__Keywords:__immunity of wireless communication systems adaptive signal processing cognitive radio aprioristic uncertainty non-gaussian noise estimation of parameters of interference

E.N. Efimov, A.M. Alhamad

Questions of increase of interference immunity of wireless communication systems on a background non-Gaussian interferences in conditions of a priori uncertainty have been considered.
The purpose of this given work is development of adaptive algorithm of detection-distinction of signals of the known form in a presence of non-Gaussian interferences presented by poly-Gaussian likelihood models, at absence of the a priori information on statistic behaviors of distribution of interference in radio channel.
On the basis of the adaptive Bayesian approach, poly-Gaussian methods and methods of an estimation of parameters mixture of distributions the generalized algorithm of detection-distinction of signals in a presence of non-Gaussian interferences in conditions a priori uncertainty concerning parameters interferences have been synthesized. The synthesized polycorrelative algorithm of signal processing combines the basic properties, characteristic for poly-Gaussian algorithms, in particular: structurally the algorithm consists of the same parallel channels, in each of channels operations characteristic for correlative algorithms are fulfilled, to this proviso the basic operation this formation of correlation integral. Irrespective of a degree of complexity of allocations of interferences, the algorithm contains a finite set of standard operations and has recurrent form. The estimation of parameters non-Gaussian interference distribution is carried out by results of an average of statistics stored on the previous steps of signals processing and formed within the limits of the basic trunk of algorithm of distinction in which the decision on useful signals is developed, thus the adapting trunk is naturally built in structure of algorithm.
The comparative analysis of the received algorithm by a method of computer modeling have been realized.
Results of modeling synthesize algorithm essential gain in interference immunity (on probability of an error) in relation to correlative algorithm in conditions of influence on a useful signal of non-Gaussian interferences and practically identical interference immunity (on probability of an error) in comparison with policorrelation algorithm, with known parameters a interferences of radio channel. Estimation parameters interferences, made in format of the basic procedure of detection - distinctions of signals, satisfies to criterion Kolmogorov even at a rigid significance value, what confirm a high efficiency works of the blocks of an estimation of parameters interferences.
The carried out analysis has confirmed what synthesized algorithm allow to raise efficiency system of signal processing and allow achieve maximum possible improvement interference immunity of wireless communication systems in conditions a priori uncertainty in a presence of non-Gaussian interferences.

References: