DYNAMIC METHOD OF LOAD BALANCING OF DATA CENTERS TAKING INTO ACCOUNT THE FRACTAL PROPERTIES OF NETWORK TRAFFIC

An approach to the development and study of a load balancing system for data centers (DC) taking into account the fractal properties of network traffic is proposed. The fractal properties of network traffic make it possible to predict, with a fairly high probability, the appearance of bursts and drops in its activity at certain time intervals, the appearance of time periods with a possible overload in the performance of servers and network equipment, which makes it possible to develop methods for efficient planning and distribution of tasks within the data center, ensuring a statistically uniform loading its functional elements, eliminating the occurrence of overloads. The dynamic method is based on the results of statistical analysis of incoming traffic, its distribution density, autocorrelation function, spectral density, and the level of fractality. The practical significance of the proposed research results lies in the substantiation and development of a method for dynamic distribution and load balancing in cloud computing, taking into account the fractal properties of network traffic. The methods and models presented in the work can be applied in existing data processing centers to increase their productivity and improve the quality of implementation of information services.

network traffic, data centers, load balancing, fractals, time series, autocorrelation function, Hurst parameter, autoregressive models

1. Mochalov V.P., Linets G.I., Bratchenko N.Y., Govorova S.V. (2020). "An analytical model of a corporate software-controlled network switch", Scalable Computing, 21 (2), pp. 337346.

2. Boev V. Kompjuternoe modelirovanie: Posobie dlja prakticheskih zanjatij, kursovogo i diplomnogo proektirovanija v AnyLogic7 [Computer modeling: A manual for practical classes, course and diploma projects in AnyLogic7] St. Petersburg, VAS Publ., 2014, 432p. (In Russian).

3. Taihoon K., Soksoo K. Analysis of Security Session Reusing in Distribution Server System.Computational Science and Its Applications - ICCSA 2006. Springer, 2006, 1045 p.

4. Khritankov A. Modeli i algoritmy raspredelenija nagruzki. Algoritmy na osnove setej SMO [Models and algorithms of load balancing. Algorithms on the basis of networks of queuing systems]. Informacionnye tehnologii i vychislitel'nye seti - Information technologies and computer networks. 2009, vol. 3. (In Russian).

5. Ivanisenko I., Kirichenko L., Radivilova T. Metody balansirovki s uchetom multifraktal'nyh svojstv nagruzki [Balancer multifractal methods considering load characteristics].International Journal "Information Contentand Processing". 2015, vol. 2, no. 4, pp. 345 -368. (In Russian).

6. Panchenko T.V. Genetic Algorithms [Text]: teaching aid /Yu.Yu. Tarasevich. Astrakhan: «Astrakhanskiy Universitet»,2007. - 87 p. (In Russian)

7. Holland J.H. Adaptation in Natural and Artificial Systems:An Introductory Analysis with Applications to Biology,Control, and Artificial Intelligence /j. H. Holland. The MITPress, Cambridge, 1992. - 211 p.

8. Michalewicz Z. Genetic algorithms + Data Structures=Evolution Programs/Z. Micha-lewicz. - New York: Springer-Verlag, 1996. - 387.

9. Tsoy Yu.R., Spitsyn V.G. Genetic Algorithm / SpitsynV.G., Tsoy Yu.R. Knowledge Representation in InformationSystems: a Tutorial. - Tomsk: TPU, 2006. - 146 p. (In Russian).

10. Mitchell M. An Introduction to Genetic Algorithms/M.Mitchell. - Cambridge: MIT Press, 1999. -158 p.

11. Periaux J., Sefrioui M. Evolutionary computationalmethods for complex design in aerodynamics // AIAA-98-0222. - Reno, 1998. - 15 p.

12. Periaux J, Chen HQ, Mantel B, Sefrioui M, Sui HT(200i) Combining game theory and genetic algorithms withapplication to DDM-nozzle optimization problems. Finite ElemAnal Des 37(5), pp. 417-429.

13. Zhirkov A Supercomputers: developing, trends, usage.Eurotech HPC solutions review [Superkomp'yutery: razvitie,tendentsii, primenenie. Obzor HPC-resheniy Eurotech] Presentautomatization solutions 2014, no. 2, p. 201.

14. Taihoon K., Soksoo K. Analysis of Security Session Reusing in Distribution Server System // Computational Science and Its Applications - ICCSA 2006. / Springer, 2006. - 1045 с.

15. Beloglazov A., Buyya R. Optimal online deterministic algorithms and adaptive heuristics for energy and performance efficient dynamic consolidation of virtual machines in cloud data centers, Concurrency and Computation: Pract.

16. Mochalov V.P., Bratchenko N.Y., Yakovlev S.V. (2018). "Analytical model of object request broker based on Corba standard", Journal of Physics: Conference Series, 1015 (2). doi: 10.1088/1742-6596/1015/2/022012.

17. Mochalov V.P., Bratchenko N.Y., Yakovlev S.V. (2018). "Analytical model of integration system for program components of distributed object applications", International Russian Automation Conference, RusAutoCon 2018, № 8501806. doi: 10.1109/ RUSAUTOCON.2018.8501806.

18. Mochalov V., Bratchenko N., Linets G., Yakovlev S. (2019). "Distributed management systems for infocommunication networks: A model based on tm forum frameworx", Compute rs 2019, 8(2). doi: 10.3390/computers8020045.

19. Mochalov V.P., Bratchenko N.Y., Yakovlev S.V. (2019). "Process-Oriented Management System for Infocommunication Networks and Services Based on TM Forum Frameworx", Proceedings - 2019 International Russian Automation Conference, RusAutoCon 2019, № 8867619. doi: 10.1i09/RUSAUTOCON.2019.8867619.

20. Vdovin P.M., Kostenko V.A. Algorithm for Resource Allocation in Data Centers with Independent Schedulers for Different Typesof Resources //j.Computer and Systems Sciences International. 2014. № 6.1.