for Jerry Richardson....and other rf gurus

About the walmart problem Mr. Cupp was discussing from Sylacauga (about 2 hours from us) - I’m hoping he’ll provide the hot line numbers as, frankly, right now, we have no 900 mhz radios we can go verify anything with. :frowning:

We have 3 customers whose jitter has done exactly what he describes - but they’re not around the walmart at all. As my good buddy captain kirk describes…the signal crosses near their path - I am going to post their radio profiles and a map of the area on our website at:

http://www.cyberbroadband.net/walmartproblem2007

The question I’m asking is based on these radio profiles, can the walmart stores and their new RFID equipment cause the jitter values I’m fixing to post? [I’m pretty sure the walmart site is LOS or close from this tower…so therefore, I’m expecting yes, walmart can cause this problem]

LUID: 005 : MAC: xx-xx-xx-xx-xx-xx State: IN SESSION (Encrypt Active)
Site Name : Bill Strandlund
Software Version : CANOPY 7.3.6 Oct 24 2005 12:06:56
Software Boot Version : CANOPYBOOT 3.0
FPGA Version : 070605 (DES Sched) P9
Session Timeout: 0, AirDelay 424 (approximately 11.80 miles (62328 feet))
Session Count: 65, Reg Count 282, Re-Reg Count 101
RSSI (Avg/Last): 1121/1113 Jitter (Avg/Last): 12/15 Power Level (Avg/Last): -84/-84
DnRate(A): 6978 DnLimit(A): 500000 UpRate(A): 22 UpLimit(A): 100000 (kbit)
LoUpCIR(D): 0 LoDnCIR(D): 0 (kbps)
Rate : VC 21 Rate 2X/1X

LUID: 006 : MAC: xx-xx-xx-xx-xx-xx State: IN SESSION (Encrypt Active)
Site Name : Jerry’s Used Cars
Software Version : CANOPY 7.0.7 Mar 02 2005 15:06:58
Software Boot Version : CANOPYBOOT 3.0
FPGA Version : 021605 (DES Sched)
Session Timeout: 18, AirDelay 224 (approximately 6.24 miles (32928 feet))
Session Count: 28, Reg Count 37, Re-Reg Count 27
RSSI (Avg/Last): 1215/1212 Jitter (Avg/Last): 5/6 Power Level (Avg/Last): -82/-82
DnRate(A): 6978 DnLimit(A): 500000 UpRate(A): 22 UpLimit(A): 100000 (kbit)
LoUpCIR(D): 0 LoDnCIR(D): 0 (kbps)
Rate : VC 22 Rate 2X/1X

LUID: 014 : MAC: xx-xx-xx-xx-xx-xx State: IN SESSION (Encrypt Active)
Site Name : William Hershberger Sr
Software Version : CANOPY 7.2.9 Jul 23 2005 01:49:03
Software Boot Version : CANOPYBOOT 3.0
FPGA Version : 070605 (DES Sched) P9
Session Timeout: 0, AirDelay 104 (approximately 2.90 miles (15288 feet))
Session Count: 75, Reg Count 48, Re-Reg Count 148
RSSI (Avg/Last): 1281/1284 Jitter (Avg/Last): 3/3 Power Level (Avg/Last): -80/-80
DnRate(A): 6978 DnLimit(A): 500000 UpRate(A): 22 UpLimit(A): 100000 (kbit)
LoUpCIR(D): 0 LoDnCIR(D): 0 (kbps)
Rate : VC 30 Rate 2X/1X

LUID: 017 : MAC: Xx-xx-xx-Xx-xx-xx State: IDLE
Session Timeout: 0, AirDelay 589 (approximately 16.40 miles (86583 feet))
Session Count: 1, Reg Count 1, Re-Reg Count 0
RSSI (Avg/Last): 1261/1265 Jitter (Avg/Last): 7/6 Power Level (Avg/Last): -81/-80
DnRate(D): 20000 DnLimit(D): 500000 UpRate(D): 64 UpLimit(D): 100000 (kbit)
LoUpCIR(D): 0 LoDnCIR(D): 0 (kbps)
Rate : VC 33 Rate 1X/1X

[lynn harris]

LUID: 018 : MAC: xx-xx-xx-xx-xx-xx State: IN SESSION (Encrypt Active)
Site Name : Tony Ellard
Software Version : CANOPY 7.2.9 Jul 23 2005 01:49:03
Software Boot Version : CANOPYBOOT 3.0
FPGA Version : 070605 (DES Sched) P9
Session Timeout: 0, AirDelay 350 (approximately 9.74 miles (51450 feet))
Session Count: 162, Reg Count 318, Re-Reg Count 220
RSSI (Avg/Last): 1177/1174 Jitter (Avg/Last): 7/8 Power Level (Avg/Last): -83/-83
DnRate(A): 6978 DnLimit(A): 500000 UpRate(A): 22 UpLimit(A): 100000 (kbit)
LoUpCIR(D): 0 LoDnCIR(D): 0 (kbps)
Rate : VC 34 Rate 2X/1X

LUID: 020 : MAC: xx-xx-xx-xx-xx-xx State: IN SESSION (Encrypt Active)
Site Name : Judy Lowery
Software Version : CANOPY 7.3.6 Oct 24 2005 12:06:56
Software Boot Version : CANOPYBOOT 3.0
FPGA Version : 070605 (DES Sched) P9
Session Timeout: 0, AirDelay 135 (approximately 3.76 miles (19845 feet))
Session Count: 46, Reg Count 60, Re-Reg Count 56
RSSI (Avg/Last): 1223/1219 Jitter (Avg/Last): 4/5 Power Level (Avg/Last): -82/-82
DnRate(A): 6978 DnLimit(A): 500000 UpRate(A): 22 UpLimit(A): 100000 (kbit)
LoUpCIR(D): 0 LoDnCIR(D): 0 (kbps)
Rate : VC 36 Rate 2X/1X

this location has 2 canopy 900 aps on 120 sectors (thanks for the suggestion 2 years ago bill the FIRST time we had this problem!)

the 2nd sector, on 906, points mostly north and slightly west. it has no problems to speak of.

i’m posting a 2nd map on our website that will show all customers on this particular panel so you can compare their locations to their signals and jitter values. it screams walmart! (in my opinions)

oh…and yes, you do see a marker for walmart distribution too - we have a huge one of those. i bet they’re rfided too!

there are other customers that should be affected according to their signal paths…that aren’t. how do ya explain that? :frowning:

it’s got to be signal levels…the stuff @ walmart is bleeding over to our tower…but not strong enough to cause folks under about -77 to drop. i pulled a spectrum the other night too - i’ll post that out there too (i didn’t see anything that really got my attention) - but our tower is nearly 5 miles from the walmart…although apparently if it’s LOS…that doesn’t really matter. and our tower is really high. (firetower on top of an already high hill)

initial specturm (spectrum-203) was run 4/23 at 2:22 am

i’m running a new one now (spectrum-current) … soon as i get it posted.
i’ve shut down our other AP at the site…

jerry…you wrote:

Ask him if there is a Tx power setting for his system. If it’s -39dB at 1/4 mile, it’s probably -30dB at the store which may be more than they need.

what do you estimate he’d be at 4 miles out? - assuming line of sight or close line of sight?

i see another uplink of ours is coming in roughly 90 degrees off of walmart at about -80 (our cold springs tower uplink) and it’s not suffering a problem…

i hate rf sometimes…

thank you for your help - and everyone else. i’ll shuddup and go to bed now :slight_smile:

When you do a link test, what is the up and down efficiency?

I’m going to guess Walmart. You need to co-ordinate with them.

I think I would go for separation in this situation. Ask Walmart to use 906. Put the AP facing away from Walmart on 915, and the AP facing toward Walmart on 924.

That should clear things up.

at 12 noon, here is what we’re looking at:

LUID: 006 : MAC: State: IN SESSION (Encrypt Active)
Site Name : Jerry’s Used Cars
Software Version : CANOPY 7.0.7 Mar 02 2005 15:06:58
Software Boot Version : CANOPYBOOT 3.0
FPGA Version : 021605 (DES Sched)
Session Timeout: 27, AirDelay 224 (approximately 6.24 miles (32928 feet))
Session Count: 16, Reg Count 20, Re-Reg Count 17
RSSI (Avg/Last): 1220/1227 Jitter (Avg/Last): 7/12 Power Level (Avg/Last): -81/-81
DnRate(A): 6978 DnLimit(A): 500000 UpRate(A): 22 UpLimit(A): 100000 (kbit)
LoUpCIR(D): 0 LoDnCIR(D): 0 (kbps)
Rate : VC 22 Rate 2X/1X

2 sec test:
Stats for LUID: 6 Test Duration: 2 Pkt Length: 1522
Downlink RATE: 2078208 bps
Uplink RATE: 147456 bps
Downlink Efficiency: 100 Percent
Max Downlink Index: 100
Actual Downlink Index: 100
Expected Frag Count: 8118
Actual Frag Count: 8118
Uplink Efficiency: 76 Percent
Max Uplink Index: 100
Actual Uplink Index: 76
Expected Frag Count: 576
Actual Frag Count: 751

Strandlund isn’t connected
Ellard isn’t connected

LUID: 021 : State: IN SESSION (Encrypt Active)
Site Name : Lynn Harris
Software Version : CANOPY 7.3.6 Oct 24 2005 12:06:56
Software Boot Version : CANOPYBOOT 3.0
FPGA Version : 070605 (DES Sched) P9
Session Timeout: 0, AirDelay 587 (approximately 16.34 miles (86289 feet))
Session Count: 11, Reg Count 33, Re-Reg Count 23
RSSI (Avg/Last): 1190/1160 Jitter (Avg/Last): 6/5 Power Level (Avg/Last): -82/-82
DnRate(A): 20000 DnLimit(A): 500000 UpRate(A): 64 UpLimit(A): 100000 (kbit)
LoUpCIR(D): 0 LoDnCIR(D): 0 (kbps)
Rate : VC 37 Rate 2X/1X

2 sec test:
Stats for LUID: 21 Test Duration: 2 Pkt Length: 1522
Downlink RATE: 2089472 bps
Uplink RATE: 147456 bps
Downlink Efficiency: 100 Percent
Max Downlink Index: 100
Actual Downlink Index: 100
Expected Frag Count: 8162
Actual Frag Count: 8162
Uplink Efficiency: 75 Percent
Max Uplink Index: 100
Actual Uplink Index: 75
Expected Frag Count: 576
Actual Frag Count: 758

being walmart 4.67 miles away or whatever i posted it was, would the noise floor be about where i think it is?
thanks for your help.

ellard is back…strandlund is still attempting to register

LUID: 014 : State: IN SESSION (Encrypt Active)
Site Name : Tony Ellard
Software Version : CANOPY 7.2.9 Jul 23 2005 01:49:03
Software Boot Version : CANOPYBOOT 3.0
FPGA Version : 070605 (DES Sched) P9
Session Timeout: 0, AirDelay 352 (approximately 9.80 miles (51744 feet))
Session Count: 34, Reg Count 100, Re-Reg Count 39
RSSI (Avg/Last): 1163/1172 Jitter (Avg/Last): 9/8 Power Level (Avg/Last): -82/-82
DnRate(A): 6978 DnLimit(A): 500000 UpRate(A): 22 UpLimit(A): 100000 (kbit)
LoUpCIR(D): 0 LoDnCIR(D): 0 (kbps)
Rate : VC 30 Rate 2X/1X

he won’t stay connected during a link test tho.

re-reg count, btw, reset at 2 am when i pulled the spectrum so this is about 10 hours worth of stats…

Switch them all to 1X.

great idea, i’ll move that whole tower to 1x.

It might help to go check the AP and antenna just to make sure there is not something physical going on.

Also, you have 2AP’s - how are they getting sync? If they are generating sync that will create problems.


If you are sure that the equipment and AP configs are right, and they are in sync then see if you can get Walmart to turn off the RFID for a few hours so you can see if your issues go away.

an update…yes, we have 2 aps, they are being synced via a sync pipe and a homemade sync cable. The real giveaway is this problem just started two weeks ago…and the good news is we just came back from walmart and walmart confirmed they installed the same rfid systems our neighbors in sylacauga alabama wrote about. They did this … two weeks ago.

I’m awaiting their next play…off to do other work now. Disabling 2x hasn’t helped a whole lot. :frowning: We’re now down to 15 where we should have 19.

Dont mean to butt in here Jerry.

I see several overlooked issues.

None of those links listed are acceptable to begin with. Anything above a -77 will become unstable especially the more the AP gets loaded. The link will fade and drop off causing a re-reg. With the links being borderline to begin with Walmart’s 900 probably increased the noise floor just enough to cause this. The solution is not just working with them on freq choice but you need to address those links.

On your path profile you show tx power to be 28dbm or .631W. This is incorrect the max tx power on Canopy 900 is 26dbm or .400W. You also show to have an antenna gain of 9dbi at both ends. What are you using? If you are only using 9dbi antenna’s MTI makes a 120 pannel rated at 12dbi for the AP that will work better and use integrated 900 sm’s rated at 12.5. Just remember 3db gain will double your power so by using 9dbi antennas you have cut the power of the radios in half.

Last thing I see you have radios using software 7.0.7, 7.2.9, and 7.3.6. Update the software so they are all on the same version especially that 7.0.7.

Good observations on the path analysis and the SW versions.

However; before two weeks ago things were fine (or at least less bad). WalMart installed the system 2 weeks ago.

These RFID systems are going to become more and more of a problem increasing the noise floor.

Im not denying the cause of the problem being walmart but my point is that walmart would not be causing a problem if the links were not borderline to begin with. Jerry you and i both know and i have seen you wright it here 6 to 10 db above the noise floor -60 to -77 anything other and your asking for trouble. Today it is walmart tomorrow it might be me with these new 900mw mini pci cards i just got in from Ubiquity. HA! HA! just kidding. and not about the cards.

especially for the 9 and 13 mile links - they should never have been made to begin with - we should have had a closer tower on that side of the county. however, they’re there, so we’re trying to work with them. we have a newer tower closer to the 13 mile link but it’s 40 degrees off - and that particular antenna is on the top of a 40 foot pole. it’d be easier to get walmart to work with us (since they have already our neighbors) than try to adjust that antenna (the pole has been damaged before too).

jerry is right, there were no problems with any of these links until two weeks ago - they’d been running (both of them) since july 2005.

13 miles is fine just use a higher gain antenna. I have several at 20 miles in the - 60 to -70 range. But you guys are missing the point. Regardless what walmart does the links are unacceptable. getting them to work with you is just a temporary solution that will come back again. Every day new wireless devices come out. The noise floor will contenue to rise. The only right solution and fix for the problem is what I have suggested.

ok, aside from going higher (which we’ve already done…gone 40 feet), what else can be done to improve a signal?

Attitude, I agree the links are marginal. Not every link will fall into the ideal range. In this case the links have been running for a long time with no issues. I have links that have been running toward the lower end without issues.


Take a look at this link - http://motorola.canopywireless.com/supp … sc&start=0

-39dB at 1/2 mile is INSANE - no amount of antenna gain is going to fix it. Using the inverse square law this means that even at 4 miles the levels are around -57dB.

RFID systems are a different beast - may as well be another 900MHz WISP operating for all the power they are pumping out. Freq co-ordination is the only solution.


Jay, you asked about how to increase the gain on a link. In any RF system, you have five parameters:

1. Link distance
2. Tx Power
3. Rx Sensitivity
4. Tx Antenna Gain
5. Rx Antenna Gain

Assuming you can’t do anything about 1, 2, or 3 your only choices are 4, and 5.

As Attitude mentioned, you could use 12dB sector antennas for your AP’s. This will increase the overall Tx power from the AP as well as increase the AP’s ability to hear a weaker SM.

For SM’s, go connectorized and then use a selection of antennas to fit the need. For closer in SM’s, you can use the 8dBi patch antennas (I don’t really like them) or 9, 13, and 18dB yagi antennas.

We use the 9dB yagis for less than 8 miles, 13dB from 8 to 16 miles, and 18dB for 16+ miles. This is not a hard and fast rule though, sometimes a higher gain antenna is needed to penetrate foliage or deal with a high noise sector, or perhaps the SM is a bit more off-axis than we’d like.

hope this helps.

Jerry the RFID system is why Freq co-ordination is just a temp fix and why better links are the only long term solution. For the most part RFID systems will replace a lot of bar code reader systems for major retailers. They are also being used for such things as pre and post school buss inspection places at several points on the buss to varify if the driver does the inspection. The noise floor is only going to increase as more systems get put in place and we only have 3 freq to use in 900 on a 6 ap site. I do agree that the links might have been stable before walmart I do not see any numbers from before. They even want to put these things on the kids to track them getting on and off the buss at the proper stop. Now a -39 is extremely excessive for & would like to know if it is at a legal level. Im sorry I am just not one for having to go back down the road to fix again.

A Little info Walmart awarded two contracts for RFID readers to Alien Technologies, and Impinj for 15,000 fixed readers and Albertsons awarded a contract for 5000 units to Symbol who Motorola just aquired. The units are 902 to 928Mhz Frequency Hopping Spread Spectrum with a output of 30dbm. Using a lefthand and righthand circular polarization antenna.

Sorry but I dont think freq co-ordination is going to solve this one.
I do respect your openion though jerry.

A couple of things on Walmart.

Signal strength was not affected only Jitter.

All customers affected showed terrible downlink stats such as 20% on link tests during interference. Uplink was 100%.

Looking at your link stats, none of those customers would stay up in our area with -77 to -82 db’s with or without the interference.

We do not install anyone with RSSI’s of 1100 and Db’s in the -77 or higher range. The re-regs would be terrible.

We try to stay above the RSSI range of 1300 with Db’s in the low -70’s.

The customers we lost from the Walmart interference all were in the RSSI range of 1600-2000 with -59 to -71.

I would install strong external antennae on those SM’s ASAP.

We use the 17 db gain from www.M2inc.com

Generally we see a signal increse of -10 db’s and 100-150 RSSI gain when compared to the integrated SM’s

Good Luck

Reay Culp