CRLB analysis for cognitive IR-UWB with imperfect sensing

Abstract

Cognitive ultra wideband (UWB) technology has ability to sense the environment, and offers an adaptive system with low transmission power and high throughput for wireless communications. Impulse radio UWB (IR-UWB) supports flexible modulations, with short mono-pulses, in which channel estimation has a critical role. However, imperfect sensing in cognitive IR-UWB introduces the channel estimation errors associated to the interfering primary users. Two main schemes of channel parameters identification are known as pilot-aided and blind methods. In this paper, the effect of primary user interference is considered for channel estimation based on maximum-likelihood (ML) criterion. Cramer-Rao lower band (CRLB) analysis is derived, and performance analysis is simulated to compare both pilot-aided and blind methods in cognitive IR-UWB system. The results verify the better performance of the pilot-aided estimation, and also show the negative effect of increasing the number of primary users and paths on the performance. At SNR=10 dB, for pilot-aided method with 10 users and 3 paths, the CRLB square root of gain is and that of delay is less than blind method. Besides, when the number of primary users is 20, and the paths are 7, this gap is for gain, and for delay