István Derka

55641948000

Publications - 6

Evaluation and improvement of parallel discrete event simulation performance predictions: A rough-set-based approach

László Muka István Derka

Publication Name: Acta Polytechnica Hungarica

Publication Date: 2016-01-01

Volume: 13

Issue: 6

Page Range: 125-145

Description:

Simulation performance prediction methods make possible the realization of performance improvement potentials of Parallel Discrete Event Simulation (PDES) methods, important in the analysis of complex systems and large-scale networks. Currently, high performance execution environments (emerging clusters and computing clouds) advance the development of quality/cost analysis capabilities of performance prediction methods. In this paper, for the evaluation and management of prediction correctness/cost, the efficacy, efficiency and effectiveness coefficients and improvement operations are defined for predictions. The performance coefficients and improvement operations are embedded in the rough-set-modeling and learning process and presented as an enhancement approach of the conventional Coupling Factor Method (CFM). A case study based on the CFM analysis of PDES of a closed queuing network model is presented. In the example, after rough-modeling and train-and-test analysis, the correctness/cost evaluation and effectiveness improvement operations are shown for series of predictions and the feedback connection to modeling refinement phase is demonstrated too.

Open Access: Yes

DOI: DOI not available

Towards the efficient simulation of telecommunication systems in heterogeneous distributed execution environments

László Muka István Derka Gábor Lencse

Publication Name: 2013 36th International Conference on Telecommunications and Signal Processing Tsp 2013

Publication Date: 2013-10-21

Volume: Unknown

Issue: Unknown

Page Range: 304-310

Description:

An introduction is given to the topics of the parallel and distributed simulation and of the modeling of telecommunications systems. Our practical modeling concept for simulation in heterogeneous execution environment is presented. Its logical topology is a star shaped network of homogeneous clusters. The load balancing and the coupling factor criteria are set up for building models of telecommunications systems so that the simulation may produce good speed-up in a heterogeneous distributed execution environment. A case study is given with the open source OMNeT++ discrete-event simulation system and its parallel CQN (closed queueing network) sample model executed by 64 CPU cores of four different types. Our criteria are heavily supported by the results of our experiments. © 2013 IEEE.

Open Access: Yes

DOI: 10.1109/TSP.2013.6613941

Investigation of the fault tolerance of the PIM-SM IP multicast routing protocol for IPTV purposes

István Derka Gábor Lencse

Publication Name: Infocommunications Journal

Publication Date: 2013-03-01

Volume: 5

Issue: 1

Page Range: 21-28

Description:

IPTV services should use an IP multicast solution for a network bandwidth efficient delivery of the media contents. PIM-SM is the most commonly used IP multicast routing protocol in IPTV systems. A short introduction to the operation of PIM-SM is given. Its fault tolerance is examined by experimenting on a mesh topology multicast test network built up by XORP routers in a virtualizcd environment. Different fault scenarios are played and different parameters of PIM-SM and OSPF are examined if they influence and how they influence the outage time of an IPTV service. A formal model is given for the service outage time of the IPTV service on the basis of the results of the experiments.

Open Access: Yes

DOI: DOI not available

Testing the speed-up of parallel discrete event simulation in heterogeneous execution environments

István Derka Gábor Lencse

Publication Name: 11th International Industrial Simulation Conference 2013 Isc 2013

Publication Date: 2013-01-01

Volume: Unknown

Issue: Unknown

Page Range: 101-107

Description:

This paper deals with the experimental testing and verification of the earlier proposed load balancing and coupling factor criteria for the conservative parallel discrete event simulation in heterogeneous execution environments whether they can ensure a good speed-up. The definition of the relative speed-up is extended to the heterogeneous systems in a natural way. This definition is used to measure the efficiency of the parallel simulation executed by heterogeneous systems. A closed queueing network is used as the simulation model, and it is executed on different heterogeneous test systems. Among several scenarios, it is demonstrated on the example of a heterogeneous system containing 87 CPU cores of 5 different types that a good speed-up can be achieved using the load balancing and coupling factor criteria. It is shown that the extension of the definition of the relative speed-up to the heterogeneous systems made it easy to judge the speed-up of parallel discrete event simulation in heterogeneous execution environments.

Open Access: Yes

DOI: DOI not available

Improving performance prediction of parallel and distributed discrete event simulation: A rough sets-based approach

László Muka István Derka

Publication Name: 11th International Industrial Simulation Conference 2013 Isc 2013

Publication Date: 2013-01-01

Volume: Unknown

Issue: Unknown

Page Range: 95-100

Description:

The rapid growth of complexity of communication networks - including grids, cloud systems and services - increases the need for the use and performance prediction of parallel and distributed discrete event simulation. The paper aims at the decreasing the overall cost of simulation by improving the performance prediction of simulation particularly for the phases with uncertain and vague information. Involving rough sets-based methods, the paper introduces how the efficiency of the well known coupling factor method of the performance prediction of parallel and distributed simulation can be improved. According to the presented results, the prediction can merely be based on the maintainable lookahead feature of the designed simulation model. Using the rough set analysis, the paper also describes how the number of simulation experiments necessary for prediction can be limited without the decrease of accuracy of prediction.

Open Access: Yes

DOI: DOI not available

Towards the modelling of the fault tolerance mechanism of the PIM-SM multicast routing protocol in an IPTV environment

István Derka Gábor Lencse

Publication Name: Esm 2012 2012 European Simulation and Modelling Conference Modelling and Simulation 2012

Publication Date: 2012-01-01

Volume: Unknown

Issue: Unknown

Page Range: 152-156

Description:

The fault tolerance mechanism of the PIM-SM IP multicast routing protocol is investigated in order to be able to model it. The analysis is done by playing different fault scenarios on a mesh topology multicast test network built up by XORP routers in a virtualized environment. Different parameters of the PIM-SM and the OSPF protocols are examined if they influence and how they influence the outage time of an IPTV service. The results of the experiments provide important factors for building a formal model of the service outage time of an IPTV service. © 2012 EUROSIS-ETI.

Open Access: Yes

DOI: DOI not available