BAYESIAN MONTE CARLO ESTIMATORS OF COMPETING TERMINALS IN AN AD HOC WIRELESS NETWORK

摘要

We treat the problem of estimating the number of competing terminalsrnin an IEEE 802.11 wireless network, based on the BayesianrnMonte Carlo signal processing techniques. It is assumed that thernnumber of users in the network evolves according to a Markovrnchain with unknown transition probability matrix and the estimationrnis based on the observed collision probabilities in the sharedrnwireless channel. The online estimator is developed, based onrnthe sequential Monte Carlo (SMC) technique for jointly trackingrnthe system parameters and the number of users sharing the channel.rnComputer simulation results are provided to demonstrate thatrnthe Monte Carlo-based estimator significantly outperforms the extendedrnKalman filter (EKF)-based estimator.