Return to Revisions

3 of 3

Improve readability and fix typos

edit approved Apr 25, 2020 at 21:43

If two hosts are using a transparent IP level authentication and encryption, is their communication subject to both MITM or replay attacks?

Encryption process for any outgoing packet to the specified "remote IP":

-original payload P is encrypted (the bytes after the IP header) using RC4 with the preestablished key KEY to obtain encrypted payload C=RC4(P, KEY)

-The encrypted payload C is concatenated with the 16 byte output of MD5(C|KEY), and the IP header is adjusted accordingly (increasing the packet total length, re-calculate the checksum) so that the outgoing packet will have payload C | MD5(C|KEY) and destination "remote IP"

Decryption process for any incoming packet from specified "remote IP":

-Check the packet payload for the MD5 authentication with the same given "key" Specifically, divide the received packet payload (those bytes after the IP header) as two parts C, A where A is the last 16 bytes of the packet payload and C is the rest.

-If MD5(C|KEY) == A, decrypt C using RC4 to obtain P=RC4(C, KEY), change the payload of the incoming packet from C|A to P, and adjust the IP header accordingly (decrease the packet total length, re-calculate the checksum) so that the incoming packet will be restored to its original form without keyed md5 authentication before sent to the receiving process

asked Apr 25, 2020 at 20:29

Koolz