Traffic Grooming in WDM Ring Networks Presented by: Eshcar Hilel.

Slides:



Advertisements
Similar presentations
Optimization Problems in Optical Networks. Wavelength Division Multiplexing (WDM) Directed: Symmetric: Undirected: Optic Fiber.
Advertisements

Optical networks: Basics of WDM
Data Communications and Networking
Optimization Problems in Optical Networks. Wavelength Division Multiplexing (WDM) Directed: Symmetric: Optic Fiber.
Wavelength Division Multiplexing (WDM) in Optical Networks: Modeled as a Graph Coloring Problem By Joshua Schoenly.
Lecture: 4 WDM Networks Design & Operation
1 EL736 Communications Networks II: Design and Algorithms Class3: Network Design Modeling Yong Liu 09/19/2007.
Apoorv Nayak Prathyusha Dasari Traffic Grooming. Agenda  Improved approaches for cost effective traffic grooming in WDM ring networks  Motivation 
1 Wide-Sense Nonblocking Multicast in a Class of Regular Optical Networks From: C. Zhou and Y. Yang, IEEE Transactions on communications, vol. 50, No.
EMIS 8373: Integer Programming Valid Inequalities updated 4April 2011.
Optical communications & networking - an Overview
Optic fiber Electronic switch the fiber serves as a transmission medium Optical networks - 1 st generation 1. Optical networks – basic notions.
9/22/2003Kevin Su Traffic Grooming in WDM Networks Kevin Su University of Texas at San Antonio.
Chapter 4 Circuit-Switching Networks
Synchronous Optical Network (SONET) Fall Semester, School of Computer Science & Engineering, Seoul National University. Professor Yanghee Choi Student.
Sonet. Definition Short for Synchronous Optical Network, a standard for connecting fiber-optic transmission systems. SONET defines interface standards.
Chapter 10 Wide Area Networks. Contents The need for Wide area networks (WANs) Point-to-point approaches Statistical multiplexing, TDM, FDM approaches.
Traffic Grooming in WDM Networks Wang Yao. WDM Technology increases the transmission capacity of optical fibers allows simultaneously transmission of.
1 Vertex Cover Problem Given a graph G=(V, E), find V' ⊆ V such that for each edge (u, v) ∈ E at least one of u and v belongs to V’ and |V’| is minimized.
EE 4272Spring, 2003 Chapter 9: Circuit Switching Switching Networks Circuit-Switching Networks Circuit-Switching Concept  Space-Division Switching  Time-Division.
CS335 Networking & Network Administration Wednesday, April 14, 2010.
Distributed Combinatorial Optimization
1 Introduction to Optical Networks. 2 Telecommunications Network Architecture.
1 Distributed Computing Optical networks: switching cost and traffic grooming Shmuel Zaks ©
Multiplexing.
FiWi Integrated Fiber-Wireless Access Networks
1 Protection Mechanisms for Optical WDM Networks based on Wavelength Converter Multiplexing and Backup Path Relocation Techniques Sunil Gowda and Krishna.
TTM1 – 2013: Core networks and Optical Circuit Switching (OCS)
9 1 SIT  Today, there is a general consensus that in near future wide area networks (WAN)(such as, a nation wide backbone network) will be based on.
Intorduction to Lumentis
1 William Stallings Data and Computer Communications 7 th Edition Chapter 8 Multiplexing.
Optimization of Wavelength Assignment for QoS Multicast in WDM Networks Xiao-Hua Jia, Ding-Zhu Du, Xiao-Dong Hu, Man-Kei Lee, and Jun Gu, IEEE TRANSACTIONS.
November 18, Traffic Grooming in Optical WDM Networks Presented by : Md. Shamsul Wazed University of Windsor.
1 Chapter 5 Multiplexing : Sharing a Medium Data Communications and Computer Networks: A Business User’s Approach.
Wavelength Assignment in Waveband Switching Networks with Wavelength Conversion Xiaojun Cao; Chunming Qiao; Anand, V. Jikai LI GLOBECOM '04. IEEE Volume.
1 Optical Burst Switching (OBS). 2 Optical Internet IP runs over an all-optical WDM layer –OXCs interconnected by fiber links –IP routers attached to.
Maximization of Network Survivability against Intelligent and Malicious Attacks (Cont’d) Presented by Erion Lin.
Network Survivability Against Region Failure Signal Processing, Communications and Computing (ICSPCC), 2011 IEEE International Conference on Ran Li, Xiaoliang.
CSC/ECE 778 Optical Networking Rudra Dutta Fall, 2007.
Το Πρόβλημα Routing and Path Coloring και οι εφαρμογές του σε πλήρως οπτικά δίκτυα Ευχαριστίες: οι διαφάνειες αυτές βασίστηκαν εν μέρει στην παρουσίαση.
Terminal Handling & Polling
1 Traffic Engineering in Multi-Granularity, Heterogeneous, WDM Optical Mesh Networks Through Dynamic Traffic Grooming Keyao Zhu, Hongyue Zhu, and Biswanath.
Multi-layered Optical Network Security
Optimization of Wavelength Assignment for QoS Multicast in WDM Networks Xiao-Hua Jia, Ding-Zhu Du, Xiao-Dong Hu, Man-Kei Lee, and Jun Gu, IEEE TRANSACTIONS.
1 Multicasting in a Class of Multicast-Capable WDM Networks From: Y. Wang and Y. Yang, Journal of Lightwave Technology, vol. 20, No. 3, Mar From:
1 Distributed Computing Optical networks: switching cost and traffic grooming Shmuel Zaks ©
1 11 Channel Assignment for Maximum Throughput in Multi-Channel Access Point Networks Xiang Luo, Raj Iyengar and Koushik Kar Rensselaer Polytechnic Institute.
NP-Complete Problems. Running Time v.s. Input Size Concern with problems whose complexity may be described by exponential functions. Tractable problems.
Communications Equipment Communication Equipments: –Equipments that are, Not a direct part of the network (like node, client, terminal, server etc.) but,
Data Comm. & Networks Instructor: Ibrahim Tariq Lecture 3.
Lecture 1 Outline Statistical Multiplexing Inter-Process Communication.
Survivable Traffic Grooming with Differentiated End-to-End Availability Guarantees in WDM Mesh Networks Proceedings of the 13th IEEE Workshop on Local.
INTRODUCTION. Homogeneous Networks A homogeneous cellular system is a network of base stations in a planned layout and a collection of user terminals,
10/6/2003Kevin Su Traffic Grooming for Survivable WDM Networks – Shared Protection Kevin Su University of Texas at San Antonio.
Optical Networking University of Southern Queensland.
Unit III Bandwidth Utilization: Multiplexing and Spectrum Spreading In practical life the bandwidth available of links is limited. The proper utilization.
Traffic grooming in WDM Networks Dynamic Traffic Grooming in WDM Mesh Networks Using a Novel Graph Model by Hongyue Zhu, Hui Zang, Keyao Zhu, and Biswanath.
1 CS612 Algorithms for Electronic Design Automation CS 612 – Lecture 8 Lecture 8 Network Flow Based Modeling Mustafa Ozdal Computer Engineering Department,
QoS-specified Traffic Grooming Algorithm in WDM Mesh Networks Bing Xiang, Hongfang Yu, Sheng Wang, Lemin Li Communications, Circuits and Systems, 2004.
Introduction to Communication Lecture (07) 1. Bandwidth utilization Bandwidth utilization is the wise use of available bandwidth to achieve specific goals.
1 Chapter 6 Reformulation-Linearization Technique and Applications.
1 Chapter 5 Branch-and-bound Framework and Its Applications.
Errol Lloyd Design and Analysis of Algorithms Approximation Algorithms for NP-complete Problems Bin Packing Networks.
Chapter 2 PHYSICAL LAYER.
4.3 Multiplexing Outlines FDM TDM.
The minimum cost flow problem
Presented by Radha Gummuluri ECE-E 641 Fiber Optic Communications
The University of Adelaide, School of Computer Science
The University of Adelaide, School of Computer Science
Optical communications & networking - an Overview
Presentation transcript:

Traffic Grooming in WDM Ring Networks Presented by: Eshcar Hilel

Distributed Algorithms, Spring Introduction Optical Networks - A new generation of networks using optical fiber transmission – Excellent medium, high BW, low error … SONET ring - synchronous optical network, currently the most widely deployed optical network infrastructure WDM Technology – wavelength-division multiplexing

Distributed Algorithms, Spring Introduction – SONET Ring SADM - SONET add/drop multiplexers can aggregate lower-rate signals into a single high- rate stream SONET ring use one fiber pair (or two for protection) to connect SADMs in the source and destination nodes

Distributed Algorithms, Spring Introduction – WDM Increases the transmission capacity of optical fibers Allows simultaneously transmission of multiple wavelengths (channels) within a single fiber One wavelength may carry Internet traffic; another may carry voice or video

Distributed Algorithms, Spring Introduction – SONET over WDM Multiple SONET rings can be supported on a single fiber pair by using multiple wavelengths The networks are limited by the processing capability of electronic switches, routers and multiplexers (not by transmission bandwidth) New aim: overcoming the electronic bottleneck by providing optical bypass

Distributed Algorithms, Spring Introduction – Optical bypass WADM - WDM Add/Drop Multiplexer allows to drop (or add) only the wavelength that carries the traffic destined to (or originated from) the node The dropped wavelength is electronically processed at the node All the other wavelengths optically bypass the node

Distributed Algorithms, Spring Introduction – WADM More optical switches may be added to support more add-drop wavelengths

Distributed Algorithms, Spring Introduction – Traffic Grooming Every wavelength needs a SADM only at nodes where it is ended Traffic typically require only a small fraction of the wavelength Traffic grooming can be used in such a way that all of the traffic to and from the node is carried on minimum number of wavelength

Distributed Algorithms, Spring Topics of Discussion Traffic Grooming - Understanding the Problem Single Exit Node Network – NP-complete problem – Special case: uniform traffic – Special case: minimum number of wavelengths All-To-All Uniform Traffic Network

Traffic Grooming Understanding the problem

Distributed Algorithms, Spring What ’ s the Problem? Unidirectional (clockwise) WDM ring N nodes: 1,2, …,N c – grooming factor r ij - number of low rate circuits from node i to node j Objective: minimize total number of SADMs

Distributed Algorithms, Spring Illustration Unidirectional ring network: N = 4 6 pairs of nodes r ij = 8: 8 OC-3 circuits between each pair c = 16: each wavelength supports an OC-48 ring Total load: 6x8 OC-3 = 3 OC-48, requires 3 wavelengths

Distributed Algorithms, Spring Illustration Traffic assignment: 1: 1↔2, 3↔4 2: 1↔3, 2↔4 3: 1↔4, 2↔3 Total: 12 SADMs

Distributed Algorithms, Spring Illustration Traffic assignment: 1: 1↔2, 1↔3 2: 2↔3, 2↔4 3: 1↔4, 3↔4 Total: 9 SADMs

Distributed Algorithms, Spring Goal – Traffic Grooming Tradeoff between efficient use of fibers and the cost of electronic equipment When no limitation on wavelengths – dedicated wavelength per connection, no multiplexing Else design traffic grooming algorithms to – Minimize number of electronics (SADMs) – Minimize number of wavelengths (efficient use of wavelengths)

Single Exit Node Network E. Modiano, A. Chio, “ Traffic Grooming Algorithms for Reducing Electronic Multiplexing Costs in WDM Ring Networks ”

Distributed Algorithms, Spring Telephone company’s central office Computational Complexity Unidirectional ring All the traffic on the ring is destined to a single exit node Denote the exit node 0 r ij > 0, for j = 0 and i = 1, …,N Note: maximum load L max =  i=1..N r i0 and minimum wavelengths W min =  L max / c 

Distributed Algorithms, Spring Computational Complexity Assume w.l.o.g. r i0 <c for all i Else fill  r i0 /c  wavelengths with  r i0 /c  *c low rate circuits, and groom the remaining (<c) circuits Theorem: The traffic grooming problem is NP-complete

Distributed Algorithms, Spring Computational Complexity Bin packing problem: What is the least number of bins (containers of fixed volume) needed to hold a set of objects (of different volumes)? The bin packing problem is an NP-complete problem.

Distributed Algorithms, Spring Computational Complexity Claim: There exist an optimal solution such that no traffic from a node is split onto two rings Proof: – Consider assignment where the traffic of some nodes is split onto 2 or more rings – Each such node have at least 2 SADMs – Accommodate the traffic on a separate wavelength – Requires at most 2 SADMs

Distributed Algorithms, Spring Computational Complexity Theorem Proof: – For any optimal solution with no split traffic: regular nodes - N SADMs; exit node - k SADMs, where k is the number of SONET rings – Problem reduced to minimizing total number of rings – Achieved by combining traffic from multiple nodes onto single ring (wavelength) – This is basically the Bin Packing problem! QED

Distributed Algorithms, Spring Special Case: Uniform Traffic r i0 = r Optimal solution does not require split traffic May groom traffic from at most  c/r  nodes on one SONET ring Number of wavelengths: W =  N/  c/r   Hence, minimum SADMs M min = N + W Not the minimum number of wavelengths!

Distributed Algorithms, Spring Special Case: Minimum Number of Wavelengths Traffic from nodes may have to be split onto multiple rings, S - total number of traffic splits Additional SADM per split Hence, #SADMs M = N + W min + S, where W min =  r*N /c  Objective: minimize the total number of splits

Distributed Algorithms, Spring Special Case: Minimum Number of Wavelengths Maximum load for ring with no split L ns =  c/r  *r W ns Maximum number of rings with no split Remaining rings contain at most c circuits: W ns * L ns + (W min - W ns )*c >= L max W ns = min{W min,  (c* W min – L max ) / (c-L ns )  }

Distributed Algorithms, Spring Iterative Algorithm Initialization: c 0 = c, N 0 = N, r 0 = r, W 0 = W 0 min Steps of loop i: – If W i ns = W i min then accommodate the remaining traffic without splitting - terminate – Fill W i rings with unsplit traffic from  c i /r i  nodes – Remaining capacity is c i+1 = c i -  c i /r i  *r i – N i+1 = N i -  c i /r i  *W i nodes needs to be assigned

Distributed Algorithms, Spring Steps of loop i (cont): – N i+1 = N i -  c i /r i  *W i nodes needs to be assigned – Fill remaining capacity c i+1 by traffic from N i+1 nodes – Remaining traffic becomes r i+1 = r i – c i+1 – W i+1 = W i – N i+1 – Continue to loop i+1 N i+1 < W i Iterative Algorithm (cont)

All-To-All Uniform Traffic Network J.C. Bermond, D. Coudert, “ Traffic Grooming in Unidirectional WDM Ring Networks using Design Theory ”

Distributed Algorithms, Spring All-To-All Uniform Traffic We show the problem can be formulated in terms of graph partition into sub-graphs: – at most c edges and per sub-graph – minimize total number of vertices

Distributed Algorithms, Spring Traffic Grooming: Reformulating the Problem N nodes of unidirectional ring C N R = N(N-1)/2 circles c – grooming factor K N - Complete graph on N vertices B λ denote a sub-graph of K N V(B λ ) (resp E(B λ )) denote its vertex (resp edge) set

Distributed Algorithms, Spring Traffic Grooming: Reformulating the Problem B λ correspond to a wavelength An edge of B λ correspond to a circle in the ring B λ is viewed as a set of circles packed in a wavelength |E(B λ )| <= c V(B λ ) correspond to the number of SADMs A(c,N) denotes total number of SADMs

Distributed Algorithms, Spring Traffic Grooming: Reformulating the Problem Input:N and c Output:partition of K N into sub-graphs B λ, λ = 1, …,W, such that |E(B λ )| <= c Objective: minimize ∑ 1<=λ<=W |V(B λ )|

Distributed Algorithms, Spring Lower Bound ρ(B λ ) = |E(B λ )|/|V(B λ )| is the sub-graph ratio ρ(m) maximum ratio of sub-graph with m edges ρ max (c) = max m<=c ρ(m)

Distributed Algorithms, Spring Lower Bound Theorem: any grooming of R circles with grooming factor c needs at least R/ρ max (c) SADMs Proof: R = ∑ W λ=1 |E(B λ )| <= ρ max (c)* ∑ W λ=1 |V(B λ )| Thus we have the lower bound: A(c,N) >= N(N-1) / ρ max (c)*2

Distributed Algorithms, Spring Lower Bound We compute ρ max (c) Theorem: If k(k-1)/2<=c<=(k+1)(k-1)/2, then ρ max (c)=(k-1)/2 If (k+1)(k-1)/2<=c<=(k+1)k/2, then ρ max (c)=c/k+1 Proof: on board

Distributed Algorithms, Spring Lower Bound Note: these sub-graphs do not have necessarily exactly c edges and so the minimum is not necessarily attained for W = W min Example: N=13 and c=7

Distributed Algorithms, Spring Discussion My opinion of the subject Your opinion of the subject (and presentation … ) That ’ s all folks!

Distributed Algorithms, Spring References J.C. Bermond, D. Coudert, “ Traffic Grooming in Unidirectional WDM Ring Networks using Design Theory ”, IEEE International Conference on Communications, May, 2003 E. Modiano, A. Chio, “ Traffic Grooming Algorithms for Reducing Electronic Multiplexing Costs in WDM Ring Networks ”, IEEE J. Lightwave Tech., Jan vol. 18(1)

Distributed Algorithms, Spring References E. Modiano, P. Lin, “ Traffic Grooming in WDM Networks ”, IEEE Communication Magazine, July 2001.