Often, customers will want a device that they can use in an enviornment where it moves around. As it does so it will work seemlessly with the various wifi networks in the area.
First, let us define what ‘roaming’ is. According to the wifi standard it’s defined by each vendor so it can be a little ambiguous. For this paper, we will define roaming as where the wifi device in question somehow senses when it’s getting out of range of it’s old access point or network and somehow determines the new one to switch to. This is done as the device travels. Typically there is some kind of heuristics to determine when to switch. Signal strength is the most obvious however other ways could be used as well. GPS could be another.
The WiMe’s way of determining when to switch access points (‘roam’) is limited. It will disassociate from it’s wifi network after it senses the signal strength drop below a certain threshold. Once this happens it will then start scanning for a network to associate with. It will choose one if it fits it’s programmed values (encryption, ssid, channel, etc.). If more than one is found with these values it may or may not select the one with the strongest signal. Also, the wime makes no effort to gracefully handle network connections (TCP, UDP, etc.). So if you’ve got a TCP connection to the wime when it switches, then that connection is broken. The WiMe will not automatically close it down and then reestablish it using the new IP scheme of the new access points. That would need to be handled by some other method, perhaps the underlying network or the application.