An Evolutionary based Dynamic A E l ti b d D i Energy Management Framework for Energy Management Framework for IP-over-DWDM Core Networks Xin Chen, Chris Phillips School of Electronic Engineering & C Computer Science t S i
O tli Outline � Introduction � Background k d � New Energy Management Design � New Energy Management Design � Simulation � Further work 2
Introduction Introduction � Benefits of Energy Saving � Economical Lower OPEX for ISPs Lower OPEX for ISPs � Environmental Lower CO 2 emissions 3
Introduction d i � Our energy management scheme combines gy g infrastructure sleeping and virtual router migration together with automatic optical migration together with automatic optical layer connection forwarding to enable resources to be used in an energy efficient resources to be used in an energy-efficient manner. 4
Background k d � Network Architecture � Network Architecture � Energy Saving Approaches � Infrastructure sleeping and Virtual Router Migration Techniques Router Migration Techniques 5
Network Architecture Network Architecture � IP over DWDM IP over DWDM We apply the wavelength continuity constraint. There is no wavelength conversion for through-traffic in the network. 6
Energy Saving Approaches S i A h � Static Mechanisms � Static Mechanisms Network planning, i.e. ILP � Dynamic Mechanisms Infrastructure sleeping, rate adaptation, f l i d i network virtualization….. 7
Infrastructure Sleeping p g � Switch off unneeded equipment during off-peak periods periods � Previous work [over 20% saving] L. Chiaraviglio, M. Mellia, and F. Neri, "Energy-Aware Backbone Networks: A Case Study" in IEEE International Conference on Communications Workshops, 2009, pp. 1–5, June 2009. 8
Infrastructure Sleeping f Sl i Some issues and limitations Some issues and limitations � � � The problem of loss of connectivity due p y to reconvergence � When to sleep / wake ? � When to sleep / wake ? How to sleep / wake ? � 9
Virtual Router Migration g � Move virtual routers among difference physical platform without degrading the service i h d di h i Wang, Yi,; Keller, E.; Biskeborn, B.; Jacobus van der Merwe, Rexford, J.; ,"Virtual routers on the move: live router migration as a network-management primitive," SIGCOMM Comput. Commun. Rev. 38, 4 (August 2008), 231-242. 10
Virtual Router Migration i l i i Some issues and limitations � When to trigger virtual router migration? When to trigger virtual router migration? � � Where to move virtual routers to? � 11
Dynamic Energy Management Dynamic Energy Management Framework � Overall Energy Management Procedure � Optical Connection Management � VRM_MOEA Virtual Router Migration – Multi-Objective Evolutionary Algorithm 12
Overall Energy Management Procedure Overall Energy Management Procedure 1 C ll 1. Collect and analyze the network status t d l th t k t t 2. Trigger VRM MOEA. (Quiet and Busy Thresholds) gg _ (Q y ) 3. Establish the new optical connections 4. Virtual router migration 5 Then Switch off (on) the corresponding physical 5. Then Switch off (on) the corresponding physical platforms and removed the unneeded optical connections connections 6. Go back to step 1 to recheck the network status 13
Dynamic Optical Connection Management � Additional optical connections are needed for Additi l ti l ti d d f forwarding the traffic to the remote virtual router(s) processing the packets t ( ) i th k t � Changes in the underlying physical network are hidden from the topology as seen by Layer-3 and p gy y y so reconvergence events are avoided 14
15 Example: Optical Connection Management
Destination Physical Platform Selection y Algorithm- VRM_MOEA � Individual chromosome representation: Chromosome length set to sum of VRs Gene VR index Gene, VR index 1 1 2 2 Allele gives PP location 16
Destination Physical Platform Selection y Algorithm - VRM_MOEA Initial Population: Pre screening procedure for selecting the variable solutions Evolutionary algorithm well- suited to real-time operation as search can be halted at any point and we only need a “good” and we only need a good solution 17
Destination Physical Platform Selection y Algorithm- VRM_MOEA T Two objective functions: bj ti f ti 1.Power Consumption: α α ∑ = α ⋅ + ⋅ + θ ⋅ β α ⋅ + β ⋅ （ ） P P N P - P P total base i lc base roadm = i 1 On PPs On PPs On PP Linecards On PP Linecards Off PPs Off PPs On ROADMs On ROADMs P α to ta l ----- The number of active PPs ----- The power consumption of β ----- The number of ROADM in the the network P P network ba se ----- The power consumption of θ ----- A percentage of the base system base system P lc power consumption a PP consumes ----- The power consumption of when it is sleeping. h i i l i a line card P ro ad m ----- The number of active line cards in N ----- The power consumption of i the i -th PP a ROADM 18
Destination Physical Platform Selection Algorithm- VRM_MOEA Al ith VRM MOEA 2. Virtual Router Migration Cost (Second objective function) � The first VRM cost component comes from hop count f from an original “home” VR location to its destination PP. i i l “h ” VR l ti t it d ti ti PP For i-th candidate solution: β β = ∑ j j Cost _ ( ) a i d g ( , g ) 0 i = j 1 d ( 1, x x 2 ) ----- A function for obtaining the distance between two PPs: x1 and x2. i g ----- j -th gene in the default network configuration 0 j g ----- j -th gene in a candidate solution. i β ----- The number of VR 19
Destination Physical Platform Selection Al Algorithm- VRM_MOEA ith VRM MOEA � The second component comes from the virtual router � The second component comes from the virtual router migration process β β = ∑ j j Cost _ ( ) b i d g ( , g ) current i = j 1 i g g ----- j -th gene in the current network j g curren t curren t configuration ----- j -th gene in a candidate solution. j g i � Therefore The overall cost of one possible solution � Therefore, The overall cost of one possible solution is : = ϕ ϕ ⋅ + ϕ ϕ ⋅ Cost i ( ) ( ) Cost _ ( ) a i ( ) (1- ) ( ) Cost _ ( ) b i ( ) ϕ ----- Weight of two cost terms 20
Destination Physical Platform Selection y Algorithm- VRM_MOEA � Fi � Fitness function: f i Strength Pareto Evolutionary Algorithm II (SPEA2) � Selection Mechanism: To rn ment sele tion Tournament selection � Crossover Operation: BLX- α crossover � Mutation Operation: � Mutation Operation: Mutation rate = 0.1 21
Strength Pareto Evolutionary Algorithm II (SP A2) II (SPEA2) The relationship between two decision vectors : Dominance , indifference The relationship between two decision vectors : Dominance , indifference 22
Strength Pareto Evolutionary Algorithm II (SP A2) II (SPEA2) � SPEA2 P � SPEA2 Procedure: d 1.Assign a strength score to each solution. The score is equal to the number of solutions it dominates. 2.Get the raw fitness value of a solution by summing up the strength score of solutions which dominate it. 3.Get the density value by K-th nearest neighbor method (K=1). 4.Add the raw fitness value and density value to obtain the fitness value. A non- dominate solution has fitness value 0. 23
Strength Pareto Evolutionary Algorithm II (SP A2) II (SPEA2) 24
BLX- α Crossover BLX α Crossover It offers an opportunity that after crossover, the offspring’s genes come pp y p g g from a slightly larger range randomly selected between the two parents’ genes. � Procedure: � Procedure: h For two Parets: G1, G2, the i-th gene of offspring is define: i = = − ⋅ − ⋅ α α + ⋅ + ⋅ α α [ [ , ] ] h h Uniform g Uniform g I I g g I I i min max = 1 2 g Min g ( , g ) min i i 1 1 = g 1 1 2 2 ----- i-th gene of G1 g Max g g ( , ) i max i i 2 ----- i-th gene of G2 g = − i I g g α ----- An user define parameter max min 25
26 � Simulator Introduction � Simulation Results Simulation
Simulator Introduction 1. Main simulation framework: Hybrid simulator 2. Network topology: a simple network topology generator p gy p p gy g 3. Traffic model: fluid flow model and daily traffic model 4. Destination physical platform selection problem: 4 D i i h i l l f l i bl VRM_MOEA 5. When to trigger VRM : Reactive mechanism 27
Simulation Results Network 6N8L 11N14L Scheme Name Energy Energy Energy Energy Consumption / day Saving Consumption / day Saving No VRM 4057200.00 0.00% 7698240.00 0.00% Quick VRM 3223327.20 20.55% 6008048.00 21.96% VRM_MOEA(0 ,1) 3134507.20 22.74% 5457832.20 29.10% VRM_MOEA(0.2,0.8) 3125365.00 22.97% 5451220.20 29.97% VRM_MOEA(0.5,0.5) 3140139.80 22.60% 5471056.80 28.93% VRM_MOEA(0.8,0.2) 3160408.40 22.10% 5490892.40 28.67% VRM_MOEA(1 ,0) 3144768.20 22.49% 5517340.00 28.33% A Average values over 5 simulations with random seeds l 5 i l ti ith d d Quick VRM chooses the best solution in a randomly generated population of candidate solutions without any evolutionary process y y 28
Recommend
More recommend
