Accurate Performance Analysis of TR UWB Systems With Arbitrary Front-end Filters

The sampling expansion approach for computing the performance of TR UWB systems has restricted validity and is very inaccurate in some cases. A new analytical method to evaluate the BER performance of a TR UWB system is derived. The method is based on expanding the received signal using the Karhunen-Loeve expansion and then evaluating the characteristic function of the receiver decision metric. This new method puts no restriction on the front-end filter type and is accurate for any arbitrary value of receiver time-bandwidth product. The analytical results are validated with simulation results, and it is found that the method is highly accurate and can evaluate BER rates down to 10^-9.