Computer Science, № 25, 2006

A class of asynchronous cellular automata (ACA) whose evolution simulates physicochemical kinetics of nano-systems is defined. It is characterized by multicell probabilistic transition rules and stochastic character of their application. To simulate real processes in real time ACAs should have huge cellular arrays and a very long evolution. So, the...

In this paper, we consider Computational Grid as market of the two commodities: CPU time and disc storage. Market agents are suppliers and users (consumers). The suggested market model refines the model proposed by R. Wolski et al. For the refined model, we discuss the possible application of the priceadjustment...

The Lattice Gas Cellular Automata (LGCA) models can be considered as an alternative to the conventional approach to the spatial dynamics simulation. The LGCA is based on a microscopic model of a physical process being simulated. Here we consider two simple LGCA models: HPP and HPPRP. They are based on...

Associative (content addressable) parallel processors of the SIMD type are ideally suited for performing fast parallel search operations being used in different applications such as graph theory, computational geometry, relational database processing, image processing, and genome matching.

The Delaunay Tessellation is used to construct a cellular automata resembling model based on data obtained from molecular dynamics experiment. Particle coordinates at each time step of required length are tessellated to compose a system of vertices and edges, which are classified to produce states and transitions. These states and...

A neural network algorithm is proposed for automation of cytological diagnostics of follicular tumors of a thyroid gland by images of an intraoperative material. Experimental results show the algorithm efficiency for constructing a medical expert system.

In the simulation of a protoplanetary disc with a power law density profile, a disc instability is detected. The instability arises only with a power law profile and is affected by the power index. Thus, the impact of initial density profile is large within the employed numerical model.

An effective algorithm is proposed for a balancing computation load in a heterogeneous multicomputer system by the self-organizing neural Kohonen network. The algorithm takes into consideration different performances of the system processors.