Abstract
Bonald et al. have recently characterized a set of insensitive dynamic load balancing policies by modelling the system as a Whittle network. In particular, they derived optimal “decentralized” strategies based on limited state information and evaluated their performance in simple example networks. In this paper, we consider the specific case of a data network where each flow can be routed on one of a set of alternative routes. By using the linear programming formulation of MDP theory we are able to analyze optimal routing policies that utilize the full global state information. In the ordinary LP formulation of MDP theory, the global balance condition appears as a linear constraint on the decision variables. In order to retain insensitivity, we impose stricter detailed balance conditions as constraints. As a further extension, the MDP-LP approach allows joint optimization of the routing and resource sharing, in contrast to the earlier work where the resource sharing policy was required to be separately balanced and fixed in advance. The various schemes are compared numerically in a toy network. The advantage given by global state information is in this case negligible, whereas the joint routing and resource sharing gives a clear improvement. The requirement of insensitivity still implies some performance penalty in comparison with the best sensitive policy.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Combe, M.B., Boxma, O.J.: Optimization of static traffic allocation policies. Theoretical Computer Science 125, 17–43 (1994)
Ephremides, A., Varaiya, P., Walrand, J.: A simple dynamic routing problem. IEEE Transactions on Automatic Control 25, 690–693 (1980)
Towsley, D., Panayotis, D., Sparaggis, P., Cassandras, C.: Optimal routing and buffer allocation for a class of finite capacity queueing systems. IEEE Transactions on Automatic Control 37, 1446–1451 (1992)
Whitt, W.: Deciding which queue to join: some counterexamples. Operations Research 34, 226–244 (1986)
Bonald, T., Proutière, A.: Insensitivity in processor-sharing networks. Performance Evaluation 49, 193–209 (2002)
Bonald, T., Prouti‘ere, A.: Insensitivite bandwidth sharing in data networks. Queueing Systems and Applications 44, 69–100 (2003)
Bonald, T., Jonckheere, M., Proutière, A.: Insensitive dynamic load balancing. In: Proceedings of ACM Sigmetrics (2004)
Krishnan, K.: Markov decision algorithms for dynamic routing. IEEE Communications Magazine 28, 66–69 (1990)
Altman, E.: Applications of Markov decision processes in communication networks: A survey. In: Handbook of Markov Decision Processes. Kluwer, Dordrecht (2002)
Tijms, H.: Stochastic Models: An Algorithmic Approach. John Wiley & Sons, Chichester (1994)
Feinberg, E.A., Schwartz, A. (eds.): Handbook ofMarkov Decision Processes - Methods and Applications. Kluwer, Dordrecht (2002)
Bonald, T., Proutière, A.: On performance bounds for balanced fairness. Performance Evaluation 55, 25–50 (2004)
Massoulié, L., Roberts, J.W.: Arguments in favour of admission control for TCP flows. In: Proceedings of ITC, vol. 16 (1999)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2005 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Leino, J., Virtamo, J. (2005). Optimal Load Balancing in Insensitive Data Networks. In: Ajmone Marsan, M., Bianchi, G., Listanti, M., Meo, M. (eds) Quality of Service in Multiservice IP Networks. QoS-IP 2004. Lecture Notes in Computer Science, vol 3375. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-30573-6_24
Download citation
DOI: https://doi.org/10.1007/978-3-540-30573-6_24
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-24557-5
Online ISBN: 978-3-540-30573-6
eBook Packages: Computer ScienceComputer Science (R0)