Source-Destination Routing Optimal Strategies Eric Chi EE228a, Fall 2002 Dept. of EECS, U.C. Berkeley.

Slides:



Advertisements
Similar presentations
Routing and Congestion Problems in General Networks Presented by Jun Zou CAS 744.
Advertisements

Markov Decision Processes (MDPs) read Ch utility-based agents –goals encoded in utility function U(s), or U:S  effects of actions encoded in.
Markov Decision Process
Dynamic Decision Processes
Data and Computer Communications Ninth Edition by William Stallings Chapter 12 – Routing in Switched Data Networks Data and Computer Communications, Ninth.
An Introduction to Markov Decision Processes Sarah Hickmott
Markov Decision Processes
Infinite Horizon Problems
Planning under Uncertainty
The Cache Location Problem IEEE/ACM Transactions on Networking, Vol. 8, No. 5, October 2000 P. Krishnan, Danny Raz, Member, IEEE, and Yuval Shavitt, Member,
SA-1 1 Probabilistic Robotics Planning and Control: Markov Decision Processes.
Detecting Network Intrusions via Sampling : A Game Theoretic Approach Presented By: Matt Vidal Murali Kodialam T.V. Lakshman July 22, 2003 Bell Labs, Lucent.
Lecture 3. Preview of Markov Process A sequence of random variables X 1, X 2,….,X n,….. such that –X i+1 is independent of X 1,….X i-1 given X i –Pr(X.
December 20, 2004MPLS: TE and Restoration1 MPLS: Traffic Engineering and Restoration Routing Zartash Afzal Uzmi Computer Science and Engineering Lahore.
Reinforcement Learning Mitchell, Ch. 13 (see also Barto & Sutton book on-line)
November 2001Dynamic Alternative Routing1 Yashar Ganjali Stanford University November 2001.
1 Hybrid Agent-Based Modeling: Architectures,Analyses and Applications (Stage One) Li, Hailin.
Discretization Pieter Abbeel UC Berkeley EECS
UCB Tools for Smart Networks Jean Walrand BITS (Berkeley Information Technology & Systems) U.C. Berkeley
Multiagent Planning with Factored MDPs Carlos Guestrin Daphne Koller Stanford University Ronald Parr Duke University.
Markov Decision Processes Value Iteration Pieter Abbeel UC Berkeley EECS TexPoint fonts used in EMF. Read the TexPoint manual before you delete this box.:
Department of Computer Science Undergraduate Events More
More RL. MDPs defined A Markov decision process (MDP), M, is a model of a stochastic, dynamic, controllable, rewarding process given by: M = 〈 S, A,T,R.
Reinforcement Learning Yishay Mansour Tel-Aviv University.
S. Suri, M, Waldvogel, P. Warkhede CS University of Washington Profile-Based Routing: A New Framework for MPLS Traffic Engineering.
Interference-aware QoS Routing (IQRouting) for Ad-Hoc Networks Rajarshi Gupta, Zhanfeng Jia, Teresa Tung, and Jean Walrand Dept of EECS, UC Berkeley Globecom.
Cost-Performance Tradeoffs in MPLS and IP Routing Selma Yilmaz Ibrahim Matta Boston University.
MAKING COMPLEX DEClSlONS
Integrated Dynamic IP and Wavelength Routing in IP over WDM Networks Murali Kodialam and T. V. Lakshman Bell Laboratories Lucent Technologies IEEE INFOCOM.
Profile-Based Topology Control and Routing of Bandwidth-Guaranteed Flows in Wireless Optical Backbone Networks A. Kashyap, M.K. Khandani, K. Lee, M. Shayman.
Flows and Networks Plan for today (lecture 5): Last time / Questions? Blocking of transitions Kelly / Whittle network Optimal design of a Kelly / Whittle.
A Call Admission Control for Service Differentiation and Fairness Management in WDM Grooming Networks Kayvan Mosharaf, Jerome Talim and Ioannis Lambadaris.
1 Call Admission Control for Multimedia Services in Mobile Cellular Networks : A Markov Decision Approach Jihyuk Choi; Taekyoung Kwon; Yanghee Choi; Naghshineh,
Flows and Networks Plan for today (lecture 6): Last time / Questions? Kelly / Whittle network Optimal design of a Kelly / Whittle network: optimisation.
1 ECE-517 Reinforcement Learning in Artificial Intelligence Lecture 7: Finite Horizon MDPs, Dynamic Programming Dr. Itamar Arel College of Engineering.
Behrouz Haji Soleimani Dr. Moradi. Outline What is uncertainty? Some examples Solutions to uncertainty Ignoring uncertainty Markov Decision Process (MDP)
Dynamic Programming for Partially Observable Stochastic Games Daniel S. Bernstein University of Massachusetts Amherst in collaboration with Christopher.
Networks of Queues Plan for today (lecture 6): Last time / Questions? Product form preserving blocking Interpretation traffic equations Kelly / Whittle.
Presenter: Jonathan Murphy On Adaptive Routing in Wavelength-Routed Networks Authors: Ching-Fang Hsu Te-Lung Liu Nen-Fu Huang.
Examination Committee: Dr. Poompat Saengudomlert (Chairperson) Assoc. Prof. Tapio Erke Dr. R.M.A.P. Rajatheva 1 Telecommunications FoS Asian Institute.
Examination Committee: Dr. Poompat Saengudomlert (Chairperson) Assoc. Prof. Tapio Erke Dr. R.M.A.P. Rajatheva 1 Telecommunications FoS Asian Institute.
1 Markov Decision Processes Infinite Horizon Problems Alan Fern * * Based in part on slides by Craig Boutilier and Daniel Weld.
Performance Evaluation of TCP over Multiple Paths in Fixed Robust Routing Wenjie Chen, Yukinobu Fukushima, Takashi Matsumura, Yuichi Nishida, and Tokumi.
Reinforcement Learning Yishay Mansour Tel-Aviv University.
1 Markov Decision Processes Infinite Horizon Problems Alan Fern * * Based in part on slides by Craig Boutilier and Daniel Weld.
1 - CS7701 – Fall 2004 Review of: Detecting Network Intrusions via Sampling: A Game Theoretic Approach Paper by: – Murali Kodialam (Bell Labs) – T.V. Lakshman.
1 ECE 517: Reinforcement Learning in Artificial Intelligence Lecture 8: Dynamic Programming – Value Iteration Dr. Itamar Arel College of Engineering Department.
MDPs (cont) & Reinforcement Learning
Flows and Networks Plan for today (lecture 6): Last time / Questions? Kelly / Whittle network Optimal design of a Kelly / Whittle network: optimisation.
An Optimal Distributed Call Admission control for Adaptive Multimedia in Wireless/Mobile Networks Reporter: 電機所 鄭志川.
1 Presented by Sarbagya Buddhacharya. 2 Increasing bandwidth demand in telecommunication networks is satisfied by WDM networks. Dimensioning of WDM networks.
Design and Analysis of Algorithms (09 Credits / 5 hours per week) Sixth Semester: Computer Science & Engineering M.B.Chandak
1 Dynamic RWA Connection requests arrive sequentially. Setup a lightpath when a connection request arrives and teardown the lightpath when a connection.
Reinforcement Learning Dynamic Programming I Subramanian Ramamoorthy School of Informatics 31 January, 2012.
Stochastic Optimization for Markov Modulated Networks with Application to Delay Constrained Wireless Scheduling Michael J. Neely University of Southern.
Department of Computer Science Undergraduate Events More
MDPs and Reinforcement Learning. Overview MDPs Reinforcement learning.
Some Final Thoughts Abhijit Gosavi. From MDPs to SMDPs The Semi-MDP is a more general model in which the time for transition is also a random variable.
Flows and Networks Plan for today (lecture 6): Last time / Questions? Kelly / Whittle network Optimal design of a Kelly / Whittle network: optimisation.
1 Minimum Interference Algorithm for Integrated Topology Control and Routing in Wireless Optical Backbone Networks Fangting Sun Mark Shayman University.
Constraint-Based Routing
Making complex decisions
Markov Decision Processes
Dynamic Programming Lecture 13 (5/31/2017).
Markov Decision Processes
Flows and Networks Plan for today (lecture 6):
Markov Decision Problems
CS 188: Artificial Intelligence Spring 2006
Hidden Markov Models (cont.) Markov Decision Processes
Reinforcement Learning Dealing with Partial Observability
Presentation transcript:

Source-Destination Routing Optimal Strategies Eric Chi EE228a, Fall 2002 Dept. of EECS, U.C. Berkeley

Basic Routing Problem Network with links of finite capacity Connection requests for various node-pairs arrive one by one A decision is made to either –deny the request or –admit the connection along a given route An admitted call simultaneously holds some capacity along all links along the route for some amount of time before departing Objective: Make decisions that minimize blocking probability

Approaches Suboptimal: Greedy algorithms –Always admit if there is space. –Choose good heuristics for where to place calls. Maximize spare capacity Minimize “Interference” Optimal: Dynamic programming –Balances Immediate gains Long term opportunity costs

Markov Decision Process State specified by a Markov Chain –Request arrivals are Poisson –Calls holding times are exponentially distributed Rewards (Costs) associated with –Residing in a state –Making a transition Transition probabilities depend on policies for a given state.

Discrete Time MDP

Bellman Principle of Optimality Given an optimal control for n steps to go, the last n-1 steps provide optimal control with n-1 steps to go. Example: Dijstkra’s Shortest Path Algorithm

Solving MDPs: Value Iteration Solve the fixed point equation. Then

Solving MDPs: Policy Iteration

Example: Symmetric Y/C X/C ’ Optimal Policy: Route to least loaded

Proof (Sketch) Prove that load balancing is optimal for any finite time to go n. (Monotone convergence allows us to take the limit.) Prove inductively that for all n, , a

Example: Unbalanced Y/C X/C   2

Example: Unbalanced Y/C X/C ’ Optimal Policy: Route to lower link until full. If full route to top link.

Comparison

Example: Alternate Routing Policy A: Route up 1 st, Route down 2 nd Policy B: Route down 1 st, Route up 2 nd Y/C X/C  2

Comparison Two policies

Literature K. R. Krishnan and T. J. Ott, "State-dependent routing for telephone traffic: theory and results," in 25th IEEE Control and Decision Conf., Athens, Greece, Dec. 1986, pp A. Ephremides, P. Varaiya, and J. Walrand. A simple dynamic routing problem. IEEE Transactions on Automatic Control, 25(4): , August R.J. Gibbon and F.P. Kelly. Dynamic routing in fully connected networks. IMA journal of Mathematical Control and Information, 7: , Marbach, P., Mihatsch, M., Tsitsiklis, J.N., "Call admission control and routing in integrated service networks using neuro-dynamic programming," IEEE J. Selected Areas in Comm., v. 18, n. 2, pp , Feb K. Kar, M. Kodialam, and T.V. Lakshman, “Minimum Interference Routing of Bandwidth Guaranteed Tunnels with Applications to MPLS Traffic Engineering,” IEEE JSAC, 1995, Special Issue on Advances in the Fundamentals of Networking, pp