how to reduce the scan duration after orphan scan failed？

The Seven

Hi all

i have a ZED switch which power supplied by battery。

when it being orphan， i want to make it go into sleep as soon as possible。。。when i don't want to changing the zgDefaultRejoinScan 。

so i changing the ZDO_StartDevice（） function as like below:

    else if ( startMode == MODE_RESUME )
    {
      if ( logicalType == NODETYPE_ROUTER )
      {//路由设备
        ZMacScanCnf_t scanCnf;
        ZDApp_ChangeState( DEV_NWK_ORPHAN );

        /* if router and nvram is available, fake successful orphan scan */
        scanCnf.hdr.Status = ZSUCCESS;
        scanCnf.ScanType = ZMAC_ORPHAN_SCAN;
        scanCnf.UnscannedChannels = 0;
        scanCnf.ResultListSize = 0;
        nwk_ScanJoiningOrphan(&scanCnf);

        ret = ZSuccess;
      }
      else
      {//终端设备
        ZDApp_ChangeState( DEV_NWK_ORPHAN );
        //set timer for scan and rejoin
      #ifdef SWITCH_BAT
        //电池按键 orphan 1s 后立刻进入休眠态
        osal_start_timerEx( ZDAppTaskID, ZDO_REJOIN_BACKOFF, 1000);
      #else
        osal_start_timerEx( ZDAppTaskID, ZDO_REJOIN_BACKOFF, zgDefaultRejoinScan );
      #endif  
        ret = NLME_OrphanJoinRequest( runtimeChannel,
                                     zgDefaultStartingScanDuration);
      }

when i found that sometimes, the ZED swicth could not rejoin ~~~ it just sending beacon request, and could not rejoin !

why ?

BR!

over 8 years ago

0 YiKai Chen over 8 years ago

Guru 735695 points

Do you use Ubiqua Packet Analyzer to check this issue?

0 The Seven over 8 years ago in reply to YiKai Chen

Guru 11440 points

it is OK ,it is my mistake!

we should pay more attention of the ZDOInitDevice（）！

// Hold down the SW_BYPASS_NV key (defined in OnBoard.h)
// while booting to skip past NV Restore.
//if ( zdappHoldKeys == SW_BYPASS_NV )
//{
zdappHoldKeys = 0; // Only once
//networkStateNV = ZDO_INITDEV_NEW_NETWORK_STATE;
//}
//else
//{
// Determine if NV should be restored
networkStateNV = ZDApp_ReadNetworkRestoreState();
//}

because there is a keypress judge！

BR!

0 YiKai Chen over 8 years ago in reply to The Seven

Guru 735695 points

It's good to know you find where the problem is.

0 The Seven over 8 years ago in reply to YiKai Chen

Guru 11440 points

No it is just one cause！

the issue have not been solved！

i found when the ZED give the MAC ACK to other ZC ， it could not rejoin its parents anymore while it parents reply it a right beacon ！

001248000cB8039B is not the parents of ZED switch！ why does it send a coordinator realigment to ZED switch ？？

there is a log ！！！

3513.1.rar

0 YiKai Chen over 8 years ago in reply to The Seven

Guru 735695 points

Packet #42 (Device Announce) shows this device joins back, doesn't it?

0 The Seven over 8 years ago in reply to YiKai Chen

Guru 11440 points

yes ， that is because i reset the device， and the device reply the MAC ACK to the right one ， so it could rejoin success！

if we do not reset it ， it could not rejoin although there is a lot of right beacon ！！！！

0 YiKai Chen over 8 years ago in reply to The Seven

Guru 735695 points

Can you specify between which packets that you do reset to a device?

0 The Seven over 8 years ago in reply to YiKai Chen

Guru 11440 points

hi yikai ！

the second orphan notification is happen when i reset the device！

this situation is like that ZED had join so many ZC， and leave without tell those ZC。

so when it sending the orphan notification ，those ZC will all reply it a coordinator realignment ！

i don't know what is ZED depend on to reply a MAC ACK to which one ZC ！（the best result is that ZED could know which is right for it ） but it seems that ZED choise randomly！

when reply the old one ， it could not rejoin any more！

BR!

0 YiKai Chen over 8 years ago in reply to The Seven

Guru 735695 points

I suggest you to set a breakpoint in ZDApp_NwkDescListProcessing to see why it doesn't send association request after receiving beacon frame.

0 The Seven over 8 years ago in reply to YiKai Chen

Guru 11440 points

it returns pNwkDesc！ but i have not seem any more associattion request ~~~~~~

if ( i == ResultCount )
{
nwk_desc_list_free();
return (NULL); // couldn't find appropriate PAN to join !
}
else
{
return (pNwkDesc);
}

0 YiKai Chen over 8 years ago in reply to The Seven

Guru 735695 points

You should keep tracking into "case ZDO_NWK_DISC_CNF:" in ZDApp_ProcessOSALMsg.

0 The Seven over 8 years ago in reply to YiKai Chen

Guru 11440 points

it run into the continueJoining but i have not seen any association request

if ( ZG_BUILD_JOINING_TYPE && ZG_DEVICE_JOINING_TYPE )
{
networkDesc_t *pChosenNwk;
if ( ( (pChosenNwk = ZDApp_NwkDescListProcessing()) != NULL ) && (zdoDiscCounter > NUM_DISC_ATTEMPTS) )
{
}
else
{
if ( continueJoining )
{
......
}
}
}

0 YiKai Chen over 8 years ago in reply to The Seven

Guru 735695 points

Do you mean it calls to ZDP_NetworkInit?

0 The Seven over 8 years ago in reply to YiKai Chen

Guru 11440 points

yes ！

it calls the ZDApp_NetworkInit() , and then go to

  if ( events & ZDO_NETWORK_INIT )
  {
    // Initialize apps and start the network
    ZDApp_ChangeState( DEV_INIT );

    ZDO_StartDevice( (uint8)ZDO_Config_Node_Descriptor.LogicalType, devStartMode,
                     DEFAULT_BEACON_ORDER, DEFAULT_SUPERFRAME_ORDER );

    // Return unprocessed events
    return (events ^ ZDO_NETWORK_INIT);
  }

when calls below function in ZDO_StartDevice(), it returns ret = FAILED

      if (startMode == MODE_REJOIN)
      {//
        ret = NLME_NetworkDiscoveryRequest( runtimeChannel, BEACON_ORDER_480_MSEC );
      }

BR!

0 YiKai Chen over 8 years ago in reply to The Seven

Guru 735695 points

There is no NLME_NetworkDiscoveryRequest source code available to do further tracking. What is runtimeChannel when this issue happens? Is it the channel where ZC is?

0 The Seven over 8 years ago in reply to YiKai Chen

Guru 11440 points

oh no ，i make a mistake ！

NLME_NetworkDiscoveryRequest returns ZSuccess， and it have not calls below function in the end
if ( ret != ZSuccess )
{
osal_start_timerEx(ZDAppTaskID, ZDO_NETWORK_INIT, NWK_RETRY_DELAY );
}

but i have not seen device send a rejoin any more ~~~

anyway， i always found that when it calls ZDApp_NwkDescListProcessing（）

ZDO_UseExtendedPANID is different with pNwkDesc->extendedPANID

0 The Seven over 8 years ago in reply to YiKai Chen

Guru 11440 points

now the problem is that when zstack run into the case ZDO_NWK_DISC_CNF，

if ( ( (pChosenNwk = ZDApp_NwkDescListProcessing()) != NULL ) && (zdoDiscCounter > NUM_DISC_ATTEMPTS) )

ZDApp_NwkDescListProcessing returns a pNwkDesc for rejoin， but zdoDiscCounter is 1 so make the device could not doing the below rejoin process！

if ( ( (pChosenNwk = ZDApp_NwkDescListProcessing()) != NULL ) && (zdoDiscCounter > NUM_DISC_ATTEMPTS) )
{
if ( devStartMode == MODE_JOIN )
{
}
else if ( devStartMode == MODE_REJOIN )
{
}
{

0 YiKai Chen over 8 years ago in reply to The Seven

Guru 735695 points

Try to set NUM_DISC_ATTEMPTS to 0.

0 The Seven over 8 years ago in reply to YiKai Chen

Guru 11440 points

it is OK, but why ? what is the NUM_DISC_ATTEMPTS and what is zdoDiscCounter?

0 YiKai Chen over 8 years ago in reply to The Seven

Guru 735695 points

Decreasing NUM_DISC_ATTEMPTS would make device sending less beacon requests for confirming join process.

0 The Seven over 8 years ago in reply to YiKai Chen

Guru 11440 points

set the NUM_DISC_ATTEMPTS as 0 or make sure device received beacon more than NUM_DISC_ATTEMPTS ！

0 YiKai Chen over 8 years ago in reply to The Seven

Guru 735695 points

Is it OK now?

0 The Seven over 8 years ago in reply to YiKai Chen

Guru 11440 points

yes ！ thanks ！

0 YiKai Chen over 8 years ago in reply to The Seven

Guru 735695 points

You are welcome and it's good to know it works.

Zigbee & Thread

Zigbee & Thread forum

how to reduce the scan duration after orphan scan failed？