Suggestions for improving signal strength

[Edward]

I currently have two routers available, one is a Cisco Linksys WRT54G (about 3-4 years old) and the other is a Cisco Linksys E2000, purchased earlier this year.

The E2000 has internal antennas, but also has gigabit Ethernet and WLAN ports, as well as two networks, the standard 2.4 GHz, plus a 5 GHz.

The WRT54G has two external antennas and 10/100 Ethernet/WLAN ports, but only offers 2.4 GHz for the wireless network.

Whichever router is being used, it is situated two floors below the laptop location.

With the E2000, I generally see signal strengths on the laptop (running only Debian Linux) between 60 and 65% using the 5 GHz network and it will get into the 70%'s using the 2.4 GHz network. But when using the 5 GHz network, the wireless interface in the laptop will use 802.11a. On the 2.4 GHz side, it can use either 802.11b or 802.11g.

With the WRT54G, the laptop will display the signal strength between the upper 60%'s to low 70%'s and the interface can use 802.11b or 802.11g.

The interface in the laptop is an Intel PRO/Wireless 2915ABG.

The Linux network manager (wicd), will also display three or four other routers in the immediate neighborhood, all of them are using any of channels 1, 6, or 11, which are the only three non-overlapping channels in the spectrum. Only channels 1 through 11 are available in the U.S. on the 2.4 GHz side This is not a problem with the 5 GHz side, since my router is the only one that wicd will display as being on the 5 GHz side and I am aware that the 5 GHz signal does not travel as far as a 2.4 GHz signal will and the 5 GHz side also uses higher channel numbers.

I previously had the router sitting on top of one of the computers down there, which has a metal case, before I read online that a router should not be anything close to metal. Then I moved it to the top of one of the desks which sited the router roughly 2 feet below the ceiling down there and with the router in either location, the signal strengths were the same.

I have since moved the router location back to the other desk which is lower than the other, but placed as far away from the computer on that desk as I could. The closest anything-metal would be one of the speakers (only 6" tall and 3" wide) which is at least two feet away from it.

I'm running out of ideas as to how to improve the signal coming up two floors. Obviously, the 2.4 GHz side is providing a stronger signal, however, where all of the other routers the laptop finds are using channels 1, 6, or 11, all the remaining channels overlap. This wouldn't be much of an issue if there weren't that many other routers the laptop is finding. Moving the router up one floor is not an option since I would have to run additional cables from the basement up to the router and I don't want cables running here, there and everywhere.

I'm not sure if there are any advantages of simply using one of the overlapping channels on the 2.4 GHz side, or simly stick with using the 5 GHz side.

Any suggestions are welcome. Thanks.

UserAgent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.1.16) Gecko/20110701 Iceape/2.0.11 (like Firefox/3.5.16)

[Antony]

Firstly, my suggest won't be cheap, but it will give you good result. It's simple: get multiple base stations to run the same network. Make sure you get the one that supports simultaneous dual-band and 802.11n.

With multiple compatible base stations, all of them to have same wireless network ID and same password, but different channel, when you travel from one end of the house to another end, the connection might be changed from one station to next, but you stay connected.

Two ways to achieve this: Roaming Network (linking base stations via ethernet cable), and WDS (linking base stations wirelessly)

To achieve Roaming Network, you will need very long Cat.6 ethernet cable, and running through one station to next following the edge of the wall (... etc), complicated, but you will get the best result. WDS will result slower performance.

As for my own home network? I have 3 AirPort Extreme Base Stations (two are simultaneous dual-band models), and 1 Time Capsule Bast Station (simultaneous dual-band). All of them are 802.11n, and Roaming (linked by Cat.6 ethernet cable across the house).

UserAgent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.6; en-US; rv:1.9.2.19) Gecko/20110707 AlexaToolbar/alxf-2.13 Firefox/3.6.19

[Edward]

I don't bring the laptop downstairs all that much, except to keep the battery in good working order, to use it until the battery warning is displayed.

The one thing I do not want to do is to run cables everywhere. I already have one CAT 5e cable running between two floors right now.

When I do bring the laptop downstairs (one floor), as that it is that much closer to the router, the signal strength is well into the 80%'s, sometimes in the 90%'s.

One option I thought of was to bring the router up one floor so that it is in the middle of the house, but again, that would involve running more cables from downstairs, up one floor, which I don't want to do.

I have reconnected the WRT54G as well as changed the channel it was using to something other than 1, 6 or 11 and moved it even further away from anything metal, which necessitated the move of the power strip by a foot or so from its old spot and the laptop (two floors up) is now displaying a signal in the low-to-mid 70%'s.

UserAgent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.1.16) Gecko/20110701 Iceape/2.0.11 (like Firefox/3.5.16)

[Antony]

If you really want to get good signal everywhere in your house, set up Roaming Network and carefully wiring the cables. You can get the cable wired professionally if you want to. This way, you will have plenty of freedom for yourself.

Then, the not as good, but still provide good coverage, set up WDS, if you don't want to run more cables. You just need to put the new access points close to the power point. Of course, you need to have wireless routers which support such setting.

If you don't want to invest at all, you can try to place your router at different location, and change the channel numbers. Remember, the 2.4GHz wireless router is fighting with many cordless phones (2.4GHz), TV, microwave oven and neighbour's wi-fi networks.

UserAgent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.6; en-US; rv:1.9.2.19) Gecko/20110707 AlexaToolbar/alxf-2.13 Firefox/3.6.19

[Don_HH2K]

I'm not sure how much this would apply to the E2000 (as from what I've read it has three antennas), but for the WRT54G and possibly the E2000 as well, simply try changing the orientation of the router. If you have a wireless card with an antenna pointing horizontally and a router with the antenna pointing vertically, the client will be out of phase with the router, and therefore "see" a weaker signal.

There's also talk on Google that orienting the antenna horizontally reduces crosstalk with other access points in the area, but I don't know if that's necessarily true or not.

If you're keen on installing DD-WRT or similar firmware on one of those routers, you can adjust the output power of the wireless radio. By default I believe the WRT54G is set to output with either 28mW or 70-something mW (I can't remember which), but I have one that's been successfully running with 116mW for a few years without issue.

A third option would be, as Antony said, to make a WDS and use one of the routers as a repeater. I'm not sure about stock Linksys firmware, but I know this is also possible using DD-WRT.

UserAgent: Mozilla/5.0 (Windows NT 6.1; Win64; x64; rv:5.0) Gecko/20110624 Firefox/5.0-x64 PaleMoon/5.0-x64

[Edward]

I'm not sure exactly where the antennas are situated in the laptop, possibly in the lid.

But right now, wicd is reporting a decent signal strength coming from the WRT54G, at 76%. That's better than the E2000 has ever given me. I wonder if the external antennas (2) have anything to do with that...

I changed the channel number to 3. The laptop and router locations (despite being two floors apart) are also the furthest away from each other, diagonally (on opposite sides of the house).

I ended up getting the E2000 in exchange, because the Netgear N600 I originally bought (which also has internal antenna(s)), had buggy firmware in that the timezone could not be changed from the Pacific (U.S.) Time Zone, which is three hours behind the Eastern (U.S.) Time Zone, plus five "levels" of Netgear's technical support could not fix the problem. More recently, I checked Netgear's web site to see if they had eventually released a new version of firmware for that model, they have not...

I looked at DD-WRT, but prior to that, Cisco released a new version of firmware for the WRT54G and I installed that instead.

EDIT: I found some specs online for both the E2000 and WRT54G. The RF power (EIRP) of the E2000 is 17 dBm, with the antenna gain (in dBi) of the Main Antenna of 1.5 dBi, the third antenna has a gain of 2.2 dBi.

The WRT54G specs reference a transmit power of 18 dBm.

This would clearly explain the stronger signal the laptop is receiving from the WRT54G.

UserAgent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.1.16) Gecko/20110701 Iceape/2.0.11 (like Firefox/3.5.16)

[PaulD]

I am not familiar with Linux, or its tools for network measurements.
Do you know about the free product InSSIDer, from Metageek?
It is available in versions for
- Windows, and for
- Linux.
--- "This is an alpha release so we would love to get your feedback."

Notebook antennae are located in the lid/cover, behind the (LCD) display. From the Hardware Maintenance Manuals that I've looked at there may be one or two. It/they seem to be routed up the side/s and across the top (top as the display is open). It is worth experimenting with rotating the computer on the Z-axis (i.e., horizontally).

I understand that In general, the signal from an antenna is strongest to the side. Hence the router antennae should be rotated so that the receivers (computers) are not in-line with the axis of the antenna rods.
BUT: reflections will distort the signal paths, so 'whatever works best for everyone' really is the solution.

UserAgent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:2.0.1) Gecko/20100101 Firefox/4.0.1

[Edward]

I am not familiar with Linux, or its tools for network measurements.
Do you know about the free product InSSIDer, from Metageek?
It is available in versions for
- Windows, and for
- Linux.
--- "This is an alpha release so we would love to get your feedback."

Notebook antennae are located in the lid/cover, behind the (LCD) display. From the Hardware Maintenance Manuals that I've looked at there may be one or two. It/they seem to be routed up the side/s and across the top (top as the display is open). It is worth experimenting with rotating the computer on the Z-axis (i.e., horizontally).

I understand that In general, the signal from an antenna is strongest to the side. Hence the router antennae should be rotated so that the receivers (computers) are not in-line with the axis of the antenna rods.
BUT: reflections will distort the signal paths, so 'whatever works best for everyone' really is the solution.

I have not heard of InSSIDer previously, but the "wicd" network manager in Linux that is used with Debian (with the XFCE and LXDE default installtions) displays the same information, the other network names assigned to routers that the laptop will pickup, the signal strengths from them, the security mode used (WPA2) as well as the channel used by each of them. Since I changed the WRT54G router's channel to 3 and physically moved it to its current location, the signal the laptop is receiving improved a good 10%. I also have a shelf with two legs on them and have now placed both the router and my LCD monitor on top of it. The router is in the same location but is now raised about 10 inches higher than the top of the desk, so I probably might see an even slightly stronger signal now.

UserAgent: Mozilla/5.0 (X11; U; Linux x86_64; en-US; rv:1.9.1.16) Gecko/20110701 Iceweasel/3.5.16 (like Firefox/3.5.16)

[iJohnE]

I am not familiar with Linux, or its tools for network measurements.
Do you know about the free product InSSIDer, from Metageek?
It is available in versions for
- Windows, and for
- Linux.
--- "This is an alpha release so we would love to get your feedback."

Notebook antennae are located in the lid/cover, behind the (LCD) display. From the Hardware Maintenance Manuals that I've looked at there may be one or two. It/they seem to be routed up the side/s and across the top (top as the display is open). It is worth experimenting with rotating the computer on the Z-axis (i.e., horizontally).

I understand that In general, the signal from an antenna is strongest to the side. Hence the router antennae should be rotated so that the receivers (computers) are not in-line with the axis of the antenna rods.
BUT: reflections will distort the signal paths, so 'whatever works best for everyone' really is the solution.

I have not heard of InSSIDer previously, but the "wicd" network manager in Linux that is used with Debian (with the XFCE and LXDE default installtions) displays the same information, the other network names assigned to routers that the laptop will pickup, the signal strengths from them, the security mode used (WPA2) as well as the channel used by each of them. Since I changed the WRT54G router's channel to 3 and physically moved it to its current location, the signal the laptop is receiving improved a good 10%. I also have a shelf with two legs on them and have now placed both the router and my LCD monitor on top of it. The router is in the same location but is now raised about 10 inches higher than the top of the desk, so I probably might see an even slightly stronger signal now.

I currently have my wireless router (not doing any routing itself, just being used as an access point through to the main router) mounted in my 2nd floor hallway, in the middle of the house. I get 100% signal (It's a WRT54G, running DD-WRT) on the second and first floor, and about 90% in the basement.

UserAgent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:5.0.1) Gecko/20100101 Firefox/5.0.1

[Edward]

I could have put the router in the middle of the house, which would indeed improve the signal regardless of where the laptop is located, but I did not want all those cables running up one floor, beyond the one cable that presently is.

I looked at those adapters that use the electrical wiring as the network, but I have yet to hear one good thing about those.

UserAgent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.1.16) Gecko/20110701 Iceape/2.0.11

[iJohnE]

I could have put the router in the middle of the house, which would indeed improve the signal regardless of where the laptop is located, but I did not want all those cables running up one floor, beyond the one cable that presently is.

I looked at those adapters that use the electrical wiring as the network, but I have yet to hear one good thing about those.

Powerline networking is convenient, and I know quite a few people that have used it problem free.

UserAgent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:5.0.1) Gecko/20100101 Firefox/5.0.1

[Edward]

How does powerline networking, work?

For example. someone has three PC's with a LAN cable from each going into the router. Instead of going into the router, each cable goes into one of these powerline adapters. How do these connections make it to the router from that point? Do those units act like routers and take care of the overall networking, with only one LAN cable eventually going into the router instead of three?

UserAgent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.1.16) Gecko/20110701 Iceape/2.0.11

[Don_HH2K]

It's similar to Wi-Fi, where Wi-Fi is technically just a different physical layer with some software extensions to wired Ethernet (hence why it was originally called "wireless Ethernet"). While there are quite a few differences in the physical layer (since air and powerlines are two very different things for a signal to travel through), the idea's pretty similar. So most of what you know about Wi-Fi would also apply to a powerline system - you have one wall brick per device, whether that device be a PC or your router, much like in Wi-Fi you have one radio per device.

UserAgent: Mozilla/5.0 (Windows NT 6.1; Win64; x64; rv:5.0) Gecko/20110624 Firefox/5.0-x64 PaleMoon/5.0-x64

[Edward]

I Googled that after I posted my reply. Apparently, the speed is limited to only 14Mbps. The wired connections from the two Athlons (32- and 64-bit) as well as Wi-Fi from the laptop are faster than that (20-25 Mbps down, on speed tests), so I don't think that type of networking would do me any good.

UserAgent: Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.1.16) Gecko/20110701 Iceape/2.0.11

[Antony]

In my opinion, using in house powerline for ethernet is stupid, and should be avoided at all cost.

In short, you are sending a smaller waves (information) into powerline which is designed to carry standard electricity AC waves (50Hz or 60Hz, depending on country). At added information (wave) is designed to add on top of the sinusoid AV wave from electricity company, making the resulted waveform no longer sinusoid.

Think about your expensive appliances, feeding them with non-sinusoid electricity is simply to shorten the lives of them.

My suggestion is still: get a cable guy and have the ethernet cable (and all other cables) properly fitted through out the house.

UserAgent: Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10.6; en-US; rv:1.9.2.19) Gecko/20110707 AlexaToolbar/alxf-2.13 Firefox/3.6.19