Recurrent Estimation of Information Streams Parameters on a Sliding Sample of Constant Volume

Abstract

Volume reduction of storing the data is one of the ways to increase the safety of information streams storage. Thus arising problem of formal representation of the data flow is shown in the form of a problem of identification of a time row with the non-stationary parameters which estimation should be carried out on a sliding sample of constant volume. In this work the recurrent method of a time row non-stationary parameters estimation is offered which, on the one hand, allows to carry out parameters estimation procedure in an «on-line» mode, and on the other hand, is essential to simplify computing procedures.

Keywords

• security storage of information
• non-stationary time series
• estimation of parameters of time series
• method of least squares