best firmware for long range transmission?

[anoukaimee]

I have an RT-AC56U on AsusMerlin v. 380.61.0 (I believe that's the newest build). My best friend, who lives in the apartment above and across from me in our old house, has generously agreed to share his bandwidth with me--I am on a fixed income. I'm going to give him my newer modem and the router to try to maximize speed for both of us.

My quick (lol) question is what is is the firmware built for range? Are there any methods to tweak the TX settings beyond the max allowed (I know this is capped by different gov'ts, but not a huge concern as a individual consumer in a semi-desperate situation--I hope no one thinks this is selfish)? I've seen that you can overclock routers. I understand that can drastically reduce the life of the router, but it might be the best option. Are there any firmware versions that let you do this? Or scripts that would optimize range power (note I don't do linux, but I do pick up things quickly and can probably figure things out if I'm pointed to a ready-made script).

Note too that I tested my setup with Acrylic, which said that my signal quality is optimal, my signal-to-noise is horrible and transmission power mediocre. I have the TX settings set to the max available. Anything I can do w/in AsusMerlin (or other firmware) to rectify this?

SIDENOTE: I posted in another forum about options like MoCA, powerline adapters, and your traditional plug-in adapters. Because of the ancient, ungrounded wiring in our old house, powerline is out; he said MoCA was improbable "because of splitters" (I think this has something to do with me having cut the cord entirely, and having no cable service in my house but I'm not totally clear)--note I do have coax outlets; and plug-in adapters weren't suggested just because of the range. I didn't inquire about bridges, etc., and am researching that. His main suggestion was to get a model with 3x3 or 4x4 to use as the main router. I know my model has beamforming and 2x2, but apparently that won't cut it. Of course, the RT-AC56 doesn't have antennae, which doesn't help, either--no tin can DIY solutions. If anyone has any other comments or suggestions on alternative solutions involving extra hardware, would love to hear them.

But in lieu of trying out options that would be expensive for me, I thought that changing my firmware to something that "supercharges" the router (!!!) would be the best option to start with. I'm a little leery of DD-WRT because it seems very complex and not entirely stable, and Open-WRT seems complex, too. I am totally open to figuring it out, though. I also saw that Padavan offers a build for RT-N builds, there's nothing specific for the RT-AC66, and I know very little about it. I've been fine with AsusMerlin, but don't know if there is something better (esp. given my test results re transmission power--see above).

Are there any scripts in Entware/Optware or specialized for some specific firmware to increase range?

I know QoS settings could help us (the last thing I want to do is inconvenience my friend, since he's basically financing my internet), but am unclear on good settings, and whether they work with access controls in AsusMerlin (read somewhere that they aren't). If not, that would be a dealbreaker; I've become quite the internet addict and having him hold on to the password after configuration and removing the main router from my house for easy reboot is necessary). In any case, I'll figure out QoS after finding a good firmware that gives maximum range.

Any opinions on something that might be good to try? Desperate for help to make this work for both of us. Thank you very much for your help!

 

[sfx2000]

My quick (lol) question is what is is the firmware built for range?

None of the above - can't beat physics...

And cross-posting won't solve that one.

 

[anoukaimee]

None of the above - can't beat physics...

And cross-posting won't solve that one.

I kind of take issue with your assumption that I am cross posting. I am asking specifically about firmware, not hardware solutions to optimize: Thanks.

 

[dlandiss]

My best friend, who lives in the apartment above and across from me in our old house, has generously agreed to share his bandwidth with me--I am on a fixed income. I'm going to give him my newer modem and the router to try to maximize speed for both of us.

Your best chance of success is to swap that router for one with external antennas and buy a set of antennas with more gain from Amazon.

The trade-off is gain versus angle of coverage. To step up the gain of dipole antennas you are actually concentrating the signal in a direction perpendicular to the antenna axis. But if your goal is "across from you" it is easy with external antennas to lean them to rest at right angles to your line-of-sight to your neighbor.

Also place both the serving router and your local client in windows facing each other.

 

[sfx2000]

The trade-off is gain versus angle of coverage. To step up the gain of dipole antennas you are actually concentrating the signal in a direction perpendicular to the antenna axis. But if your goal is "across from you" it is easy with external antennas to lean them to rest at right angles to your line-of-sight to your neighbor.

The downside to high gain antennas is that it also makes the AP more prone to interference from adjacent networks, and that the benefit is only one way - e.g. more gain (or higher power) at the AP just unbalances the link, the clients need to change as well.

 

[dlandiss]

the benefit is only one way - e.g. more gain (or higher power) at the AP just unbalances the link, the clients need to change as well.

Negative on that for antenna gain--reciprocity principle. The same gain applies for reception.

It is true that higher tx power suffers the imbalance that you note. In some cases I would imagine that it is not noticeable, because within certain limits a lower s/n ratio simply causes slower transmissions, and the typical Web case is for the client to send a short URL and the server to return a big page and lots of photos. The slower client-to-server link may not even be significant.

 

[sfx2000]

Negative on that for antenna gain--reciprocity principle. The same gain applies for reception.

Ends up being a net loss as it skews the link budget for the remote client - it sees the AP as being closer than it is, and there is an element of open loop power estimates - so the client can reduce Tx power, and thusly, increases Tx errors, impacting thruput or forcing a MCS speed reduction.

802.11 was designed with a specific link budget in mind...

 

[anoukaimee]

Hmmm. Thank you all for the detailed instructions. It's a shame I didn't splurge for the AC66; I didn't foresee a special need for antennae. I'm still not completely clear on why MoMA won't work, but that's ok.

We're getting workable reception, and if I configure the QoS not to impede on his bandwidth, as well as DNS and maybe tweaking the router firmware, I'm happy not to stream media. Usually watch TV at his place anyway. Better off, maybe, to just use for internet access and calls over Project Fi, and look into radio technologies to get reliable signals for NPR, BBC, etc. Would probably do well to just entirely cut the cord, frankly.

Thank you again!

 

[kvic]

Ends up being a net loss as it skews the link budget for the remote client - it sees the AP as being closer than it is, and there is an element of open loop power estimates - so the client can reduce Tx power, and thusly, increases Tx errors, impacting thruput or forcing a MCS speed reduction.

I'm not sure exactly how much substance in SFX's explanation. His original assertion apparently ignores the possibility that the clients may install a higher gain antenna as well e.g. possible on PCIe / USB WiFi cards.

If "open loop power control" designed in 802.11, it might be good news. Users can adjust TX power on client side through parameters exposed by drivers. That's certainly fashionable on Windows and Android...at the expense of battery life for mobile devices.

Back to OP's problem, I read else where that it's an old building with 8 apartments on more than one floor. Higher gain antenna is a very good and inexpensive suggestion if applicable. Ruling that out maybe a cheap WiFi repeater off e-Bay, placed in OP's apartment. Another option is a point-to-point WiFi bridge..may get that off e-Bay as well. OP could get increasing reliability of WiFi access respectively.

 

[dlandiss]

Ends up being a net loss as it skews the link budget for the remote client - it sees the AP as being closer than it is, and there is an element of open loop power estimates - so the client can reduce Tx power, and thusly, increases Tx errors, impacting thruput or forcing a MCS speed reduction.

802.11 was designed with a specific link budget in mind...

I am not privy to the link budget implementation--the way power reduction is usually explained to the public is that the router will reduce power if the highest power is not necessary to maintain the greatest necessary bandwidth. That explanation ignores the possibility you have brought up that a CLIENT might reduce power based on incoming RSSI from the SERVER.

In any case that effect has had no perceived impact in my own experiments. Adding gain antennas to my access point improved performance quite noticeably in a brick outbuilding 62 feet away. The client link budget might be in the 802.11 specs, but in practice the antennas have extended our usable range in several installations around St. Louis.

 

[anoukaimee]

Back to OP's problem, I read else where that it's an old building with 8 apartments on more than one floor. Higher gain antenna is a very good and inexpensive suggestion if applicable.

It sounds like opinions are mixed here, but since posting, I have read about a few "hacks" that will enable you to get wireless from non-password protected WiFi providers blocks away (coffee shops, library, etc.). The most common solution has been a high gain antenna mounted outside the house, and so I thought this could be a solution to our problem. The articles are dated, though, so I'm not sure how high gain an antenna I would need for maximal range/thoroughput (ideally, including 5.2, but that's no big deal at this point). How strong an antenna would be needed?

I've read a little and these seem to require antenna built into the router itself (which I don't have in my RT-AC56). Would it be possible to buy a decent range extender and then attach the powerful antenna to the range extender antenna?

Also, I'm almost positive that the answer will be "no," but I might as well ask: how effective are the DIY Yagi or other various solutions listed here http://www.dxzone.com/catalog/Antennas/WiFi/?

Thanks again.

 

[sfx2000]

I'm not sure exactly how much substance in SFX's explanation. His original assertion apparently ignores the possibility that the clients may install a higher gain antenna as well e.g. possible on PCIe / USB WiFi cards.

Because we cannot assume that all client stations are equal - simply put...

WiFi only has about 20dB of range to work with in any context - and big antennas never work - e.g. gain increases all signals - both good (Eb) and background (No) - so Eb/No is a wash... more power at the AP, bad idea, as this unbalances the link with the client.

Not all "tribal knowledge" is correct...

 

[kvic]

The most common solution has been a high gain antenna mounted outside the house, and so I thought this could be a solution to our problem.

I've seen people done that with off the shelf device, picking up wifi hotspot ~80m away with line of sight. Sustainable speed ~8Mbps. Those hotspots aren't free but come with cell plans. Interesting solution for ppl in my region who stuck in old buildings with only 6Mbps ADSL at outrageous price. I think you don't have to go down the path of such a device though.

I've read a little and these seem to require antenna built into the router itself (which I don't have in my RT-AC56). Would it be possible to buy a decent range extender and then attach the powerful antenna to the range extender antenna?

RT-AC56U does't have external connectors for antenna. Though on the circuit board it has tiny industry standard connectors from where it's not difficult to attach external antennas. I also don't think you have to go down this path but the possibility is there.

Given the size of your building, an easier solution IMO is that: bring your RT-AC56U to your friend's place. Locate a few candidate spots that are convenient as well as providing good WiFi coverage at his place. Then bring your smartphone (with Acrylic) back to your apartment. For each candidate spots in your friends' apartment, locate the sweet spots in your place that give you the strongest signal.

After a few iterations, you shall find the best balanced spot to place RT-AC56U in your friend's apartment, and the best spot in your apartment where you receives one of the better signals. Now if possible borrow (or buy) a WiFi repeater/extender and place at this spot. Preferably this spot is also picked to give you WiFi coverage throughout your apartment.

Shall not be that tedious to locate such two spots. The end result will be your friend with your RT-AC56U have both wired and wireless Internet access. The WiFi signal is extended to your apartment through the WiFi extender/repeater placed in your apartment. Yourself will get Internet access wireless through the extender.

 

[sfx2000]

If "open loop power control" designed in 802.11, it might be good news. Users can adjust TX power on client side through parameters exposed by drivers. That's certainly fashionable on Windows and Android...at the expense of battery life for mobile devices.

It's standardized as TPC, and RxAGC is the basis for Tx power outside of the standards...

 

[sfx2000]

It sounds like opinions are mixed here, but since posting, I have read about a few "hacks" that will enable you to get wireless from non-password protected WiFi providers blocks away (coffee shops, library, etc.). The most common solution has been a high gain antenna mounted outside the house, and so I thought this could be a solution to our problem. The articles are dated, though, so I'm not sure how high gain an antenna I would need for maximal range/thoroughput (ideally, including 5.2, but that's no big deal at this point). How strong an antenna would be needed?

Interesting question - let's say you satisfy the link budget - you run into another problem....

TIME

Nothing is instant, and being let's say 1000 meters away from that coffee shop - speed of light gets into the way, along with the WiFi card processing packets and sending acks - and the phy trying to keep all the frames straight...

It'll work, just not work well - 802.16, along with LTE and 2G/3G technologies take this into consideration - while folks might do a 30KM wifi connection, that same STA won't work with an AP 2 meters away...

 

[dlandiss]

WiFi only has about 20dB of range to work with in any context - and big antennas never work - e.g. gain increases all signals - both good (Eb) and background (No) - so Eb/No is a wash

Interesting viewpoint.

I have usable Wi-Fi signals from -70 to -10dB--60dB of dynamic range.

The "interfering signals" issue is more problematic with analog communications. Same-channel interference is moderated by the spread-spectrum nature of Wi-Fi, and many of the networks I service who are asking for more range are in areas whose spectrum is not as crowded so signal strength, not interference, is the limit.

Adjacent-channel interference can be troublesome, but from here I see only one radio on channel 2 and none on channel 3, so channel 1 is quite suitable for us.

I am beginning to realize a trend on this forum. If a writer makes a comment about theory, a quick answer returns that theory means little because in practice there are so many variables. But if a writer makes a comment about practical observations, a quick answer appears that theory prohibits what he claims to observe.

 

[dlandiss]

I'm not sure how high gain an antenna I would need for maximal range/thoroughput

Your agreement with your friend to consensually share one ISP account is probably contrary to the ISP's service agreement, but I'm not telling. However, stealing Wi-Fi from the coffee shop is not a good idea.

Even if it were legal your question cannot be answered without knowing more about the setup. Let's say your router gives great performance 30' away. The OEM antennas are usually thought to have about 2dBi gain. If you could build an antenna with 20dBi gain, it would be 18dB better than the OEM antennas.

Since each 6dB should double both the transmission and reception range you might then get a decent two-way signal 240' away.

 

[kvic]

Your agreement with your friend to consensually share one ISP account is probably contrary to the ISP's service agreement, but I'm not telling.

Interesting point. Strictly speaking they aren't sharing the account i.e his friend need not share the username/password to login ISP (say PPPoE) other than himself. But from ISP perspective apparently they want to maximise profit by restricting connection sharing. I won't be surprised if there are relevant clauses that define what's a household and network could only be access from within the household. I never read the fineprints. To the extreme I guess if VPN back to home then out to Internet might be breaching ISP contract?

 

[kvic]

It's standardized as TPC, and RxAGC is the basis for Tx power outside of the standards...

I meant to say you were telling partial fact in order to support your argument. Not the full story. Perhaps skewed in your favour a bit.

I learned from BBS experts (just picked up the term from your other thread..no offence to experts in the wild) that TPC only mandatory for a sub-band of 5GHz. Given OP facing a challenge radio environment, I'm not sure if she wants to go with 5GHz between RT-AC56U and the WiFi extender. But she still has the bottom and top bands to choose from in 5GHz avoid possible crowded 2.4GHz. If most likely need to resort to 2.4 for better penetration where TPC won't apply (?).

Now comes the part of rxAGC for 2.4GHz. The missing dots for some of us are that how the TX power algorithm works. If "it sees the AP as being closer than it is, and there is an element of open loop power estimates - so the client can reduce Tx power", then from AP's perspective, it senses the client is farther away. AP boosts its TX power a bit more to help. Now it goes into a downward spiral until I guess the TX power reaches its max allowed on AP. That doesn't sound right.

Also drivers allow users to adjust TX power on WiFi cards. How does that play in the TX power control algorithm? Apparently the adjustment can reduce/increase max tx power allowed within regulations that can offset the auto reduction effect in your argument (?).

 

[dlandiss]

Strictly speaking they aren't sharing the account i.e his friend need not share the username/password to login ISP (say PPPoE) other than himself

I suspect one modem = one account. ISP account not equal to email account etc.