INVESTIGATION OF PECULIARITIES OF RANGE ALGORITHMS AND DETERMINING OPPORTUNITIES FOR USE IN NGN NETWORKS

Abstract

The basic properties of ranks and rank statistics are analyzed. The property of invariance of ranks on the example of a harmonic signal with a random initial phase is considered. An example of signal ranking is considered. It is shown that the distribution of ranks is invariant to the type of initial distribution only if the elements of the ranked samples are statistically homogeneous. In all cases, the elements of the samples are statistically independent, i.e. their distribution does not depend on the element number. The nonlinear transformation characteristic corresponds to the integral impedance distribution function, and the ranking according to the algorithm is equivalent to the actual distribution of the sample, which, depending on the situation, may be a disturbance of the disturbance or mixture of the interference signal. The restrictions on the scope of rank procedures in comparison with optimal invariant transformations are analyzed. First of all, it is necessary that, in the absence of a signal, the ranked sample was statistically homogeneous. This ensures a stabilization of the level of false alarms with an a priori unknown distribution. But it is equally important that the appearance of the signal destabilizes the homogeneity of the sample. Despite the reduction in the number of informative readings, the sensitivity of the distribution of ranks in the presence of a signal not only did not decrease, but even increased, which is an essential principle for organizing the procedure for detecting a signal based on ranks. Flexibility of the ranking procedure provides the possibility of solving a wide range of signal detection tasks under conditions of nonparametric a priori uncertainty.

Authors and Affiliations

Olha Tkachenko

Keywords

Related Articles

Use of some mathematical regularities in processing radar information in the interest of prevention of emergency situations of terrorist character

The possibility of using some mathematical regularities in the processing of radar information is considered in the work, in order to shorten the time of identification of dangerous targets in order to prevent emergencie...

FEATURES OF ARCHITECTURAL AND FUNCTIONAL-STRUCTURAL DESIGN OF SYSTEMS AND NETWORKS OF EXTERNAL STORAGE AND DATA PROCESSING

The justification for the choice of formal and informal methods for describing the systems and networks of data storage and processing is given. The peculiarities of application of the considered methods in architectural...

Analysis of efficiency of signal filtering algorithm built on the basis of L-group operations.

Signal filtering (signal extraction) problem is formulated within signal space with L-group properties. We consider algorithm of narrowband (in radioengineering sense) signal filtering that is based on the method of lin...

ANALYSIS OF INTERACTION OF RISK DAMAGE IN THE CASE OF TECHNOGENIC ACCIDENTS IN THE CONCEPT OF ACCEPTABLE RISK

The methods of estimating the damage from the occurrence of an industrial accident considered. For the quantitative description of the consequences of an accident, the notion of damage is used, which is considered as an...

STRUCTURE OF RESOURCE SUPERVISING INFORMATION TECHNOLOGY

The feature of the resource supervising tasks is that a substantial part of the information on the resources has signs of imperfection. This is manifested in ambivalent data (information about the same object/process is...

Download PDF file
  • EP ID EP461109
  • DOI -
  • Views 91
  • Downloads 0

How To Cite

Olha Tkachenko (2018). INVESTIGATION OF PECULIARITIES OF RANGE ALGORITHMS AND DETERMINING OPPORTUNITIES FOR USE IN NGN NETWORKS. Телекомунікаційні та інформаційні технології, 141(1), 127-133. https://europub.co.uk/articles/-A-461109