an unlimited download account with another ISP.

    This observation about transfer times does not apply to data sources within the local LAN, where wired connections may sometimes give a very substantial improvement over wireless for activities such as file transfer. For example, installing files or applications on a Mac using remote disc sharing can be much faster using a wired ethernet connection, reducing times to just a few minutes instead of hours.

    Some users may find that wired connections are much better for them because the communications links are not shared with other users, or at least not significantly so. Households where every user connects wirelessly, and many users are regularly consuming media on demand may suffer from poor wireless connectivity as each user's share of the available bandwidth reduces as more users try to access the same wireless access point.

    Another poster recently noted the bottle neck in some powerline adapters, which are notionally capable of 500 mpbs, but use 10/100 Mbps ethernet to transfer data either from the source or to the final destination, so transfers will be limited to 100 Mbps or lower.. At the current time the major bottlenecks are still with the link to one's ISP, but in a few years this hopefully will no longer be the case.
    I have found tremendous variability between my home and my former office, which were only 2 miles apart, and ostensibly had the same file transfer rates.

    Comment

    • Quarky
      Full Member
      • Dec 2010
      • 2621

      #3
      Interesting issues.

      I'm generally in favour of wired connections, because of low data rates in my area.

      however I do have a WiFi connection to my shack, at the bottom of the garden, about 40 meters away. This is effected with a highly directional patch antenna coupled to one of the transmit ports of the router.

      I think WiFi transmit signal boosters are commonly available.

      Comment

      • Dave2002
        Full Member
        • Dec 2010
        • 17842

        #4
        Originally posted by Oddball View Post
        Interesting issues.

        I'm generally in favour of wired connections, because of low data rates in my area.

        however I do have a WiFi connection to my shack, at the bottom of the garden, about 40 meters away. This is effected with a highly directional patch antenna coupled to one of the transmit ports of the router.

        I think WiFi transmit signal boosters are commonly available.
        I have not had much success with WiFi signal boosters and repeaters - indeed, I'm not sure if I've ever got one to work. I have some TP-Link devices which can be configured as repeaters, but I've only managed to get them to work as access points. In order to boost the signal if wired connections are not feasible, and if the wireless is difficult to use, then I use Powerline adapters together with an additional AP (Access Point) to extend the range of the Wi-Fi. This usually works quite well, though some configuration is needed.

        Some people do successfully use highly directional (parabolic reflectors?) aerials. I once helped someone to design a system to transmit over a kilometre - that's what I think he said the distance was. We calculated the power requirements from the Friis formula - http://en.wikipedia.org/wiki/Friis_t...ssion_equation - and practical derivations from that, and he obtained a suitable transmitter and receiver. I did point out that operating at high power might be illegal in the EU, and that the correct way to operate was to keep the power down at the transmitter low enough while still being high enough to ensure satisfactory operation.

        He later reported that it all worked, though I don't think he bothered to turn the power down. As he was living on a small island somewhere in the Mediterranean it seems unlikely that he would have any major problems with "the authorities" - and by using that approach he was able to get internet connectivity at affordable prices in a couple of houses owned by members of his family.

        I would not recommend the use of high power transmitters for this or similar purposes in the UK.

        Comment

        Working...
        X