They are similar in that they both use / permit key rotation and (in the case of WPA-PSK) both start with a "seed key" and ultimately generate dymanic keys.
WPA2 / 802.11i use AES for encryption, WEP with TKIP still uses RC4. RC4 was never intended for use with this type of encryption application. The original problem with WEP, what makes it weak by comparison (they way it was implemented for wireless), is that the initialization vectors are static.
The system was utilized such that it became easy to predict; by feeding it some known values and watching for an expected result, the encryption could be "figured out."
AES is much more recent, and does not have these weaknesses. WPA using TKIP does not have have these weaknesses.
The main weakness for WPA, specifically the PSK flavor, is if a weak key/passphrase is used (ones with "dictionary" words and/or short length).
WEP with TKIP is much more secure than plain old WEP, but it still has a few exploitations that do not exist with WPA with TKIP.
The primary difference is, I believe, how the encryption is initialized.
Good Luck
Scott
