Having problems with wireless signal from a LINKSYS wireless router. Got it located in a bedroom on middle floor of a three floor house. Signal strength is poor/negligible donwstairs in the living room. I saw a couple of suggestions whilst browsing google groups to move the antennae on the router.
Does this actually make any difference? Should I have 1 vertical and 1 horizontal as one poster suggested?
to tune the antennas i sit one computer where the signal isnt go great.. and transfer a large file (like 1gig) in something I can monitor the throughput with, to that computer. and I watch the throughput while playing with the antennas till i get it as fast as it'll go.
Have you tried/considered one of the third party firmwares for your device, if available? The hardcoded power output in the default linksys firmware on mine is 28mw but the AP is capable of putting out over 100mw with other firmware builds.
It switches between the two antennas for both transmit and receive. An inherent problem exists because the received signal strength is used to determine which antenna is selected, and that may not be the best antenna for the next transmission. In receive mode the radio periodically switches antennas to determine if there is a better received signal. The next transmission, even if it is to a different client, will be on the last antenna selected for a good received signal.
One example that demonstrates the potential problem, would be if two highly directional antennas are used, one pointed North and one pointed South, to two different clients. If the last received signal is from Client North and the next transmission goes to Client South... the data packets sent to Client South will not be received until Client South, for whatever unrelated reason, happens to make a transmission (which will cause the receiver to lock onto the South antenna for the next transmission).
Such a system will work very well for periods of time (while all data goes to one of the two clients), and will have dropouts and long delays at other times (any time data is being sent to both clients).
A couple of conclusions about wireless network topology and configuration can be drawn from the above. One is that if there are multiple clients, the two antennas should be identical. Another is that if there is only one client (e.g., a roving laptop at home), then there could be two very different antennas to provide coverage in two very distinct areas of the house.
Changing the orientation of the antennas probably amounts to having two very distinct antennas, because the difference between horizontal and vertical polarization can be much more than 25 dB of signal strength (which is a *lot*). But, by the same token... in most locations there are so many objects which reflect 2.4GHz signals that the actual difference between horizontal and vertically polarized signals by the time they get from the transmitter to the receiver is probably less than 6 dB. And that means having one antenna vertical and the other horizontal might actually provide just about the same benefits as having them separated by 6 inches! Don't do that with a point-to-point link, but it is worth trying in a home or office setting.
All one has to do is try it, and use software that shows signal strength. Just keep in mind that it takes 20-40 seconds for the effects to register because the radio reports a running average rather than an instantaneous signal level.
Wrong. There were some early wireless access points that did this, but they are long gone. These days, both antennas do transmit and receive. Search Google Groups in this newsgroups for "diversity receive reception" for some long detailed guesswork I posted on the topic.
Wrong. Since both antennas are *SWITCHED* for diversity receive, only one is on at a time. Combining these doesn't result in any tx power increase or rx receiver gain. If it did, you would certainly see commerical products offering such an arrangement.
Funny you should mention that. That has been something I've always wanted to do. It was going to be an RF sensitive "eyeball" that would allow me to see RF flowing in circuitry. We'll, there's a small problem. The receptors in your eyeballs are about 1000 wavelengths of light across. The equivalent eyeball that works at 2.4Ghz would be about 125 meters across for each sensor. That would be one huge eyeball. Another way of looking at it is if it could be built with the same scale as a human eye, then the ratio of wavelengths would be the same. 12.5cm (2.4GHz) / 700nm (light) = 180,000 times as large, resulting in one HUGE eyeball. I don't think so...
Sorry. My venture capital funding for my microwave eyeball project ran out.
Only one antenna is used to transmit. Both are used to receive: the router compares signal strength from each antenna and uses the stronger of the two. Might even be smart enough to properly combine the two for maximum, but I doubt it. In any event, adjusting antenna orientation may help.
Changing the router position - even by a few inches - may help as well.
Radio energy at these frequencies bounces off obstacles and often re-combines with itself to produce local hot or deadspots. It would be dandy to have magic goggles that could display field strength. Maybe Jeff will loan you his.