[ Date Index ] [ Thread Index ] [ <= Previous by date / thread ] [ Next by date / thread => ]
On 29/07/10 16:54, Neil Winchurst wrote: > > So, generally not very good then. Slow and expensive. Just wondered. It > seems that nobody would use it unless forced to. That being so, there is > not much likelihood of much improvement in the near future (because > there are not many users, so not worth while for anybody to spend time > and money on research). For geostationary satellites the main issues are speed of light based, and the theoretical physicist in me thinks they could spend a lot of money researching their way around that one. Bandwidth from satellites is competitively priced, even cheap on the down-link, since the technology for high speed satellite data is well established, and once it is up there it is all solar powered. So "slow" is the wrong term, it is high latency. Of course for things you are use to doing that make lots of round trips for (like web browsing) high latency means slow - although good websites try and minimize these kinds of round trips since they make pages slower with any technology. It is not that hard to change a web-server so that all the data in a web page is delivered in one hit if you don't care too much about bandwidth costs, so solutions exist - perhaps I should start marketing some?! Indeed some of the multicast distribution networks were using satellite last time I looked, if you want the same streaming data over all of the USA or all of EMEA, and latency doesn't matter much then satellite is a good technology. Think distribution of data for video on demand services. > Is that a fair description of the current situation? There is huge money being spent on satellite communications, but mostly by the military (ours and the US amongst others). We've had at least three major companies try and fail at the low earth orbit satellite to satellite communications business. Search for ICO-Teledesic for background. The big problem is the nearer you get to the ground, the more satellites you need for the same coverage. So the costs escalate fantastically as the latency decreases, and the more satellites the more complex the routing topology and the faster the routing tables have to update. The prime market are mobile people a long way from, or unable to use, existing static infrastructure. Which means basically the military, ships, rigs, aircraft. Our military are doing it themselves (well paying for a commercial solution as a consortium). My guess is that if this ever becomes a realistic technology it will be piggybacked on top of military infrastructure. Like GPS, it is unlikely commercial operators could be sure of recouping the huge investment, but if the military do it via a private consortium they'll have an interest in having the consortium sell it to reduce the exorbitant costs. Also there is interest in high altitude balloon and solar plane technologies for relaying communications, but probably these will only help is specific local comms cases - think really big pop-concerts, battle field comms, where lots of comms needs to be deployed. The kind of situation where mobile companies deploy temporary base stations already. One thing that needs bearing in mind, is that solar activity can impact satellite comms. We don't know what the impact of a coronal mass ejection like 1859 would have on modern satellites. Sure serious men in white coats have done their best to make satellites as robust as possible, but useful data on the 1859 event is pretty scant, and obviously no man-made satellites were in orbit at the time. -- The Mailing List for the Devon & Cornwall LUG http://mailman.dclug.org.uk/listinfo/list FAQ: http://www.dcglug.org.uk/listfaq