USING EXCESS CODE TO DESIGN FAULT-TOLERANT TOPOLOGIES

and in modernizing an already existing one. Particular attention is paid to it when building multicomputer systems or clusters. The most interesting ways to increase fault tolerance is to use the topological structure of the system to bypass the malfunction and use one or another element of the system to replace the faulty. Of course, this requires the development of a specific topology. The article deals with the development of fault-tolerant versions of popular topologies, such as quasiquantum and hypercube, based on the excess code 0/1 /-1. Target setting. An important part of any multicomputer system or cluster is its topological structure. This structure defines the routing of messages in the system, speed of message transmission, and level of fault-tolerance of a system. The article proposes a method for increasing fault-tolerance based on the use of excess code. Actual scientific researches and issues analysis. Synthesis of topologies such as the hypercube or the de Bruin topology is well studied and described now, there are papers consider methods for increasing the fault-tolerance that based on usage of additional nodes that duplicate current nodes. Other papers consider using a tree-based routing to improve fault-tolerance of the system. Uninvestigated parts of general matters defining. Now the possibilities of using excess code 0/1/-1 for creating new fault-tolerant topologies based on existing synthesis methods are unconsidered. The research objective. The task is to describe the synthesis of fault tolerant topologies, the consideration of the possibilities of using their features and the analysis of the main characteristics in comparison with each other and with classic versions based on binary code. The statement of basic materials. The synthesis of hypercube and de Bruin topology is described on the basis of the usual binary code and the redundant code 0/1 / -1, the possibilities of using redundancy are considered, first of all, to increase the fault-tolerance, a comparative analysis of all of these topologies is carried out. Conclusions. The analysis of characteristics is performed, the main advantages and disadvantages of the proposed topological structures are highlighted, suggestions for their improvement are made.

Data bus modelling in city buses

The paper characterizes ICT networks used in urban buses. It describes the multiplex system manufactured by Continental VOD and the ELSY multicomputer system manufactured by Thoreb, Differences in the structure and functioning of both systems were presented. Based on the Thoreb system, a test station for the modelling of a complex data bus used in city buses has been designed and built.

THE CONGESTION OF GENERALIZED CUBE COMMUNICATION PATTERN IN LINEAR ARRAY NETWORK

Generalized cubes are a subclass of hypercube-like networks that are employed by a number of parallel algorithms. Linear array is common topology of multicomputer system. So congestion is very useful to improve algorithm performance for implementing generalized cube communication pattern on linear array. This paper addresses the congestion of generalized cube communication patterns embedding into linear array topology. For this purpose, the maximum m-induced subgraph of generalized cube is determined firstly, which is very important for discussing the congestion. Then an embedding scheme is described, and the congestion is shown to attain the minimum, which guarantees the optimality of the proposed scheme.