Using queuing theory to describe adaptive mathematical models of computing systems with resource virtualization and its verification using a virtual server with a configuration  similar to the configuration of a given model

Martyshkin, Alexey I.; Pashchenko, Dmitry V.; Trokoz, Dmitry A.; Sinev, M.P; Svistunov, B. L.

doi:10.11591/eei.v9i3.1714

Cited by 13 publications

(3 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It , , , h r m m , [8,[13][14][15][16]. For each specific case there are certain optimal correlations of values of the algorithm parameters, which reduce the number of working steps during decision making.…”

Section: Discussionmentioning

confidence: 99%

Control algorithm for decision making in multi-criteria optimization problems

Grishko

2022

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

The main feature of the multi-criteria optimization of the design processes of complex electronic devices is that they are all interconnected, heterogeneous in nature and, as a rule, non-linear. That is, by changing some of them by varying the corresponding parameters, we inevitably change other parameters of the system, and not always in the right direction. The purpose of the work is to propose a way to organize the procedure for finding the optimal solution to multicriteria design problems. When developing and creating new schemes, structures and systems, the task of determining the optimal values of the parameters of the designed technical device or system is reduced to the task of finding the extremum of a certain function, called the quality function. Usually, it is a non-linear function of many independent variables, which, as a rule, cannot be investigated by analytical methods. In this situation, it is advisable to search for a solution to the problem using special algorithms. At the early stages of designing complex radio electronic devices, an algorithm for finding the optimal solution based on the application of search optimization theory methods is proposed. The paper describes the procedure for finding the optimal solution, the distinctive feature of which is that during the search for a solution, information is accumulated and a search direction vector is constructed. This vector offers options for changing weights for subsequent search procedures. The algorithm developed on its basis is used in the information system for the complex analysis of the designs of radio-electronic facilities. The use of random search provides the algorithm with global properties, reduces the number of objective function calculations by one step, which is especially noticeable with a large number of optimized parameters. Allows you to quite simply include any restrictions on the set of valid values of the vector of optimized parameters, which is not always possible when using regular search methods. The proposed algorithm is simply tuned to search for a global extremum by selecting the appropriate parameters. For each specific case, there are some optimal ratios of the values of these algorithm parameters, which, when searching for a solution, reduce the number of working steps. The developed algorithm is less critical to the optimal values of its own parameters, which makes it much easier to tune it when solving practical problems of optimizing design processes.

show abstract

Section: Discussionmentioning

confidence: 99%

Control algorithm for decision making in multi-criteria optimization problems

Grishko

2022

J. Phys.: Conf. Ser.

View full text Add to dashboard Cite

show abstract

“…Average service times are calculated according to expressions previously presented in papers (Lopes et al, 2019). Thus, for the basic model ϑ AT = 140 ns, for the model under study ϑAT = 20 ns (according to the specification for the Avalon bus (Martyshkin et al, 2020;Martyshkin, 2021). The values of the average service time in the memory controller are the following: ϑRB = ϑWB = 10 ns.…”

Section: Methodologmentioning

confidence: 99%

Experimental research of a shared memory subsystem with limited queue length for specialized reconfigurable multiprocessor systems

Martyshkin

Martens-Atyushev

2022

Ind. Jour. Manag. & Prod.

View full text Add to dashboard Cite

Recently, reconfigurable systems based on field programmable logic devices (FPLDs) have been widely used in high-performance computing. The paper discusses issues related to the experimental research of a shared memory subsystem with a limited queue length of specialized reconfigurable multiprocessor systems using the developed mathematical modelling method. The paper presents the results of the method proposed by the authors for modelling multiprocessor systems based on open queuing networks with limited queue lengths. Based on these conditions, as well as the architectural features of the investigated processor-memory subsystem, expressions are calculated to estimate the exchange time and the resulting delays at each exchange stage. During the research, the main attention was paid to the dependence of the increase in the number of processor nodes in the processor-memory subsystem. As a result, the data obtained showed that the processor growth significantly affects the exchange time, creating a significant load on the common bus, as well as increasing delays at the stages when request transfer operation from the processor to the memory is performed. At the same time, the inadequate behaviour of experimental results and inaccuracy of their values when using the basic modelling method are explicitly tracked, which is reflected in the obtained graphs. Computational experiments were carried out to calculate the probabilistic-temporal characteristics of the "processor-memory" subsystem using the developed mathematical modelling methods. Based on the experimental results, it was determined that the delays occurring in subsystem's nodes and the time of exchange between the processor and memory modules depend on the query parameters and the processor-memory subsystem’s architectural characteristics.

show abstract

“…Parallelization of an original CA is based on the use of the results of the CES determination, which is the original algorithm. The CES determining process is performed, in turn, based on the use of a direct transition table (DTT) of the NDA for all particular events implemented in the CA Martyshkin et al, 2020;Nikolaevich et al, 2020;. Before starting the NDA DTT construction, all possible pairs of particular events are found in the original CA, which will correspond to the following four values: , , , ,з ,c i j j j…”

Section: Methodsmentioning

confidence: 99%

Parallel decomposition of control algorithms for computational processes based on the use of nondeterministic automaton logic

Pashchenko

Martyshkin

Trokoz

et al. 2021

Nexo Revista Científica

Self Cite

View full text Add to dashboard Cite

The paper deals with the issues of decomposition of control algorithms for the processes in parallel computing systems and the use of automaton models. When designing parallel processing systems, an important task is the formal presentation of process control algorithms since they allow achieving a packaged solution to the problems of specification, development, implementation, verification, and analysis of complex control systems, including the control of interacting processes and resources in parallel computing systems. It is especially necessary to use formal methods to verify complex information processing systems by model testing. One of the methods for the formal description of control algorithms is based on the use for these purposes of the nondeterministic automaton (NDA) logic, which is a method that allows one to present control algorithms for information processing in the form of systems of canonical equations describing all particular events implemented in the algorithm. The advantage of such a language is that all transitions in the control system are described not in terms of system states, but in terms of particular events, the simultaneous existence of which determines all states and transitions in the system; this allows avoiding a "combinatorial explosion" in the state space to the possibilities of means verification. Purpose of the paper: research of control algorithms for parallel computing systems using the NDA apparatus. The development and research object is parallel decomposition of control algorithms for parallel computing systems using automatic models.

show abstract

Using queuing theory to describe adaptive mathematical models of computing systems with resource virtualization and its verification using a virtual server with a configuration similar to the configuration of a given model

Cited by 13 publications

References 0 publications

Control algorithm for decision making in multi-criteria optimization problems

Control algorithm for decision making in multi-criteria optimization problems

Experimental research of a shared memory subsystem with limited queue length for specialized reconfigurable multiprocessor systems

Parallel decomposition of control algorithms for computational processes based on the use of nondeterministic automaton logic

Contact Info

Product

Resources

About