Question wifi signal strength on 2.4 vs 5 Ghz

buddhaseeking

Honorable
Jan 25, 2016
46
0
10,530
The setup is given below, before asking the question:
  • Router = Asus RT-AC68U (A dual band great router with combined speeds of up to 1900 Mbps )
  • Only 4 wireless clients in a single family home residential area, all operating in N mode currently.
  • Client in concern now is a desktop (Win 10 + Linksys 2.4 Ghz USB wifi adapter).
  • Client desktop (16 GB ram + SSD 2015 dell server) is separated by 4 drywalls (25 ft straight line) from router.
  • Signal strength reported by "netsh wlan show interface" command shows 80 to 85% signal.
  • Bandwidth seen in Windows "wifi STATUS" message box varies from 35 Mbps to 108 Mbps.
  • GMon (a wifi monitoring utility on android) reports RXL ( signal dB ) around -60 near client.
  • There are about 6 other routers seen by GMon between signal strengths of -70 to -80 dB.
  • There are 12 other routers seen in neighborhood with further weak signals in client area.
Now, the issue is with 5 GHz. In order to improve my bandwidth and ping to game servers, i bought a TP-Link AC 1200 PCI network card. Tried it in desktop and found that i get weaker signal strength even though download speeds improved (as seen by speed test). With new network card "netsh wlan show interface" command shows a signal strength between 35 to 50% at client. Also the ping to the game server increased (completely negating my faith that a PCI card and network - AC would improve situation). The desktop is placed on floor and the PCI card wound up being lower than the USB network adapter, even though there are couple of antenna on the PCI card.

I am going to return the Tp- Link (because it is also causing sudden computer freezes; perhaps not a good driver). I am not concerned about data bandwidth because the game does not need a big bandwidth but needs low lags. Game requires less than a 1 Mbps data transfer and my cable modem provides 72 Mbps / 6 Mbps speeds. I do not think i am gaining anything by going to wifi AC as opposed to running my router in pure N protocol mode. N300 running at 5 GHz would give me plenty (provided the signal strength can be boosted at the 5 Ghz). I suspect the situation did not improve with new card, due to poor signal strength of 50%.
  1. Buy a wifi N600 or N300 PCI network card + separate antenna set that can be placed about 3 feet away from desktop at proper height/location.
  2. Buy a USB network adapter that has antenna separate from USB (so that i can place antenna on top of table).
  3. Buy a signal amplifying antenna (5 or 9 db ?) and attach to the PCI network card.
  4. Budget is a consideration, so anything over $50 means that i have to live with the 20 or 30 ms additional lag
  5. Running an ethernet line is out of consideration due to various other issues.
 
If you are going to buy something new I would instead try powerline network equipment especially since you have issues with wifi. You want the av2 units I suspect you will be limited to ones that have the 600 number by your budget. The best ones are 1200 number. Obviously you get nowhere near those speeds. Avoid the older av200 and av500 units but they do work if your budget is really constrained. You should see 30-50mbps in most houses on av2-600 units.

The key disadvantage to pci cards is the antenna are so close to the machine which has lots of metal to block the signal. You can get antenna extension cables but you only want fairly short ones since you lose a lot of signal in the cables. This is why USB tends to be easier BUT usb is designed for the portable market where size and battery use are a concern. You have to really read the specs on USB equipment to be sure it is transmitting at the maximum legal power.

Ignoring 802.11ac/80211n since that only represents how much data they can put in a signal. The actual signal levels are purely a function of radio output power and how much is absorbed. 2.4g tends to not be absorbed by the walls as much as 5g. Kinda why you can see more neighbors 2.4g signals than 5g signals. In theory at least all devices transmit at the legal maximum levels so replacing one device with another generally make very little difference. There are differences because very slight difference in antenna placement can make large differences.

What makes this complex is people intermix the signal levels and the data transfer rates. This makes the question massively complex to compare since the method you attempt to encode the data can be very different at the same signal levels.
 
  • Like
Reactions: kanewolf

buddhaseeking

Honorable
Jan 25, 2016
46
0
10,530
There are differences because very slight difference in antenna placement can make large differences.

This hit the spot. I could see signal strength gain 10% simply by orienting the antenna on the new TP-link network card. BTW, since the USB 3.0 ports on this desktop are available both in front and rear, i have put the USB dongle in the rear. So, there is no metal in between the router and the USB client. Direct line between router and USB dongle only has drywalls and wooden cabinets, as obstructions. The main problem is that GMon sees only signal strength of - 75 (RXL value) in the 5Ghz band, whereas i get -60 RXL in 2.4 Ghz band. If i could manage to to boost the 5Ghz signal (from current 50% to 80% as reported in netsh wlan command) at the client, there would be no worry of collisions on wifi from neighbour's. Reading on the net, says that i need a signal of -60 or better for a good wifi connection. So, i was thinking of the following options overnight :
  1. find a wifi N (or AC) network card (or USB) that has a hi gain antenna as the replacement. (5 dbi OR 9 dbi ??)
  2. instead , i can replace the antenna on the ASUS RT-AC68U router with high gain antenna, though i read somewhere that this would damage the internal parts of the router (which i can not afford to damage, since this router is over $140).
The power line equipment you mentioned are going to act similar to a wifi repeater ?? I have read that repeaters and bridges would cut my bandwidth and hence i have avoided them. Will read about them today, and await your response. Thank you.
 
Replacing antenna tends to be old information. When a router used only transmit only a single stream of data it used to be simpler. Now that devices transmit multiple signal at the same time it is much harder to just replace antenna because they are very carefully tuned based on their size and placement. In most cases it make very little difference when used inside a house. Outside with only the air absorbing the signals it is much simpler to say but inside the walls tend to be able absorb lots of signal.

Powerline equipment looks like a ethernet cable. They make model with wifi on the remote end but that is same as plugging a AP into a ethernet cable. This will increase you cost so I would stay with the ones without the wifi unless you do not have ethernet on your machines.

The key reason you use them for people that play games is the latency is consistent. It does not get lag spikes like wifi does because of external interference. It all depends on the house so you never really know if powerline is going to be faster or slower than wifi. Both can get rates in the 400mbps range in good conditions.
 
  • Like
Reactions: buddhaseeking

buddhaseeking

Honorable
Jan 25, 2016
46
0
10,530
Thank you. There are couple of things that i do not know for sure (even though i have computer background).... Perhaps you have worked in hardware networking and know these things ......

  1. In a real world situation, where the ping to the internet server is in the order of 30 ms to 70 ms range, and my ping to the wifi router is only 2ms, is any upgrade really going to make a difference ? The hop to the router and back is only 4 ms total in roundtrip, and that is less than 10% of lag. Even if i make the signal stronger at client and move into 5 Ghz band, i may only cut the round trip to router as 2 ms total.
  2. Is there anything i can setup in the router (which is running ASUS-merlin firmware), to create a static route to the internet server that has a fixed IP address for gaming ? Will that be a better way to reduce the ping to that server from my router ?
  3. Finally, even if i cut the lag by half, i would net only about 25 ms gain. Is that really going to make a difference ? I think the problem actually may lie at the gaming server in processing the game data, and that does not seem to be so good in doing that.
 
In general the games adjust for the delays and unless it is really high like 200ms they do a pretty good job. They artificially delay people with really fast times to compensate for slower users but they only to that to a point. Of course they never tell the faster users that the ping to the server is not their actual delay because some is being added artificially.

You pretty much can't do anything to reduce the delay outside your house. It is mostly a measurement of distance.

The main reason not to use wifi is that its delays is unstable at times. If it is a couple ms all the time it works great. It has the problem that if it get interference it will keep trying to retansmit data which causes large delays. It is more the random delay spikes either in your wifi or in the internet that cause problems. The server can not adjust for delays that change all the time. They try to make what you see on the screen represent what is going on in the server. They project future positions based on the delay and send you that rather than the current positions.
 
  • Like
Reactions: buddhaseeking