OPTIMUM FLOW DISTRIBUTION IN THE NETWORK WITH ADAPTIVE DATA TRANSFER

Abstract

This paper considers a fundamentally new model of data network construction for servicing the flows given by the gravitational matrix with restrictions on the time of information transmission. The proposed network model differs from the known ones because it has an expanded capacity of node and channel resources, combining the lowest possible cost of channels and switching nodes. As part of the development of the proposed model and algorithm for allocation of flows in a full mesh network, the developed mathematical apparatus provides a high degree of reliability and survivability of the synthesized network as a whole. One of the used approaches to the solution of the linear programming problem is based on the choice of the target function, the type of which is determined by the consumer conditions of the synthesis of a particular data network. Within the framework of the article, it is established that the linear programming problem, for each specific case, has an admissible, practically realizable solution with the optimal choice of the target function without additional simplex transformations. An important obtained result of the research is the simplicity of flow control, which is in direct dependence on the strict ordering of the proposed structure, because it was possible to connect by analytical dependences the flows in the branches and the path data transfer flows. The obtained analytical results can be used as a basis for statistical algorithms of information flow control.