Performance analysis of a land mobile satellite system using importance sampling

Land mobile communication using satellites has become widespread. However, estimation of its average bit-error rate is still challenging due to the complexity of various system components, such as the nonlinearity of traveling wave tube amplifiers (TWTAs), intersymbol interference, spread spectrum under fading, or jamming conditions. The analytical approach using closed form solutions is very difficult if not impossible. Monte Carlo (MC) simulation can offer an alternative method of performance estimation; however, the use of MC simulation is often limited by an excessive computational burden. Importance sampling (IS) is an efficient simulation method that can greatly reduce this computational overhead although it requires additional modeling and a careful biasing scheme. We propose an integrated IS model which combines error event simulation, a conditional IS method, and an asymptotically efficient method for the effective estimation of the error rate. Simulation results conditioned on different channel parameters such as the backoff of a TWTA, jamming signal strength, and elevation angle of the low-Earth orbit satellite are provided. Comparisons between the proposed method and the MC method show that a great reduction in simulation time, as well as increased accuracy, can be achieved with the IS method.