N.I. Ignatik - Software Engineer, NPK BIOMEDIS (Moscow) S.V. Solovyev - Gr.Ph.D. in information technology, CEO NPK BIOMEDIS (Moscow) V.D. Tsygankov - Ph.D. (Eng.), Corresponding Member of the IIA, Chief Science of the SCE BIOMEDIS (Moscow) E-mail: embrion10@list.ru

One of the new and promising avenue of creating effective systems and encryption devices is the use for this purpose neural network approach and neuro-computers. Encryption using neurocomputer \"EMBRYO\" refers to high-speed and parallel processing techniques of digital information. Possible only actual scope neurocomputing encoder implemented in the form of a miniature neurochip can be encrypted command, telemetry and all kinds of graphic information coming via channels of communication with the ground, with un-derwater drones, etc. Unlike traditional artificial neural networks (ANN), we offer the option of the \"EMBRYO\" virtual neurocomputer for practical application for a reliable high-speed encryption and decryption of valuable information. Virtual neurocomputer (NC) of the \"EMBRYO\" may be applied as a \"one-time pad\". Neurocomputer is implemented in hardware in the form neurochip the FPGA company \"ALTERA\" or software heteromorphic (heterogeneous morphology, scheme-diverse) heterochrony (synchronous and asynchronous parallel streams of different scale at different times and data signals) neurophysiologically brain-like structure. Neurocomputer Is a device or a universal transmitter of information. In the input sensor array from sensors receptors of any type data or information comes to be a universal treatment. In particular, in the present case, for encryption and decryption. Associative neural network is shown in binary words of H. In the block MN is minimized to a simple network of motor neurons of the executive, represented by digital codes at the group level Y. In such a network, the number of motor neurons in a layer no more than MN = n + 1. The energy potential of the neural network is implemented in the form of a working pulse generator that defines by this number of lifetime or the lifetime of the impulse activity of neurons in the network, therefore, the accuracy and efficiency of information processing.

 

1. Gridin V.N., Solodovnikov V.I, Evdokimov I.A. Primenenie nejjrosetevogo podkhoda na osnove LVQ-seti dlja shifrovanija tekstovojj informacii // Sistemy vysokojj dostupnosti. 2011. № 1. T. 7. S. 65-68.
2. Medvedev V.S. Nejjronnye seti. MATLAB 6.0. M.: Dialog MIFI. 2007.
3. Cygankov V.D. Virtualnyjj nejjrokompjuter «EHMBRION». M.: SINTEG. 2005.