Adaptive data verification for resource-constrained systems
A system and method for adaptively verifying data in resource constrain systems. The adaptive data verification mechanism employs the proper mode of verification adaptively to balance cost/performance requirements plus security requirements. The algorithm uses a belief level for the validity of a received message, and assigns the belief level to a scale between a bona fide message at one end of the scale and a malicious message at an opposite end of the scale. Depending where on the scale the belief level falls determines which validation mode will be used to authenticate the message. In an alternate embodiment, the belief level relative to a scale and the amount of data waiting to be processed in a buffer are both used to determine which mode will be used to validate the message.
1. A method for authenticating messages received at a receiver device, said method comprising:
establishing a belief level for senders of messages that are received at the receiver device as to the messages validity;
assigning the belief level to a scale where one end of the scale indicates a bona fide sender or message that will be accepted and the other end of the scale indicates a malicious sender or message that will be discarded;
selecting an authentication mode based at least in part on the belief level, wherein one authentication mode includes a blind accept mode for a belief level at or around the bona fide end of the scale, a discard mode for a belief level at or around the malicious end of the scale and at least one mode therebetween; and
performing an authentication process on the message depending on which mode is selected, wherein the at least one mode between the blind accept mode and the discard mode includes modes for validating a PKI signature, validating a lightweight signature and/or performing a consistency check based on local information, and wherein the PKI signature mode is selected if the belief level is near the malicious end of the scale and the lightweight signature mode is selected if the belief level is near the bona fide end of the scale.
2. The method according to claim 1 wherein if the blind accept mode or the discard mode are not available as modes of verification, other verification modes are employed and ranked according to their strength where a weakest mode may be selected at or around the bona fide end of the scale and a strongest mode may be selected at or around the malicious end of the scale.
3. The method according to claim 1 wherein one of the at least one modes is a consistency check mode that determines the authentication of the message based on local information around the receiver device.
4. The method according to claim 3 wherein sensors, GPS data or physics-based models are used to determine the local information.
5. The method according to claim 1 further comprising determining how much unprocessed data is present in a buffer at the receiver device to determine which of the at least one modes will be selected in combination with the belief level, wherein more unprocessed data in conjunction with a belief level near the malicious end of the scale moves the mode towards the discard mode and less unprocessed data in conjunction with a belief level near the bona fide end of the scale moves the mode towards the blind accept mode.
6. The method according to claim 1 wherein the receiver device is onboard a receiving vehicle and the message is transmitted from another vehicle.
7. The method according to claim 1 wherein the modes between the blind accept mode and the discard mode include a mode for performing only a cryptography check, a mode for performing only a consistency check and a mode for performing both a cryptography check and a consistency check.
8. The method according to claim 1 wherein the at least one mode between the discard mode and the blind accept mode is a plurality of different modes that are used to authenticate the message and have different cost and performance levels, where the cost and performance level increases as the mode moves along the scale towards the malicious end.
9. The method according to claim 8 wherein each mode corresponds to a range of belief levels along an interval of the scale.
10. A method for authenticating messages received at a receiver device, said method comprising:
establishing a belief level for senders of messages that are received at the receiver device as to the messages validity;
assigning the belief level to a scale where one end of the scale indicates a bona fide sender or message that will be accepted and the other end of the scale indicates a malicious sender or message that will be discarded;
selecting an authentication mode based at least in part on the belief level, wherein one authentication mode includes a blind accept mode for a belief level at or around the bona fide end of the scale, a discard mode for a belief level at or around the malicious end of the scale and at least one mode therebetween; and
performing an authentication process on the message depending on which mode is selected, wherein one of the at least one modes is a consistency check mode that determines the authentication of the message based on local information around the receiver device, and wherein sensors, GPS data or physics-based models are used to determine the local information.