Joint Data and Tag Precoder Optimization for MIMO Physical Layer Authentication with Embedded Fingerprinting

We consider precoder optimization for authentication in a wireless multiple-input-multiple-output (MIMO)system that embeds a fingerprint at the physical layer. The optimization relaxes the assumptions that the same precoding matrix will be used for both data and tag and that the intended receiver (Bob)is able to perfectly estimate the data. The sum mean-square error (MSE)estimate of the tag is derived as a function of data and tag precoder matrices, which is then minimized under individual data and tag power constraints. Simulation results show that the proposed precoder optimization provides much better authentication probability than the equal power allocation. Moreover, by tuning the tag power, the desired tradeoff between authentication probability and key leakage (as measured by key equivocation at an eavesdropper)can be achieved with the proposed method.