V.V. Chibissov - Student, Moscow Institute of Physics and Technology

The Hopfield neural network is simulated with the help of active wireless network (AWN). AWN differs from simple wireless sensor networks by their nods, which contain not only sensors, but also active elements (actuators). The Hopfield network hardware model is built for a small number of chaotic transmitters, each integrating the evolution equations, transferring data trough the whole network and indicating its state with light diode. The realization is used for observing and investigating the neurons dynamic.

 

1. Dmitriev A.S., Efremova E.V., Lazarev V.A., Gerasimov M.JU. Sverkhshirokopolostnaja besprovodnaja samoorganizujushhajasja prjamokhaoticheskaja sensornaja set // Uspekhi sovremennojj radioehlektroniki. 2013. № 3.S. 19-30.
2. Special Issue on Wireless Sensor and Actuator Networks // IEEE Trans. on automatic control. 2011. V. 57.№ 10.
3. Roska T., Chua L. The CNN Universal Machine an Analog Array Computer // IEEE Trans. on Circuits and Systems. 1993. V. 40. P. 163-173.
4. Hopfield J.J. Neural networks and physical systems with emergent collective computational abilities // Proceedings of National Academy of Sciences. Biophysics. April 1982. V. 79. № 8.P. 2554-2558.
5. Hopfield J.J. Neural with graded response have collective computational properties like those of two-state neurons // Proceedings of National Academy of Sciences. Biophysics. May. 1984. V. 81. P. 3088-3092.
6. Hopfield J.J. Learning algorithms and probability distributions in feed-forward and feed-back networks // Proceedings ofNational Academy of Sciences. Biophysics. December 1987. V. 84. P. 8429-8433.
7. Hebb D.O. The Organization of Behavior. New York: Wiley,1949.