Abstract
In this chapter, VNF scalability via joint resource scaling and migration is studied in a local network segment, to meet a probabilistic delay requirement in the presence of non-stationary traffic with changing traffic statistics. A change point detection scheme determines boundaries between stationary traffic segments in an online manner as new traffic samples arrive, and provides a triggering signal for VNF scalability. Under the fBm traffic model assumption, the resource demand for a probabilistic delay guarantee is predicted for each newly detected stationary traffic segment, based on traffic parameter learning with Gaussian process regression and fBm resource provisioning model. Given the predicted resource demand, a dynamic VNF migration problem is formulated as a Markov decision process (MDP) with variable-length decision epochs, to maximize the long-term reward integrating load balancing, migration cost, and resource overloading penalty. The MDP is solved by a reinforcement leaning (RL) approach. Specifically, a penalty-aware deep \(\mathcal {Q}\)-learning algorithm demonstrating advantages in reducing training loss and increasing cumulative reward is employed to learn the adaptive VNF migration actions under the dynamics in change points, resource demand, and background traffic.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Notes
- 1.
Note that the i.i.d Gaussian assumption is used to detect change points. For traffic parameter learning, we do not rely on such an assumption.
- 2.
The value of i 0 should be smaller than the minimum value of the most probable run lengths at any detected change points.
- 3.
Here we refer to the VNF packet processing rate as the computing resource demand for simplicity. The CPU computing resource demand in cycle/s can be mapped from the resource demand in packet/s with the processing density model in Chap. 3.
- 4.
A Q-Q plot is a probability plot, which compares two probability distributions by plotting their quantiles against each other. If the two probability distributions are similar, the points in the Q-Q plot will approximately lie on the diagonal line y = x.
References
Krithikaivasan, B., Zeng, Y., Deka, K., Medhi, D.: ARCH-based traffic forecasting and dynamic bandwidth provisioning for periodically measured nonstationary traffic. IEEE/ACM Trans. Netw. 15(3), 683–696 (2007)
Fraleigh, C., Tobagi, F., Diot, C.: Provisioning IP backbone networks to support latency sensitive traffic. In: Proc. IEEE INFOCOM, pp. 1871–1879 (2003)
Kim, J., Hwang, G.: Adaptive bandwidth allocation based on sample path prediction with Gaussian process regression. IEEE Trans. Wirel. Commun. 18(10), 4983–4996 (2019)
Cheng, Y., Zhuang, W., Wang, L.: Calculation of loss probability in a finite size partitioned buffer for quantitative assured service. IEEE Trans. Commun. 55(9), 1757–1771 (2007)
Kim, H.S., Shroff, N.B.: Loss probability calculations and asymptotic analysis for finite buffer multiplexers. IEEE/ACM Trans. Netw. 9(6), 755–768 (2001)
Adams, R.P., MacKay, D.J.: Bayesian online changepoint detection. Tech. rep., University of Cambridge, Cambridge, UK (2007)
Murphy, K.P.: Conjugate Bayesian analysis of the Gaussian distribution. def 1(2σ2), 16 (2007)
Bishop, C.M.: Pattern Recognition and Machine Learning. Springer, New York (2006)
Bayati, A., Asghari, V., Nguyen, K., Cheriet, M.: Gaussian process regression based traffic modeling and prediction in high-speed networks. In: Proc. IEEE GLOBECOM, pp. 1–7 (2016)
Williams, C.K., Rasmussen, C.E.: Gaussian Processes for Machine Learning, vol. 2. MIT Press, Cambridge (2006)
Mnih, V., Kavukcuoglu, K., Silver, D., Rusu, A.A., Veness, J., Bellemare, M.G., Graves, A., Riedmiller, M., Fidjeland, A.K., et al.: Human-level control through deep reinforcement learning. Nature 518(7540), 529–533 (2015)
Xu, Z., Tang, J., Meng, J., Zhang, W., Wang, Y., Liu, C.H., Yang, D.: Experience-driven networking: A deep reinforcement learning based approach. In: Proc. IEEE INFOCOM, pp. 1871–1879 (2018)
Schaul, T., Quan, J., Antonoglou, I., Silver, D.: Prioritized experience replay. In: Proc. ICLR’16, pp. 1–7 (2016)
MAWI Working Group Traffic Archive (2021). http://mawi.wide.ad.jp/mawi/. Accessed 14 July 2021
Abry, P., Sellan, F.: The wavelet-based synthesis for fractional Brownian motion proposed by F. Sellan and Y. Meyer: remarks and fast implementation. Appl. Comput. Harmon. Anal. 3(4), 377–383 (1996)
Kobayashi, H., Mark, B.L.: System Modeling and Analysis: Foundations of System Performance Evaluation. Pearson Education India (2009)
Qu, K., Zhuang, W., Ye, Q., Shen, X., Li, X., Rao, J.: Dynamic resource scaling for VNF over nonstationary traffic: a learning approach. IEEE Trans. Cogn. Commun. Netw. 7(2), 648–662 (2021)
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2021 The Author(s), under exclusive license to Springer Nature Switzerland AG
About this chapter
Cite this chapter
Zhuang, W., Qu, K. (2021). Dynamic VNF Resource Scaling and Migration: A Machine Learning Approach. In: Dynamic Resource Management in Service-Oriented Core Networks. Wireless Networks. Springer, Cham. https://doi.org/10.1007/978-3-030-87136-9_4
Download citation
DOI: https://doi.org/10.1007/978-3-030-87136-9_4
Published:
Publisher Name: Springer, Cham
Print ISBN: 978-3-030-87135-2
Online ISBN: 978-3-030-87136-9
eBook Packages: Computer ScienceComputer Science (R0)