Title: Routing and Wavelength Assignment and an Analytical Framework for Mobile Agent based RWA in WDM Opti
1Routing and Wavelength Assignment and an
Analytical Framework for Mobile Agent based RWA
in WDM Optical Networks
2Outline
- Introduction
- Routing and Wavelength Assignment
- Mobile-Agent based Routing
- Analytical Framework
- Summary and Future Work
- References
3Introduction
- Backgrounds
- To support increase of Internet traffic
- WDM (wavelength division multiplexing) Optical
Network as high-speed transport networks - WDM concept
4Wavelength Routed Optical Network
?1 ?2
Lightpath
Optical Cross-Connect
5Routing and Wavelength Assignment
- Problem statement
- Given
- A network topology
- Set of connection requests to be established.
- Required
- To determine lightpath for each connection
- Physical route
- Assigned wavelength
- Constraints
- Wavelength continuity constraint A lightpath
must use the same wavelength on all the links
along its path from source to destination
(loosened by wavelength conversion) - Distinct wavelength (capacity) constraint All
lightpaths using the same link (fiber) must be
allocated distinct wavelengths.
6Routing and Wavelength Assignment
- Static RWA
- A set of connection requirements is given
- Remains for a relatively long period of time
- Mixed-integer linear program NP-hard
- Dynamic RWA
- A sequence of lightpath requests arrives in some
random fashion. - Algorithms must be executed in real time.
- Performances of algorithms are measured in terms
of blocking probability.
7Static RWA wavelength-continuity constraint
- Integer linear programILP
- Minimizing the flow in each link
- Minimizing the number of wavelength too
- Maximizing the number of established connection
too
8Static RWA with wavelength conversion
- Minimizing the flow in each linkILP
- Full convertible network
- Not needed and resource waste
- Alternate
- Sparse location of wavelength converter
- Sharing of converters
- Limited-range wavelength conversion
9Routing(1/3)
- Fixed Routing
- Fixed-Alternate Routing
Shortest path only
Routing table
10Routing(2/3)
- Adaptive routing
- More flexible
- Ex adaptive shortest-cost-path routing
- Wavelength-converted networks
- Continuous routing table update
- EXleast-congested path routing
- Similar with Fixed-Alternate Routing
11Routing(3/3)
- ILP formulation for static lightpath
- NP-complete
- Approximation
- reducing search space
- randomized rounding
- Fault-Tolerant routing
- Using backup lightpath
- In fixed-alternate routing
- Choosing alternate path(disjointed)
- In adaptive routing
- Dynamically established
12Static wavelength-assignment
- Sequential Graph Coloring
13wavelength-assignment Heuristics
- Random Wavelength Assignment
- Randomly chosen
- First-Fit
- All wavelengths are numbered
- Searching from lower number
- Least-Used(LU)
- Select the wavelength that is least used
- Most-Used(MU)
- Select the wavelength that is most used
- Outperform LU FF
14wavelength-assignment Heuristics
- Min-Product
- Multi-fiber networks
- Same with FF in single fiber networks
- Less than multi-fiber version FF
- Select wavelength that minimize
- Least-Loaded(LL)
- Multi-fiber networks
- Select the wavelength that the largest residual
capacity on the most loaded link along path P - Outperform MU FF
15wavelength-assignment Heuristics
- MAX-SUM
- Maximizing remaining path capacities after
establishment - Traffic matrix is known in advance
- Number of fibers on which wavelength j is unused
on link I - Wavelength j capacity is most congested link
value - Path capacity is sum of all wavelength capacity
- Select wavelength j which maximize the
16wavelength-assignment Heuristics
- Relative Capacity Loss(RCL)
- Select the wavelength j that minimize the
relative capacity loss - RCL is better than MAX-SUM in most cases
17wavelength-assignment Heuristics
- Protection method for long lightpath
- Whether or not the connection request can be
assigned - Protect the multiple fiber link connection
- Wavelength Reservation(Rsv)
- Reservation scheme
- Protecting Threshold(Thr)
- Single-hop connection is assigned wavelength
- Only if the number of idle wavelength is at or
above the given threshold
18Simulation
19Simulation
20Mobile Agent based Routing
- Currently in the Internet
- OSPF, BGP
- Based on shortest path algorithm
- Requirements
- One or more paths between end nodes
- Fast/immediate re-establishment of paths after
failures - Optimal resource usage Adaptation to topology
traffic load - Inherent robustness and stability
21Mobile Agent based Routing
- Ant Routing
- Small simple mobile co-operating agents (ants)
- Indirect communication
- One specie per task/policy
- Collective problem solving
- Applied to Internet Routing, Ad hoc Routing, RWA
B
S
D
A
22Mobile Agent based Routing
- Ant based RWA
- An ant-based approach for dynamic RWA in optical
WDM networks, Son Hong Ngo, X. Jiang, S.
Horiguchi, Photonic Network Communications 2006.
23Mobile Agent Based Routing
- Analysis on Mobile Agent-based Routing
- Analysis on a Mobile Agent-Based Algorithm for
Network Routing and Management, John Sum, H.
Shen, C. Leung, G. Young, IEEE Transactions on
Parallel and Distributed Systems March 2003. - Agents Population Analysis
- New Analysis on Mobile Agents based Network
Routing, Wenyu Qu, H. Shen, J. Sum, Elsevier
Applied Soft Computing 2005. - Success Probability Analysis
24Analytical Framework
- Analysis on Mobile Agent based Routing and
Wavelength Assignment - Blocking Probability
- Agent Population
- Differences from previous works
- Considered Wavelength Assignment in WDM Networks
- Blocking Probability Analysis
25Analytical Framework
- Mobile Agent Movement
- When a request is made, some agents created and
set out to find the destination. - Jump to random neighbor.
- Do not move to visited node.
- Do not move to full link.
- Analysis Case
- Wavelength Conversion
- Convertible
- Non-convertible
- Wavelength Assignment Heuristics
- Random
- First Fit
26Analytical Framework
- Network Model
- N number of nodes
- L number of links
- W number of wavelength in a link
- k agents per a request
- m request per unit time in a node
- T request duration
- D average distance (hop count) between node
pairs
27Analytical Framework
- Expected number of connection
- Total link-wavelength used
- Total link-wavelength
- Prob a link wavelength is used
- Prob a link is full
28Analytical Framework
- Blocking Probability Convertible
- Prob an agent can jump from source
- Prob first jump find destination
- Prob second jump find destination
- Prob i-th jump find destination
29Analytical Framework
- Blocking Probability Convertible
- Blocking Probability Prob every agent fail in
d jumps
30Analytical Framework
- Blocking Probability Non-Convertible, Random WA
- Consider wavelength continuity constraint
- Prob a wavelength is available in i-hop path
- Prob available wavelength exists in i-hop path
w1
w2
w3
w4
31Analytical Framework
- Blocking Probability Non-Convertible, Random WA
- Prob i-th jump find destination and available
wavelength exists - Blocking Probability
32Analytical Framework
- Blocking Probability Non-Convertible, First Fit
- Wavelength continuity constraints
- Number of connection in a link
- Number of cross connections in a lightpath
- Proba connection use wavelength i PW(i)
33Analytical Framework
- Blocking Probability Non-Convertible, First Fit
- PW(i) can be calculated recursively
- Number of cross connections in an i-hop lightpath
- Prob available wavelength exists in i-hop
lightpath can be calculated in case by case
manner
34Analytical Framework
- Agent Population
- When j and i are neighbors
- Expected number of agents running in each host is
less than or equal to - Number of agents propagating through any link is
less than or equal to
35Summary and Future Work
- Routing and Wavelength Assignment
- Static RWA
- Dynamic RWA
- Mobile agent based RWA
- Dynamic
- Distributed
- Analysis on Mobile Agent based RWA
- Blocking Probability
- Agent Population
- Future work
- More Intelligent Agent
36References
- R. Ramaswami and K. N. Sivarajan, Routing and
Wavelength Assignment in All-Optical Networks,
IEEE/ACM Transactions on Networking, vol. 3, no.
5, pp. 489-500, Oct. 1995. - S. Subramaniam and R. A. Barry, Wavelength
Assignment in Fixed Routing WDM Networks, Proc.,
ICC 97, Montreal, Canada, vol. 1, pp. 406-410,
June 1997. - E. Karasan and E. Ayanoglu, Effects of
Wavelength Routing and Selection Algorithms on
Wavelength Conversion Gain in WDM Optical
Networks, IEEE/ACM Transactions on Networking,
vol. 6, no. 2, pp. 186-196, April 1998. - X. Zhang and C. Qiao, Wavelength Assignment for
Dynamic Traffic in Multi-fiber WDM Networks,
Proc., 7th International Conference on Computer
Communications and Networks, Lafayette, LA, pp.
479-485, Oct. 1998. - L. Li and A. K. Somani, Dynamic Wavelength
Routing Using Congestion and Neighborhood
Information, IEEE/ACM Transactions on
Networking, 1999. - Zang, H., Jue, J.P., Mukherjee, B. A review of
routing and wavelength assignment approaches for
wavelength-routed optical WDM networks. Opt.
Netw. Mag. 2000. - John Sum, H. Shen, C. Leung, G. Young Analysis
on a Mobile Agent-Based Algorithm for Network
Routing and Management, IEEE Transactions on
Parallel and Distributed Systems March 2003. - Wenyu Qu, H. Shen, J. Sum New Analysis on Mobile
Agents based Network Routing, Elsevier Applied
Soft Computing 2005. - Son Hong Ngo, Xiaohong Jiang, Susumu Horiguchi
An ant-based approach for dynamic RWA in
opticalWDM networks, Photonic Network
Communications 2006.