1. Advisor: Professor Yeong-Sung Lin Student: Yeong-Cheng Tzeng (曾勇誠)
Fair Intra-TAP Temporal and Throughput (Delay) Routing Algorithms in Mesh Networks
Advisor: Professor Yeong-Sung Lin
Student: Yeong-Cheng Tzeng (曾勇誠)

2. Outline Introduction Problem Description Related Work Motivation
Notation
Problem Formulation
Relaxation
Subproblem and Solution Approach

3. 1. Introduction WiFi networks have become increasingly popular:
Clients need to be in the immediate vicinity of the Internet HS
Have to deploy hot spots at well-chosen locations
Wireless Mesh Networks:
An extension of WiFi
The TAPs are not directly
connected to the Internet

4. 1. Introduction (cont’d)
If the protocol is poorly designed
Severe unfairness (starvation)
Low bandwidth utilization
Issues:
Fairness
Backhaul deployment

5. 1.1. Related Work Mesh Networks I. F. Akyildiz, X. Wang, and W. Wang.
Wireless Mesh Networks: A Survey. Computer Networks Journal (Elsevier), 47(4), 2005.
Present a survey on recent advances and open research issues in WMNs
Point out an important research topic:
Revise the design of MAC protocols based on TDMA or CDMA

6. 1.1. Related Work (cont’d) Fairness in Mesh Networks
V. Gambiroza, B. Sadeghi, and E. Knightly,
“End-to-End Performance and Fairness in Multihop Wireless Backhaul Networks" in Proceedings of MobiCom 2004.
Chain topology
Temporal fairness (max min channel access time)

7. 1.2. Motivation
To propose a fair resource allocation algorithm that achieves:
Optimize backhaul deployment
Minimize maximum end-to-end delay
Assign bandwidth to the links in the WMN
Min max delay fairness

8. 2. Problem Description

9. 2.1. Problem Description Given
The network topology includes TAP set and link set.
The capacity of all links.
The set of all candidate backhauls.
The total budget to build up the backhauls.
The cost of building backhaul.
The set of all candidate paths from each TAP to backhaul.
Data rate of each TAPs required to be transmitted.

10. 2.1. Problem Description Objective: Subject to: To determine:
To minimize the maximum end-to-end transmission delay.
Subject to:
The backhaul assignment with budget constraints will be considered.
Routing constraints will be considered.
Tree constraints will be considered.
The TAPs must route to the “to-be-determined” backhaul with delay time and capacity constraints.
To determine:
Backhaul deployment
Routing assignment
Bandwidth allocation

11. 2.2. Notation Given Parameters Notation Description V
The set of TAPs which is also the set of candidate backhauls, where v  V. B
The set of candidate backhauls, where b  B. θ
The budget to build up the backhaul. cb
The cost to build the backhaul b.
Pbs
The set of paths from original TAP s to destination TAP b or vice versa, where p  Pbs.
puv
Indication function, which denote link uv on the path p.
Cuv
The capacity of link uv. as
Required data rate of TAP s. ε
Estimated error.

12. 2.2. Notation (cont’d) Decision Variables Notation Description b
1 if TAP b is selected to be a backhaul; otherwise 0.
zbs
1 if TAP s connects to wired network via backhaul b; otherwise 0. xp
1 if path p from TAP, s, to TAP b is selected; otherwise 0.
yuv
1 if link uv is on the tree; otherwise 0.
bsuv
Bandwidth allocation to TAP s on link uv.
tsuv
Data transmission delay of TAP s on link uv. T
Maximum end-to-end delay from each TAP to associated backhaul.

13. 2.3. Problem Formulation
Objective function
subject to

14. 2.3. Problem Formulation (cont’d)

15. 2.3. Problem Formulation (cont’d)

16. 2.4. Relaxation
We can transform (IP) into LR problem where Constraints (1), (2), (3), (6), (8), (9), (10), (12), (13), (14), and (16) are relaxed.
Optimization Problem (LR):

17. 2.4. Relaxation (cont’d)
subject to

18. 2.4. Relaxation (cont’d)

19. 2.5. Subproblem and Solution Approach
Subproblem1 (related to decision variableηb)
subject to
solution

20. 2.5. Subproblem and Solution Approach (cont’d)
Subproblem2 (related to decision variable zbs)
subject to
solution
1. Decompose into |V| independent subproblems
2. Calculate the coefficient of zbs
3. For all coefficients in ascending, we assign the first zbs=1 , others 0.

21. 2.5. Subproblem and Solution Approach (cont’d)
Subproblem3 (related to decision variable xp)
subject to
solution
(SUB 3) can be decomposed into |V| independent shortest path problems

22. 2.5. Subproblem and Solution Approach (cont’d)
Subproblem4 (related to decision variable yuv)
subject to
solution

23. 2.5. Subproblem and Solution Approach (cont’d)
Subproblem5 (related to decision variable bsuv)
subject to

24. 2.5. Subproblem and Solution Approach (cont’d)
Subproblem6 (related to decision variable tsuv)
subject to

25. 2.5. Subproblem and Solution Approach (cont’d)
Subproblem7 (related to decision variable T)
subject to

26. The End