Jamming Detection on Block-fading Channels

We consider communication in the presence of hostile jamming and block-fading. The jamming signal is modeled by a Gaussian two-state hidden Markov model. Both the communication signal and the jamming signal experience fading that is constant over each packet transmission. Under this scenario, the performance is very sensitive to errors in detecting which symbols are jammed and in the estimates of the jamming and fading parameters. Thus, it is crucial for the receiver to be able to accurately determine both the set of jammed symbols and the parameters of the jamming and fading. In this paper, we present several techniques to perform this detection and estimation. These problems are made considerably more difficult because we assume that the none of the parameters of the jamming signal are known a priori to the receiver. Maximum a posteriori (MAP) estimation and generalized likelihood ratio tests (GLRT) are considered first. Then a lower complexity ad-hoc technique is developed that is based on estimating the density function of the received symbols. Simulation results are presented to compare the performance of the proposed detection and estimation techniques.