"Why High Power Routers Don't Improve Range"

[Bulldog]

Good points succinctly made. One of the things I love about SmallNetBuilder is how you effectively defuse marketing with facts.

So, if I understand you correctly, if I'm basically happy with my venerable (and no longer supported) D-Link DIR-625 router, except that I could benefit by faster throughput and fewer dropped connections, I should first try upgrading the router's antennas before considering a new router?

Or, in a larger sense, would it be correct to say that if you have a router that is up-to-date, standards-wise (i.e., 802.11n), you should only consider a new router if the new router has new features that you desire (like guest access, DLNA server, dual-band, etc.)?

 

[BrianV]

Please note that amplifying RF signals with an amplifier and/or optimized antennae on the receive path will increase the sensitivity of the receiver resulting in an overall enhancement of effective range for ALL devices on the network (only if you have amplification on both chains). Even if the small, low-power device isn't amplified, the amplified AP can effectively hear the smaller device from further away due to its superior receive sensitivity. Furthermore, in an infrastructure network, all client traffic goes through the AP, it can help extend the range of all client communication. As Tim explains, amplifying allows the AP to YELL louder, but it doesn't allow the other stations to YELL louder. However, an AP with equally as sensitive receiver can hear better (like a hearing aid) so at the end of the day, the devices can be further apart (increased range).

As such, overall range is increased with amplifiers, but one must determine that the actual product has amplifiers, and that receive sensitivity on the AP is similarly as boosted.

Here's a great PDF on understanding how receiver sensitivity affects Wi-Fi performance:
http://www.tropos.com/pdf/technology_briefs/tropos_techbrief_rx_sensitivity.pdf

Although it's a bit dated more for 11g and mesh type issues (economy rather than full range), the science remains the same. Page 3 specifically shows the issue with standard 15 dbm clients and how high-end, sensitive APs do show a net gain. Amplifying a receive antennae path increases sensitivity.

PS - I disclaim that I work for Buffalo Technology

 

[thiggins]

To improve wireless coverage and reliability the first and cheapest method is to relocate the router. Second would be to add higher-gain antennas to the router / AP.

 

[Bulldog]

Thanks to you both.

 

[Unregistered]

Directional

To improve wireless coverage and reliability the first and cheapest method is to relocate the router. Second would be to add higher-gain antennas to the router / AP.

In my parents home, I have swapped out the antennas for higher gain antennas and was wondering what you mean by directional. Are you literally referring to the front of the router v. the back of the router as being the directional area (given that the posts are positioned as most would expect, vertically)? Or is there another way to determine direction? If I can increase directional stability by merely choosing which direction the router faces, I'd be thrilled to death. Also, the router is currently on the second floor with almost ALL of the connecting devices on the ground floor. Would the router respond better directionally if turned upside down?

 

[thiggins]

You would need to look at the antennas specifications or description to tell whether it is directional.

Omni directional antennas have a signal pattern like a doughnut, with the antenna sticking up through the center of the hole. See this diagram and this article.

 

[stevech]

May I add that...
You don't get antenna gain (increased range) for "free"- it comes at the inconvenience of directionality.

An antenna can be directional in the horizontal (H) or vertical (V) plane (or a combination of both)....

Ye ole rooftop TV antenna, is directional mostly on the horizontal meaning it must "point" at the transmitting antenna (at the TV's transmitter site).

An omnidirectional antenna with good gain is omni- in the horizontal plane but narrow in the vertical - like a doughnut shape. The flatter the doughnut the higher the gain.

A (parabolic) dish antenna is directional on both H and V - it is very much like a flashlight due to it's reflector (not quite parabolic in shape). A TV Satellite dish like you see today is a part of a parabolic dish - they chop off part to make it smaller and lose some gain in doing so, but enough remains.

So the idea in directional antennas is "gain" comes from H and/or V beam focusing.

An antenna that is purely omnidirectional for H and V has a spherical shaped gain pattern. The common little stick (rubber ducky) antenna is a "dipole" antenna which has a pattern more like a dimpled sphere than a doughnut.

Of course, antenna gain is bi-directional - applies to both transmissions and receptions - which is why antenna gain is superior for range extension than are transmitting or receiving amplifiers.

 

[Hornhonker]

In ham radio land we call high power systems "alligators", all mouth and no ears

 

[stevech]

Interesting to note that public safety (police/fire/EMS) radio systems are severely "Unbalanced" - meaning the high-spot repeaters are 100 Watts or more and the handheld radios are 5W or less.

So lots of such systems have coverage-fill-in receivers scattered here and there, with their audio sent via leased phone lines to the repeater where the "best" audio among these is "voted" on by a smart audio switch then sent to the system.

With this, the unbalance is remedied to some extent.

That industry uses the terms "talk-out" (from repeater) and "talk-in" (from handheld/mobile). The talk-in is always the weakest link. Often, naive officials "buy-off" the system (authorize payment) based solely on the talk-out coverage. Oh my.

In WiFi, with a high powered access point/router, one has the same unbalanced link.

Using antenna gain does not create an imbalance.

 

[Unregistered]

Nice article, thanks.

Yes, high power routers don't extend the operaring zone. Same in practice for wireless-N.

But for a given downlink throughput, they do increase the range within the operating zone. Same for wireless-N.

This is pretty useful, given the main use for wifi is still just to transmit an asymmetric internet connection from router to client, and that the main issue users face is getting decent speeds to do it in certain areas/rooms, rather than a totally dead connection.

To follow your analogy, for downlink, the client does not need to yell back as loudly, it just needs to be heard.

I understand your use of the term "range" and it is indeed strictly correct. I just wonder if perhaps the distinction might end up being lost on some people...

 

[stevech]

I don't recall for 11n's modes, but for 11b/g, I recall that the client and access (router) device must use the same modulation mode (data speed) in both directions. This makes the from-client weak signal the "weakest link".

The per-client modulation mode can vary quickly according to propagation conditions, with the displayed speed being perhaps a snapshot in time.

 

[jamie128]

I found this a very informative article and can now understand how increasing antenna gain narrows the beam. The only part I could not understand is how that would help with receiving signals.

So lets say my phone is just out of reach of a typical omnidirectional signal, however, if I have a nice directional antenna on my router that blasts out a concentrated beam that is say in the region of my smartphone, my smartphone will now presumably try to reply using its non concentrated omnidirectional antenna, so how will the signal from the phone reach all the way back?

Thanks.

 

[thiggins]

The directional nature of higher gain antennas works for both transmit and receive.

 

[azazel1024]

Picture it like cupping your hands around you mouth and your ear at the same time (if you could). What you are pointed at both sounds louder and you are directing more of your voice at it to, so it is getting more of your sound. Everything outside of the "cone" that your hands are making gets less sound/sounds quieter.

That is roughly the same principal of how a directional antenna works with radio.

 

[sfx2000]

I don't recall for 11n's modes, but for 11b/g, I recall that the client and access (router) device must use the same modulation mode (data speed) in both directions. This makes the from-client weak signal the "weakest link".

The per-client modulation mode can vary quickly according to propagation conditions, with the displayed speed being perhaps a snapshot in time.

11n supports asymmetric MSC rates on uplink/downlink - 11ac removed this as part of the data rate/error correction/spatial stream cleanup/simplification.

 

[Grievous Angel]

Hate to bump such an old thread . . but I was wondering if anyone has had positive results using higher gain antennas on, say the N66/AC66/AC68 routers. There are any number of products offered on Amazon (3 packs for 20 bucks) . . but have very mixed reviews.

If anyone has any success replacing the stock antennas on these routers, I'd like to hear it. Particularly, comments on how it improved range . . and if you got good results, what model you bought.

 

[stevech]

Antenna gain improves both transmission and reception; this is an important distinction from the title of this thread. The point is that a thousand watt WiFi router/AP doesn't help the reverse direction signal from the client device. A typical handheld is 0.03 Watt. A typical WiFi router in the higher speeds is about the same.


Generally, changing antennas to get 2-5dBi more gain isn't very significant. This is because the signal attenuation from client to access device (WiFi router, AP, etc.) is many 10's of dB. So improving antenna gain by a few dB is small in comparison to the path loss.

Also, many manufacturers overstate the antenna gain.

A device with an integral 14dBi directional antenna is significant, but at the expense of 360 deg. coverage. High gain omni-directional antennas, say, 12dBi, have a narrow vertical beamwidth - doughnut shaped. That is a problem if roof mounted. Or in a 2-3 story house. And such an omni- is 3-4 ft. long.

 

[Grievous Angel]

Antenna gain improves both transmission and reception; this is an important distinction from the title of this thread. The point is that a thousand watt WiFi router/AP doesn't help the reverse direction signal from the client device. A typical handheld is 0.03 Watt. A typical WiFi router in the higher speeds is about the same.


Generally, changing antennas to get 2-5dBi more gain isn't very significant. This is because the signal attenuation from client to access device (WiFi router, AP, etc.) is many 10's of dB. So improving antenna gain by a few dB is small in comparison to the path loss.

Also, many manufacturers overstate the antenna gain.

A device with an integral 14dBi directional antenna is significant, but at the expense of 360 deg. coverage. High gain omni-directional antennas, say, 12dBi, have a narrow vertical beamwidth - doughnut shaped. That is a problem if roof mounted. Or in a 2-3 story house. And such an omni- is 3-4 ft. long.

Thanks for the info. I think I'm going to leave it alone. Based upon pictures I've seen from this site, the stock antenna have the "donut" shaped pattern--so, for a one story house, all three should be vertical, right?

Some have suggested the 45 degree angle for the two outside antennas, but I would expect that to be for a multi-story installation. Correct?

Sorry. . . I know little to nothing about antennas, gain, etc. I have my router sort of centrally located in a roughly square house--but it does have to broadcast through brick and glass to get to the rest of the house (and, for the most part, it does). It's basically located in what used to be a covered porch, converted to a sunroom, in the center along the back of the house (roughly 3000sqft, one level). So it has to contend with what used to be exterior windows and brick veneer to get to the rest of the house. I could move it, but I do have four devices wired directly to it in this location (it's my "man room" and has all the good stuff in it).

Again, it generally reaches most of my house:
- Wife can do wifi and stream Roku HD (even in 5gz) two rooms away
- son can do wifi and netflix from . . about three rooms away
- son can do Xbox live without issue from two rooms away
- data hog daughter is fine one room away
- wife streams video to tablet on exercise bike in the garage, probably the furthest reach in a straight line from the router to any other point in the house.

So I don't know why I'm even talking about this . . It's about all I can do not to go out and buy an R7000 or AC87U RIGHT NOW. I wasn't an enthusiast before flipping to Asus and finding Merlin and this site . . but I'm stable at the moment and figure that's worth a lot.

 

[azazel1024]

Not on those routers, but the AC1750 I have came with, what appears to be, 3 external dual band 3dBi antennas. I swapped them for some 5dBi antennas I scavenged off some N600 routers I had. Performance is a fair bit better. At my kitchen table I have enough signal strength to actually get a 5GHz connection now (a really crappy one, but Windows wouldn't connect before, InSSIDer would see it at around -88-90dBm, now I am getting -84dBm) now I can get ~3MiB/sec down, 1MiB/sec up). In most other locations, INCLUDING close to the router I am seeing 10-20% increase in throughput compared to the 3dBi antennas. A rather nice increase. That is on 5GHz. 2.4GHz barely nudged. Same performance close to the router (the same "sad" 28-29MiB/sec on 2.4GHz 40MHz and 14.2MiB/sec w/ 20MHz). Further from the router there are a couple of locations where performance is up 10-20%, but mostly its closer to a 5% increase in most locations. I did spend a heck of a lot of time testing it, but it was a quick swap.

My N600 AP went from 5 to 7dBi antennas and after finding a better spot for the AP, the swap between 5-7dBi is showing a pretty similar difference, about 10-20% better on 5GHz in most locations, only maybe 5% or so on 2.4GHz in most locations (though the furthest location sees around a 15% increase on 2.4GHz). Close in and line of sight, no difference though on 2.4GHz or 5GHz for the AP. Though at my dinning room table 10ft away and through just the side of the book shelf that the AP is sitting on, 5Ghz increased about 10% going from 5dBi to 7dBi (effectively no change for 2.4GHz). Though took a HUGE amount more testing, as the previous location the swap between antennas actually lead to a backslide in performance in 2 of the 5 test locations, almost no difference in 1 and only minor boost in the other 2 locations. So testing each band Rx then Tx, 3 times to average it, times 5 locations, times two antenna sizes and finally times 3 locations to try to find a better spot for the AP. Lots of hours (I think close to 5 hours).

Anyway, don't necessarily expect a HUGE, huge difference. Even with my AC1750 router, there are a few locations where performance is little changed. In MOST I see a good 10-20% increase on 5Ghz and as I mentioned, a slight increase in range on 5GHz (but 2.4GHz is MUCH better than 5GHz there, so it doesn't really matter) and 2.4GHz (in both cases, doesn't matter, my N600 AP covers where the "extended" range would be anyway). However, moving my butt around in my basement where the AC1750 router is, still line of sight, there are spots where performance is within 1-2MiB/sec between the old and new antennas. Most spots are more like 3-5MiB/sec faster, especially once a wall gets in the way and there are one or two spots where it is 5-8MiB/sec faster, still line of sight. I'd say 70% of locations are at least 10% faster, maybe 10% of spots where there is no real difference and 10% where it is 20% or more faster (for example, those couple of spots close in, performance went from 52MiB/sec to 61-62MiB/sec, a couple of spots medium/far went from 16MiB/sec to 20-21MiB/sec).

So, depending on EXACTLY what you need and want, bigger antennas might help out.

However, the higher the gain, the worse the coverage above and below the router will be. The only spots I care about covering above and below the router are very close to the router, so in some ways, the more directional coverage might be helping by keeping the signal level from being too high (there is such a thing), allowing for better beamforming or similar. However, I am not trying to centrally place the router and cover a floor above and below (some of the area in the floor above, but since the router at sitting by an outside wall, I am able to tilt the antennas back about 10-15 degrees to provide better coverage above the router and still provide good coverage in the basement where the router is. That antenna tilt provided the best performance in the basement AND in the areas on the floor above the router interestingly enough, it wasn't a compromise setup).

Higher gain antennas almost certainly would reduce performance, at least above the router, if I were to go with 7/8/9dBi anntenas due to the compressed beam width.

My AP, 7dBi antennas are fine, because I ONLY care about coverage on my main floor (not including the 2-3 rooms on the main floor that I want/need/have my router covering that is below those rooms). In fact, they are kind of of a benefit, so that they aren't covering the basement due to their narrower beam width. Things are setup with seperate channels already, but it still helps encourage faster roaming.

I either of these were trying to cover an entire house, at least a multistory house, Having gone to higher gain anntenas might have been detrimental in some locations. Since I already had them deployed in such a way where narrower beam widths were not an issue (or were of benefit), the higher gain anntenas were able to provide modest benefit (for the $10 it cost for the 7dBi antennas, which also freed up a pair of 5dBi to go with the other two 5dBi antennas I had to go on the AC1750 router) I'd consider it a very cheap way for ME to boost my wireless throughput in my house. I spent a lot of time testing though, if you consider time money, it was NOT very cheap. I spent maybe 5hrs testing with my AP and at least 2hrs testing with my router to both verify the gains and also to tweak positioning of the router and AP (router was pretty fixed, but I did find that moving it about a foot on the shelf it was sitting on actually boosted performance in most locations a small amount (sure, only 1-2MiB/sec, but it was consistent and repeatable).

 

[azazel1024]

Thanks for the info. I think I'm going to leave it alone. Based upon pictures I've seen from this site, the stock antenna have the "donut" shaped pattern--so, for a one story house, all three should be vertical, right?

Some have suggested the 45 degree angle for the two outside antennas, but I would expect that to be for a multi-story installation. Correct?

Sorry. . . I know little to nothing about antennas, gain, etc. I have my router sort of centrally located in a roughly square house--but it does have to broadcast through brick and glass to get to the rest of the house (and, for the most part, it does). It's basically located in what used to be a covered porch, converted to a sunroom, in the center along the back of the house (roughly 3000sqft, one level). So it has to contend with what used to be exterior windows and brick veneer to get to the rest of the house. I could move it, but I do have four devices wired directly to it in this location (it's my "man room" and has all the good stuff in it).

Again, it generally reaches most of my house:
- Wife can do wifi and stream Roku HD (even in 5gz) two rooms away
- son can do wifi and netflix from . . about three rooms away
- son can do Xbox live without issue from two rooms away
- data hog daughter is fine one room away
- wife streams video to tablet on exercise bike in the garage, probably the furthest reach in a straight line from the router to any other point in the house.

So I don't know why I'm even talking about this . . It's about all I can do not to go out and buy an R7000 or AC87U RIGHT NOW. I wasn't an enthusiast before flipping to Asus and finding Merlin and this site . . but I'm stable at the moment and figure that's worth a lot.

Both vertical in your case. Or at least generally so. You could play with it a tiny bit, because it might just setup a better constructive interference pattern of the radiation. Setting the sides at 45 degrees is a lot better for a multistory house to increase range and performance for devices above and below. However, it'll tend to significantly reduce performance for devices on the same floor once you move far away. This is because 1-2 of the spatial streams (for things that are 2 spatial streams or higher clients, like some laptops and the rare tablet) is going to end up being very weak. Effectively it is a compromise. Get the best coverage all around the router as possible, but it'll reduce the speed of everything around the router compared to the optimal antenna arrangement for each location (which also tends to mean that the other locations have the worst antenna arrangement).

For your situation, why not play with it? So long as you don't mind spending a few bucks and feel like doing a little testing, it might not hurt at all to get 3 7dBi or 9dBi antennas. Since your entire house is a signal story and fairly large, you might see a decent benefit.

I don't disagree with Stevtech that the path loss a huge inside going through walls and stuff, at the same time, if you look at the MCS rates at least advertised by most router manufacturers, it generally only takes about 3dB increase in signal strength to step up an MCS rate (MCS rate is the signaling bit rate from the router/client. Example, 300Mbps, or 54Mbps). So even if the path loss is large enough that it doesn't extend coverage in to an entirely new room by going from 5dBi to 9dBi antennas, it might increase the performance at existing locations by a noticable and meaningful amount. As per my lengthy anecdotes, stepping up 2dBi on my antennas, I am typically seeing 10-20% increase in 5GHz performance once I move away from the router/AP and 5-10% 2.4GHz performance. I'd imagine if I had jumped up 4dBi the gain would have been even more (though it might also mean losses for the router and both router and AP have antennas as big as can realistically fit in the spaces they are located as the ceiling is right above the router antennas and the next bookshelf is right above the AP with only 1-2inches to spare).

You aren't likely to triple your speeds, or even double them, or extend coverage half a dozen rooms, but if all you are expecting is mild gains, stepping up a size or two on antennas, especially on a single floor property, is likely to get you that.

That said, what gets the biggest gain is moving from a single centrally located router/AP to TWO equally seperated APs/routers to cover you area. Both between much higher signal strength through the property, on average, plus you have the benefit of seperate bandwidth. Set different channels on each so that they don't overlap, and clients on one side of the house are not using up airtime/wireless bandwidth from clients on the other side of the house. Win/Win.