 | 1 | initial version |
Packet loss is one of the significant issues with 802.11 wifi; especially in crowded networks with heavy RF usage.
Signal strength - you want to be 'close' to the devices, but not too close... maybe 1-2 meters away. One way to analyze the data is - can you correlate the loss? In a 4-way EAPOL handshake for WPA (or are you really using WEP? If so, please migrate to a secure solution!) you get two frames in each direction. The correlation part is - are they usually the client frames, or the APs, that are lost? Then look at signal strength and try to get an RSSI of -40 to -50 for both of the devices.
As well, you could have an issue with your capture mechanism not able to collect certain frames due to modulation differences. Based on the description, doesn't sound as if that is the issue but it depends on the specific frames you do get and the ones that are missing.
Passing USB devices like this into a VM to collect traffic, I find, to be very sensitive, especially with VirtualBox. It does work sometimes, but you might want to try a native install of Linux to have more resources available to the capture process. You might be able to boot from a USB on the PC so being native is usually possible.
