I’m working on point to point and Pmpt , at 5G , I’m using antenna’s which supposed to work on the range 5.7-5.8 ..
1 - I notice that I can get the best signal strength outside that range !!! like 5320 , 5400 , …
I just wonder is this normal ??
2 - Signal strength , Ack time out , CCQ … which parameter is the reference for the other ? i can get good signal strength but very bad Ack time out or bad CCQ … and vice versa …
samsoft,
For 1: When you say “best signal strength”, do you mean by just a few db or do you mean by a very large amount? A few db can be explained by the superior propigation of signals at lower frequencies, also, depending on what card you are using for a radio, it has been shown that the Ubiquiti cards don’t have linear power output across the entire 5 GHz spectrum and while I’m not certain about the R52/H cards, I would venture to guess that they are somewhat similar.
Running radios on antennas made for higher frequency isn’t something I would recommend, at least not at high power or for long periods of time. Going higher than recommended is more acceptable than going lower due to the radiation point of the signal on the antenna. Going to a lower frequency requires a longer distance from antenna base to the point of radiation and in your case, with a tuned antenna, you could experience reflected power and high VSWR values, both of which could do serious damage to your radios. Going to a higher frequency than the antenna is designed for will just cause the point of radiation to be before the designed point and thus the antenna will lose efficiency and you can expreience harmonic issues but the reflected power and SWR won’t be as damaging.
For 2: I’m by no means an expert on this but from observation, as CCQ value increases so does bitrate, as bitrate increases, tx power will fall on most radios. If the CCQ value cannot be maintained at a high point at the new bitrate, the radio will fall back to the lower bitrate with a higher CCQ and likely a higher TX power. Poor CCQ values can also be derived from high interference, fresnel issues, multipathing and other problems that will result in poor performance. If your ack is set to dynamic, which I’m guessing it is, the radio will attempt to adjust until it finds the optimal ack for the link. If there is multipathing, it can have troubles with this adjustment and will result in poor performance even with good signal strength. CCQ is also a function of your ack timeout as if your radios are not in sync with ack timeout, they will have low CCQ, at least from my expreience. Personally, on a Point Multipoint, I would set my ack timeout slightly higher than the dynamic value attained with only the most distant radio connected. While the clients closer to the AP will experience slightly lower throughput, the radio will not spend all day trying to adjust for each client. All of this is just from my experiences and observations so take it with a grain of salt, the MikroTik guys might tell you I’m full of it, and they would most likely be correct.
Tony
nice info thanks ..
if XR cards acting like this , wht do you suggest for 5G ?
the difference sometimes about 10db ..
I put a fix value for ack , i even tried less than what i got in dynamic valuse , and its working ..
the main point is to get the max throughput , the problem is somtime i found the signal acceptable about -70db but tha ack may be 100-300 us !!! so the ccq is less than 40/40 !!!
Personally I am a big fan of the XR cards. I’ve yet to see anything except great performance where I have used them. -70 isn’t what I would consider great signal, it is right in the margin for 54 mbit and therefore can have fallbacks with very little environmental change. I’m sure if you were to graph your signal level over a 24 hour period you would see 3-6 db of difference throughout the day. 10db is a rather large variance between frequencies though, I’m a little surprised at that value. I don’t know how inaccessable your antennas are but I would suggest if you can gain 10db, replace your 5.7-5.8 with some wideband antennas, at least on your ptp link. I’m guessing you’re using dishes but some more details would be great so I can try to make a better suggestion. What is the distance on these links? Clear LOS?
Like I said, setting your ack static on all members of the network at a single value is the best IMHO. The difference between 100 and 300 us on your ack will cut your bandwidth by about 20% I would say from my testing. I don’t have any radios in front of me right now to test but I can give you some more detailed observations on ack differences this weekend. 100 to 300 is a very wide range though, there is no reason your radios should be giving dynamic values this far apart. There is a table under the wireless section in the manual that gives some values that other people have tested, you might want to give those a check and see where you fall distance wise and what you should be seeing. Also, you might want to try setting your radios to a static data rate at 18 mbits or something like that and see if your CCQ rises, then start raising data rates with static settings and see how high you can get before your CCQ falls off, run tests each time for speed and see if you actually find better throughput at a lower data rate with a better CCQ.
Just some tips, I don’t know if it’s actually going to help.
Sounds like interference.
I’m using sector antenna 90 degree at the main site , connected to 2 R52H , the clients are using flat antenna’s with R52H some of them XR5 .. distances from 1km to 8km inside a big city , LOS not 100% for all of them !!! some of them yes ..
is setting the static ack should be enough at the main site only , or it should be static at all sites ?and if i should put them all static ack , is it the same value for all of them ?the dynamic values are : 28us for the 1Km , 41us for the 7Km …
i already put static data rate to 6Mbps to get more distance , band 5G-10Mhz , Peek reading is about 2700 kbps .
i keep testing and found the lower freq i put the better results i get !!! now its 5200 Mhz …
Given what you just posted, I would have to agree with 0ldman. Especially when you say big city. There could be tens if not hundreds of 5 ghz signals in a 7 km radius from your broadcast point and with you include that there isn’t LOS to some areas, it gets even worse. If 41 us is what you are having success with on your longest link, you should set 41 us for everything. If you do some detailed reading on ack timeouts for madwifi drivers, you’ll find that they suggest ALL clients be static.
To cite the madwifi wiki:
"The bottom line is that you need to determine the distance between the radios (or the maximum distance in the case of a mobile installation), calculate the time of flight of the packets in microseconds, and then set the ACK timeout to a little more than a round-trip time as the CTS timeout as well as the Slot time to the one-way time. These settings are available in /proc/sys/dev/wifiX as slottime, ctstimeout, and acktimeout. The easiest way to change these settings is using the athctrl utility provided with the driver. For example, athctrl -d 15000 sets these parameters appropriately for stations located 15000 meters apart (approx 9.4 miles). Note that it is important that all stations that are communicating with each other use the same value. So if you have an access point in a point-to-multipoint set-up where one client is 10000 meters away and the other is 15000 meters away then you should run athctrl -d 15000 on all three nodes. "
While this applies to linux using the athctrl adjustment, the general idea should apply to all networks. MT plays dynamic attempts to find the idea ack for max throughput without packet loss PER CLIENT as I understand the design. That means, if you have a client at 4 km talking to your base and a client at 7 km talking, it will either switch back and forth or there will be loss to the 7 km client.
The entire network needs to have the same ack timeout and should be configured static in a multipoint network, IMHO. All of our multipoint networks are configured that way and while the closest users don’t see the maximum throughput possible, they have a solid, stable link without hurting the people at long distances. We played with dynamic in our network for a while and I wasn’t really impressed with it.
When you say lower freq gives you better results and you say big city, I’d guess once again, interference. 5.8 GHz is the most commonly used UNII band. In the US, 5725 - 5825 MHz UNII 3 - Generally used by WISPs - may not cause interference to other services, and must accept interference that may cause undesired operation. As classified by the FCC. If you check around though, cordless phones are usually 5.8 ghz, as are most other 5 ghz household devices.
There is no reason you should need static 6Mbps with the signal levels you are talking about, we have run 54 mbit links over 60km with ~-65 db signal. If you want more throughput, crank some of your locations up to at least 12. My 2 cents(which is even more worthless with the state of the US dollar).
After reading this tread I went in looking for a way to set the ack time to static in all my MK equipment and the options i found were either “automatic” or “indoor” and could not find the place to set it manually…
Any hints? some times things are so obvious that i have missed them more than once… 
thanks again!
nemecio , simply type the number you want …
expunge , thanks alot for this informations , but my questions is there are 2 clients in opposite direction connecting to the same Omni antenna , the first is at 4km , ack is 31 , the second is at 2.5km ,ack >200 !!! same equipment except the first using 22db flat , the second using 19db flat !!!
we have run 54 mbit links over 60km with ~-65 db signal
I read here and there about such numbers ,like : 2 points at 2 mountains >300km distance over the sea !!! its an ideal setup using 2M dishes !!! i’m talking about big cities , crowded city , lots of signals , how to win the big cities competition ??
sure , even D-Link ( 65mWatt ) can make a perfect bridge in a … desert …or over the sea …
Don’t pay much attention to the fact that it says 5.7-5.8. There really is hardly any difference between an antenna rated at 5.1 ghz and one at 5.8. In fact, I have hard that many manufacturers label the antennas differently when they really are identical…
At any rate, I use pacwireless grid dish antennas for PTP links all the time. They all say 5.8 on them, but often the best performance and signal is attained in the 5.1 to 5.3 range. So if works better, just keep it. If it’s not working at 5.8, it’s most likely interference, like has already been said.
Joe
it may be interference , but when I scan or run freq. usage i dont find any other broadcasting at 5.8 !!
Could be Alvarion, Canopy, Wimax, etc…
Just because a scan doesn’t show a wireless network doesn’t mean there’s no noise. Could be a cordless phone near your AP.
Don’t know about your area, but around here most of the cell phone companies run 5.8 ghz backhauls between towers. Doubt it would interfere that much since they usually use fairly directional high gain dishes, but it also wouldn’t show up in a scan.
yes , there is many towers of cordless phones ( senao ) in the city !!!
but do they use 5.8 or 2.4g ?
WOW - What a GREAT explanation!!!
I learned a lot!
Thank you,
Rod
