Open Access Open Access  Restricted Access Subscription Access

Impulsive and Switching Load Balancing Model and Stability Analysis


Affiliations
1 School of Computer Science and Engineering, Northeastern University, Shenyang - 110819, China
 

Load imbalance in a distributed system causes low efficiency. Aiming at the dynamics of resource and load running status, we propose an impulsive and switching load balancing model with time delay based on control theory. In order to describe various current states of a node, we construct corresponding subsystems according to the dynamics of a node’s resources. The sub-system switching is triggered by an impulsive signal which can decrease the communication overhead among nodes. The model reallocates loads in light of their real-time running statuses, which improves the efficiency of dynamic load balancing. We deduce the sufficient condition for asymptotic stability of the model by using the Lyapunov–Krasovskii function, and simulation by Linear Matrix Inequality (LMI) to verify effectiveness of the model. Experimental results produced by a shared platform based on block chain demonstrate that the proposed model can make the balanced system speedy, which verifies its feasibility.

Keywords

Asymptotic Stability, Dynamic Load Balancing, Distributed System, Impulsive and Switching System, Time Delay.
User
Notifications
Font Size

  • Meng, Q. Y., Qiao, J. Z. and Lin, S. K., A delay-based dynamic load balancing method and its stability analysis and simulation, Proceedings of European Conference on Parallel Computing, Ischia, 2010, pp. 192–203.
  • Hsiao, H. C., Chung, H. Y. and Shen, H., Load rebalancing for distributed file systems in clouds. IEEE Trans. Parall. Distr. Syst., 2013, 24, 951–962.
  • Dhakal, S., Hayat, M. M., Pezoa, J. E., Yang, C. D. and Bader, D. A., Dynamic load balancing in distributed systems in the presence of delays: a regeneration-theory approach. IEEE Trans. Parall. Distr. Syst., 2007, 18, 485–497.
  • Tang, Z., Birdwell, J. D., Chiasson, J., Abdallah, C. T. and Hayat, M., Resource-constrained load balancing controller for a parallel database. IEEE Trans. Control Syst. Technol., 2008, 16, 834–840.
  • Tang, Z., White, J. and Chiasson, J., Closed-loop load balancing: comparison of a discrete event simulation with experiment. Proceedings of the 2005 American Control Conference, Portland, 2005, pp. 2721–2726.
  • Liu, X., Zhong, S. M. and Ding, X. Y., Robust exponential stability of impulsive switched systems with switching delays: a Razumikhin approach. Commun. Nonlinear Sci. Numer. Simul., 2012, 17, 1805–1812.
  • Shim, H. and Tanwani, A., Hybrid-type observer design based on a sufficient condition for observability in switched nonlinear systems. Int. J. Robust Nonlin., 2014, 24, 1064–1089.
  • Liu, X. W., Robust H∞ filtering for switched discrete-time systems with time-delays. Proceedings of the 26th Chinese Control Conference, Zhangjiajie, 2007, pp. 660–664.
  • Liu, J., Liu, X. Z. and Xie, W. C., Delay-dependent robust control for uncertain switched systems with time-delay. Nonlinear Analy. Hybri. Syst., 2008, 2, 81–95.
  • Egorov, A. V. and Mondi, S., Necessary conditions for the exponential stability of time-delay systems via the Lyapunov delay matrix. Nonlinear Anal. Hybri., 2014, 24, 1760–1771.

Abstract Views: 302

PDF Views: 127




  • Impulsive and Switching Load Balancing Model and Stability Analysis

Abstract Views: 302  |  PDF Views: 127

Authors

Tinglei Zhao
School of Computer Science and Engineering, Northeastern University, Shenyang - 110819, China
Jianzhong Qiao
School of Computer Science and Engineering, Northeastern University, Shenyang - 110819, China
Shukuan Lin
School of Computer Science and Engineering, Northeastern University, Shenyang - 110819, China
Yanhua Wang
School of Computer Science and Engineering, Northeastern University, Shenyang - 110819, China

Abstract


Load imbalance in a distributed system causes low efficiency. Aiming at the dynamics of resource and load running status, we propose an impulsive and switching load balancing model with time delay based on control theory. In order to describe various current states of a node, we construct corresponding subsystems according to the dynamics of a node’s resources. The sub-system switching is triggered by an impulsive signal which can decrease the communication overhead among nodes. The model reallocates loads in light of their real-time running statuses, which improves the efficiency of dynamic load balancing. We deduce the sufficient condition for asymptotic stability of the model by using the Lyapunov–Krasovskii function, and simulation by Linear Matrix Inequality (LMI) to verify effectiveness of the model. Experimental results produced by a shared platform based on block chain demonstrate that the proposed model can make the balanced system speedy, which verifies its feasibility.

Keywords


Asymptotic Stability, Dynamic Load Balancing, Distributed System, Impulsive and Switching System, Time Delay.

References





DOI: https://doi.org/10.18520/cs%2Fv115%2Fi2%2F258-267