Jamming and counter-measure strategies in parallel Gaussian fading channels with channel state information

Consider a parallel Gaussian channel with M subchannels each of which experiences independent flat fading. There exists a smart jammer which is capable of intercepting the feedback channel message such that channel state information (CSI) measured and sent back by a receiver can be perfectly known to the jammer, as well. We formulate a novel two-layer two-person zero-sum game. Under this zero-sum game framework, the following three fundamental questions are addressed. The first one is about whether to hop or to spread power over M frequency bands given the full CSI at both transmitter and jammer sides. We prove that spreading versus spreading is the Nash equilibrium point. The second question is about the impact of sending back CSI on the overall throughput, to feedback or not to feedback, considering the presence of a smart jammer capable to exploit CSI for its malicious purpose. We show that possessing the full CSI enables a better means to defend against the jammerpsilas attacks. The last question is about whether the amount of feedback information can be reduced given the mutual restrictions between transmitter and jammer. We prove that the receiver should also get involved by adaptively sending CSI without compromising the sum rate at the equilibrium point.