I had the exact scenario set up.
Netgear Router set up as an access point.
Mobile device connected to the access point.
BCM9WCDPLUS114 flashed with apsta connected to the access point.
I did not see the access point going down or the mobile device being disconnected.
I tested the application using WICED 3.4.1 SDK.
What I feel is that there may have been a IP address conflict in your case.
Can you verify the IP addresses?
Hope that helps.
Would you give more clear description what you meet ? And what you expect to do ?
There are no SDK V3.2.1 ! Or you could try APSTA of SDK 3.4.1 as AB Mail!
Why you have to connect both AP and STA interface on the same time ? In most of IoT application , it did not need bring up both interface!
If you expect the WM-N-BM-14 work like a router, maybe you should find a router solution to do such kind of job.
We have test AP interface and STA interface separate, both could work well and stable!
And remain you the WM-N-BM-14 is only one of single radio (RF), It mean the AP role and STA role are share with same Radio interface !
We are working on this issue and will update you once we have a patch for you to try. We are still debugging the root cause.
1.The mobile device should be connected to the AP of WICED board and not the router.
2.Bring up the STA interface of WICED board and connect to a router as a client.
3.As soon as this happens(both the interfaces of WICED should be up now),the mobile device seems to drop the connection with WICED's AP.
We are talking to your engineering management directly for this issue. Please work with them for a resolution. Once a clear resolution is available then we will communicate that to the community. Since your management has requested direct BRCM support and not via community at this point we will take this discussion offline and discuss.
The issue seems to be related to a common ARP cache for both the interfaces in SDK 3.1.2. I tried with a link local IP range(169.254.1.0 through 169.254.254.255) on AP interface and did not notice UDP packet miss now. My guess is that the AP interface ARP cache entry is being dynamically removed as soon as a new interface is started with a common IP range.
I am facing the exact same issue. I am using SDK 3.3.1. Tried with a different IP range as well as you said but the issue still seems to persist. It is reproducible every time but only with certain routers/Access Points. This fact is in sync with a comment above that asks to replace the router.
We have tried to isolate the issue. It seems that when the IOCTL command is sent to the Network processor (WLC_SET_SSID), the network processor connects to the AP and any client connected to our device is forcibly disconnected.
In our case, this is particularly a problem because we need a client to be connected to the WICED device. The client sends WiFi credentials and we have to respond whether or not we could connect successfully. This success/fail communication is not possible because the client has been disconnected.
If you use the "wiced_wifi_enable_powersave_with_throughput ", do not use it.