• State
  • Locked Locked
  • Replies 11 replies
  • Subscribers 40 subscribers
  • Views 1149 views
  • Users 0 members are here
Support feedback
Options
Options
Related

This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

wl1271 missing APs when scanning

Dennis O'Brien
Prodigy 170 points
Other Parts Discussed in Thread: AM3715, WL1271, TPS65950

We have a design based on the AM3715 and WL1271 module.  We have found that the scan results returned from the wl1271 do not always contain strong access points (APs) in the list.

Sometimes, the strong AP will be missing from scan results for over 5 minutes.

This appears to be happening in any environment we've tested.  We do see the problem get worse in systems where a single device is serving up multiple APs.

We have seen reports of similar wl1271 issues on e2e.  One report stated that the module may be not waiting long enough an a channel to see all the APs.  Given this, we've tried to adjust 'max dwell time' in the wl1271 driver.  We have shown that we can adjust the dwell time, but even a 1 second dwell time doesn't appear to correct the problem.

Is there anything we can do to improve our situation?

Thanks in advance...

Dennis O'Brien, EQware

  • 0
    TI__Intellectual 1710 points

    Hi Dennis,

    Can you give some information on the setup / platform you are using ? 

    What Driver / FW versions are you using? 

    Regards,

    Barak

  • 0 in reply to Barak Schlosser
    Prodigy 170 points

    Barak,

     Thanks for responding so quickly.

    The product is a handheld instrument being sold to the education industry for science courses.  See Vernier Software and Technology, LabQuest2.

    The platform is based on the AM37 EVM.  It includes the AM3715, TPS65950, and the WG7310 from Jorjin (a WL1273 module) connected to one of the OMAPs SDIO interfaces.

    We are using the wilink-6.1.0.0.149 package from TI.  It contains:

    TIWLAN: 3468.372157: CHIP VERSION... set 1273 chip top registers

    TIWLAN: 3468.378688: Working on a 1273 PG 2.0 board.

    TIWLAN: 3468.383418: Starting to process NVS...

    TIWLAN: 3468.387721: NVS found, EEPROM Image addr=0xcef6f200, EEPROM Len=0x0x1d4

    TIWLAN: 3468.395137: Chip ID is 0x4030111.

    TIWLAN: 3468.399287: FEM Type 1

    TIWLAN: 3468.402491: Starting to download firmware...

    TIWLAN: 3468.454798: Starting to download firmware...

    TIWLAN: 3468.471492: Starting to download firmware...

    TIWLAN: 3468.476435: Starting to download firmware...

    TIWLAN: 3468.485163: Starting to download firmware...

    TIWLAN: 3468.491938: Starting to download firmware...

    TIWLAN: 3468.497584: Finished downloading firmware.

    TIWLAN: 3468.502253: Firmware running.

    TIWLAN: 3468.525538:

    TIWLAN: 3468.527583: --------------------------------------------------------------------

    TIWLAN: 3468.535548: Driver Version  : WiLink_Driver_6.1.0.0.130

    TIWLAN: 3468.541346: Firmware Version: Rev 6.1.0.1.349

    TIWLAN: 3468.546260: Station ID      : 01-23-45-67-89-AB

    TIWLAN: 3468.551356: --------------------------------------------------------------------

     

    Again, thank you in advance.  Your support is very much appreciated.

     

    Dennis O'Brien, EQware

  • 0 in reply to Dennis O'Brien
    TI__Intellectual 1710 points

    Hi Dennis,

    When observing this problem, what is the number of SSIDs reported by the device? 

    Is the same APs missing all the time? 

    Regards,

    Barak 

  • 0 in reply to Barak Schlosser
    Prodigy 170 points

    Barak,

    First - remember that we are observiing this problem inside the driver, just after it recieves the scan results from the module.  We also see the negative affects with in the NetworkManager.

    Typically, we see from 9 to 12 SSIDs.  But at other times we'll see only a few, and at times only one or even zero.  The signal-to-noise for most APs seem good.  We are not really worried about stations with poor signal quality.  It is the missing "good signal-to-noise" stations, that is the problem - they should always be seen.

    We miss different APs - not always the same.  Again, I'm only talking about the "strong" stations...

    I should note that this problem appears to be worse in environments where newer type APs are supporting multiple SSIDs (some supporting 4 SSIDs).

    Thanks,

    Dennis O'Brien, EQware

     

  • 0 in reply to Dennis O'Brien
    TI__Intellectual 1710 points

    Hi Dennis,

    When using multiple SSIDs on the same AP - as long as the MAC address is different for each SSID, the device should not have any issue displaying these SSIDs.

    Can you confirm eash SSID has a unique MAC address? 

    Regards,

    Barak

  • 0 in reply to Barak Schlosser
    Prodigy 170 points

    Barak,

    Our logs show that the mac addrs are different for the different SSIDs in the same AP.

    Dennis

     

    btw - do I have to continue to put 'tags' on these communications...?

  • 0 in reply to Dennis O'Brien
    Prodigy 170 points

    Adding tags...

     

    See last post...

     

  • 0 in reply to Dennis O'Brien
    TI__Intellectual 1710 points

    Hi Dennis,

    Can you share the commands you are using to re-create this issue and the value for "dwell time"?

    Please confirm the STA is disconnected from any AP when running the test.

    Regards,

    Barak 

  • 0 in reply to Barak Schlosser
    Prodigy 170 points

    Barak,

    The NetworkManager on our system places the driver in periodic scan mode.  Most of our testing has been done with the default "max dwell time" - which we think is 60ms.  Although, it's sort of hard to tell - we're looking at SME_SCAN_MAX_DWELL_DEF.

    We have changed this to longer, 1000ms.  Using a wifi sniffer, we think we see that this did change the dwell time - but again it's hard to tell with a sniffer.

    Yes - the STA is disconnected during our tests.

    Dennis

  • 0 in reply to Dennis O'Brien
    TI__Intellectual 1710 points

    Hi Dennis,

    I will further investigate and update you.

    Regards,

    Barak

  • 0 in reply to Barak Schlosser
    TI__Intellectual 1710 points

    Hi Dennis,

    In order to investigate the problem you encounter, I suggest you make a packet capture of the scan operation in one channel and then compare it with the actual list of APs on the channel. You better do it on a channel with the APs with Multiple SSIDs (which you indicated showed more problems). it will help identify if there is a problem (i.e. Probe responses are received by the chip but are not forwarded to the driver) or there is something else going on. 

    Some basic instructions on Scan commands and analyzing scan captures can be found here:

    http://processors.wiki.ti.com/index.php/OMAP_Wireless_Connectivity_WLAN_Scan

    If you have a packet capture indicating a problem please share it - so I can further assist you. 

    Regards,

    Barak