Does An AC Router Improve N Device Performance?

[sfx2000]

While antenna does not resolve 2.4 GHz issues, it's still really key for any well performing wireless. No matter it is Wi-Fi or cellular.

AP antenna bi-directional gain is the only way to enhance client uplink radio link budget, if clients are what they are (phones etc). Proper antenna (directional, vertical, horizontal gain) allows focusing AP transmitted energy more to target area and helps to reduce utilization elsewhere. Directionality helps also to reduce impact of interference from unnecessary directions (for example from upstairs and downstairs). The more dense environment, the more it has impact. It's also important to remember that antenna gain does not end at 0 dB. Many small end user terminal antennas actually attenuate signal both ways. Antenna may have omnidirectional nature and effective average gain may be - 3dB due to implementation losses.

A lot of people found these two blogs on antennas useful.

Antenna design is critical, without a doubt - but at the same time, well designed antennas can be done internal or external - but again, belief bias by SOHO consumers equate big honking di-poles, and lots of them, as being better...

A well designed internal array can, and often does, perform better...

 

[azazel1024]

One advantage to external antennas, even fairly small 2-3dBi ones is that you can reposition the antennas to more effectively cover the area where you need wifi coverage. Internal antennas limits your ability to do that without very creative solutions. I don't disagree that a well designed internal array can sometimes be better, but if I had to pick which design was more likely to perform better, I'd pick an external antenna design as being most likely to be better than an internal design. That isn't a rule though, just a strong guideline.

 

[azazel1024]

One thing I forgot to mention, the most important part of ANY wireless network though is testing. You won't know which is better until you've tested it.

You can confidently claim that 20MHz 2.4GHz or 40MHz 2.4GHz is better, but you won't know until you try both under as many "typical usage" scenarios as you can. Same thing with better antennas*, access point placement, antenna alignment, etc.

Yeah, who wants to spend half a day testing their wireless, but since I have done that, on many occasions with many routers, I've found it can pay off HUGELY. Compared to "how I would have aligned my antennas, where I would have put my WAPs and what antennas I might have stuck on them", I have easily 20-30%** better performance averaged out across my entire home by right sizing my antennas, aligning them the way I have them instead of straight up and down, or all at varying angles or what not, as well as where exactly I've placed my router and wireless access points.

I wouldn't say that is life changing, but to me it is well worth the 20-30hrs of wireless testing I've done on various products as I buy/acquire them and implement them in to my network, as well as the maybe $20 on antennas I've bought or salvaged off other products.

*Bigger isn't always better, I saw very nice gains on my Archer C8 by going from the 3dBi to 5dBi dual band omnis on it, but trying 7dBi, saw a performance decrease in almost all test locations and both bands. A little better than the 3dBi on average (about 5% better 2.4GHz and 5GHz), but a fair amount worse than the 5dBi (about 5% worse on 2.4GHz and 15% worse on 5GHz) averaged out. The 7dBi on my AC1200 and WDR3600 sees about 5% better average 2.4GHz performance and 15% better 5GHz performance, I suspect based on testing that larger antennas would not improve performace on the AC1200 or WDR3600 on my indoor location and would likely hurt performance, but 9dBi might improve performance on my outdoor WDR3600 since the elevation difference is such that you'd likely be well within the HPBW of even 9dBi omnis in almost all use cases.

**Which also means in some cases, it is only as little as 10-15% faster, but in some cases it is easily 50% faster.

 

[sfx2000]

One thing I forgot to mention, the most important part of ANY wireless network though is testing. You won't know which is better until you've tested it.

You can confidently claim that 20MHz 2.4GHz or 40MHz 2.4GHz is better, but you won't know until you try both under as many "typical usage" scenarios as you can. Same thing with better antennas*, access point placement, antenna alignment, etc.

I'm a systems engineer - I've done the models, and done the testing - the data backs me up...

Most significant change for WiFi, whether it is 2.4GHz or 5GHz, is location of the AP in relation to the clients...

And again, I'll respectfully disagree with you on big antennas and wide channels on general principles - I do appreciate and understand your comments, and for some, your lay testing is good information, no doubt... we all want to have the best performance on our wireless networks - this is why we have options...

 

[azazel1024]

Well, certainly the location is the most imporant point. Since that is only moderately controllable in a lot of circumstances, you need to fall back on your next options for maximizing network performance.

Wide channels only seem to be a downside at very long ranges or if interference is present. In any location I've tested with a pretty large amount of consumer client and base station gear, the only time I've seen wide channels perform worse is if range was very high to the base station or interference was moderate to extreme. The lower the level of interference, the further from the base station you can be, before wide channels perform worse than narrow channels.

Same with antenna gain, the higher the interference, you either want to go very high gain if you can "isolate" out the interference source, or you want to go very low gain, to reduce how much of the noise is picked up. I've seen both strategies be good. I have a friend who lives just down a hill from an apartment complex. Swapping on 7dBi antennas instead of the 3dBi he had made a huge positive difference, because the apartment complex is about 60ft up and 100ft over from his house. Another friend who lives in a condo, swapping some 2dBi antennas on to his N66 helped out some.

 

[sfx2000]

@ Azazel1024 - I'm going to roll back to my diplomatic standards hat... in other words, I'm being really nice while violently disagreeing with you on this topic...

Doesn't mean we can't do lunch. More alike than different...

I hear you, and I appreciate your findings - but I will disagree on general principles... I know my stuff, and perhaps you know yours as well... ultimately, you're wrong, IMHO... but I respect that...​

That being said, we all appreciate your contributions here... please keep them coming, and I appreciate our discourse...

sfx

 

[azazel1024]

I am not trying to be a mule post, but then explain why (or what is your experience with) wider channels almost always only result in a very small gain. That isn't my experience at all, and it isn't the most extensive in the universe, but it also isn't some sheltered limited experience either. I fully understand the loss of signal strength, yet you are utilizing twice the spectrum and because of the larger contiguous spectrum and OFDM, you have greater than double the potential signaling rate (150Mbps per stream instead of 72Mbps per stream).

With the exception of fairly long range or fairly high levels of intereference (co-channel intereference within 20dB of the client/base station RSSI), I have never once tested a setup where 40MHz channels on 2.4GHz or 5GHz didn't result in at least a gain of 50% in real, utilizable bandwidth short of some fairly specific limitations with the client or base station (for example, client bus being limited to less than signaling rate, or base station being port limited below signaling rate) and in most cases it is roughly double 20MHz channels, except at reasonably long range (signal levels below -70dBm) where 20MHz begins to catch up and tends to eventually surpass 40MHz (seemingly in the -80 to -85dBm range, but I haven't tested too much out at those levels).

I am all about being a good neighbor with wifi when it is appropriate and wide channels when you have actual nearby neighbors utilizing the same general spectrum doesn't benefit anyone. However, if you don't have those concerns, I don't see much of a reason to use narrow channels unless you have to stretch range to the absolute maximum and/or have a situation where you have dense deployments with multiple overlapping base stations and large numbers of clients.

Again, maybe all of my experience really is an edge case, but I at least don't feel like it is with doing or helping out with at least a dozen different people's setups from fairly rural like mine, to friends living in apartment complexes or townhouses with dozens of nearby networks (and yes, in those dense setups, 20MHz generally performed far better than 40MHz. In all of the reasonably suburan or rural setups, 40Mhz performed far better).

 

[Steve40th]

I just bought the V2 of this router. Can someone recommend settings to get it going. Its in auto right now, suing N wifi cards only till 7260 card shows up.

 

[azazel1024]

V2 of WHICH router?

 

[Steve40th]

V2 of WHICH router?

tp link Archer C7

 

[azazel1024]

I'd turn UPnP off. Turn off things like the print server unless you are using it. Run something that'll show you what channels your neighbors might be using and then move the furthest from them. You'll want to select channel 1, 6 or 11 for 2.4GHz. If you don't have any nearby neighbors, use 40MHz, if you do, use 20MHz. 5GHz, select the higher channels and 80MHz. Make sure remote administration over the internet is off (it is by default, shows as 0.0.0.0 for the IP). Use WPA2/AES for the wireless security and at least an 8 character long pass phrase that isn't a word or doesn't contain a word.

 

[System Error Message]

I think the final and easiest answer to this thread is no. Theres so much extended discussion that isnt even relevant. Better chips and firmware mean faster performance but it has nothing to do with the fact that a newer protocol is used.

 

[thiggins]

I think the final and easiest answer to this thread is no. Theres so much extended discussion that isnt even relevant. Better chips and firmware mean faster performance but it has nothing to do with the fact that a newer protocol is used.

I don't think the data supports your statement. I started with the highest performance N router as the basis for comparison. Perhaps there are other N routers that would produce similar improvements. But the industry has moved on to AC.

The data do support the statement that AC routers CAN improve N device performance.

 

[System Error Message]

i did say if better chips were used. Obviously newer chips and better firmware tend to be better.

 

[Steve40th]

I went from a Linksys E3000 to a TP link Archer C7 and , even though its theoretical via windows, shows over 360 Mbs sometimes 405, vice 130s before.. Not accurate, but something changed.

 

[zakoh]

Replaced a TP-Link WDR-3600 with a TP-Link Archer C8 (AC1750), and the improvement was huge, even in 2.4GHz band. Moreover, in locations where 5GHz connectivity was either unusable or very marginal, the 5GHz data transfer rates are somewhere in the neighborhood of a 100mbps wired lan. Very good improvement. I don't have many n-only devices left, but I am going to do a test later with a Laptop that has a 1x2 MIMO wireless-n card. Can't imagine how one could possibly recommend an n-class router. It's not worth it, even $50 a pop (what I paid for the WDR-3600, it's a big paperweight now).

 

[Steve40th]

Antenna design is critical, without a doubt - but at the same time, well designed antennas can be done internal or external - but again, belief bias by SOHO consumers equate big honking di-poles, and lots of them, as being better...

A well designed internal array can, and often does, perform better...

Antenna within your computer is critical too? External antennas available?

 

[sfx2000]

Antenna within your computer is critical too? External antennas available?

On built-in wifi radios, typically not - external USB dongles, there are options, as well as PCI cards for desktops...

I'm sure there may be options for ExpressCard and CardBus cards, but does anybody even sell them in the general big-box stores anymore?

 

[htismaqe]

The thing about external antennas, or USB dongles, is that they can be repositioned, which often makes a LOT of difference. For example, when I was monkeying with the Asus RT-AC68U, the antennas HAD to be in a \|/ configuration with the two outside antennas at a 45-degree forward angle. Anything other than that and 5Ghz coverage to half my house was non-existent. With the one desktop PC we have here, we use a Linksys AE1000 with a USB cradle extension, which literally means the different between the radio being inside of a desk or hanging on the wall 4 feet above the desk.

 

[sfx2000]

The thing about external antennas, or USB dongles, is that they can be repositioned, which often makes a LOT of difference.

Exactly... with a lot of PC Towers, there's no good RF solution with a couple of dipoles out the backside...

All in One Desktops - it depends, some are good perhaps, some aren't - similar to Laptops with regards to internal cards and antennae...

USB adapters - it keeps the RF local to the dongle... can extend/move as needed...