Waveform design for sequential detection with subspace interference

This paper addresses the problem of adaptive waveform design for target detection using sequential hypothesis testing. In the proposed scheme, the transmit waveform is determined based on previous measurements. The sequential framework is especially adequate in cases of composite hypothesis testing, where the different hypotheses depend on unknown parameters. With each new observation, additional statistical information regarding the unknown parameters is obtained with decreasing uncertainty. We derive a sequential test based on the generalized likelihood ratio test (GLRT), which is applied for target detection in the presence of subspace interference, and propose a waveform design scheme based on the Kullback-Leibler divergence. It is shown by simulations that the proposed scheme requires fewer observations compared to non-adaptive waveform transmission in order to achieve desired error rates.