Method for estimating fingerprint parameters of communication radiation source based on improved intelligent particle filtering

Abstract

The present invention discloses a method for estimating fingerprint parameters of communication radiation source based on improved intelligent particle filtering, firstly constructing a full-link fingerprint mechanism model, simulating a complete physical process of a signal from a transmitter through a channel to a receiver by combining mathematical operators. Based on this model, establishing a state space equation, adopting an improved particle filter algorithm to perform parameter estimation. In a resampling stage, dividing particles into a large-weight particle set and a small-weight particle set according to weight sizes, performing a crossover operation and a mutation operation on small-weight particles, the mutation operation is implemented by introducing a random perturbation obeying normal distribution on randomly selected dimensions. Finally, estimating radiation source fingerprint parameters based on an optimized particle set. The present invention effectively improves accuracy and stability of parameter estimation.