CASPIAN JOURNAL

MANAGEMENT AND HIGH TECHNOLOGIES

Read

Prikhodko Maxim Alexandrovich  // Caspian journal : management and high technologies. — 2012. — №1. — pp. 47-52.

Prikhodko Maxim Alexandrovich - Candidate of Physical and Mathematical sciences, Moscow State Mining University, 119991, Russia, Moscow, Leninsky avenue 6, spex19@mail.ru

This work considers the problem of optimal data flow redirection in distributed multi-agent heterogeneous information processing systems. It is shown that optimality of data flow redirection is determined by available nodes capacity and quality of communication channels. It is proposed to choose nodes for data flow redirection on the basis of the weights equal to ratio of available capacity to the speed of data transfer through the corresponding communication channel. The nodes with the minimal weights are chosen so that their total capacity is not less than the redistributed surplus. Further redistribution is made if there are more available modes or the last node with the largest weight among the chosen is not fully used. Algorithm of data flow redirection is described.

Key words: information processing,information processing system,multi-agent system,distributed system,agent,counter-agent,intellectual agent,rival agents,data flow,redirection,heterogeneous information,optimization,node,heterogeneous information