Successfully reported this slideshow.
We use your LinkedIn profile and activity data to personalize ads and to show you more relevant ads. You can change your ad preferences anytime.

(Slides) A Method for Distributed Computaion of Semi-Optimal Multicast Tree in MANET

2,265 views

Published on

Takashima, E., Murata, Y., Shibata, N., Yasumoto, K. and Ito, M.: A Method for Distributed Computaion of Semi-Optimal Multicast Tree in MANET, IEEE Wireless Communications and Networking Conference (WCNC 2007), pp. 2570-2575, DOI:10.1109/WCNC.2007.478 (March 2007).

http://ito-lab.naist.jp/themes/pdffiles/070314.eiichi-t.wcnc2007.pdf

In this paper, we propose a new method to construct
a semi-optimal QoS-aware multicast tree on MANET using
distributed computation of the tree based on Genetic Algorithm
(GA). This tree is sub-optimal for a given objective (e.g.,
communication stability and power consumption), and satisfies
given QoS constraints for bandwidth and delay. In order to
increase scalability, our proposed method first divides the whole
MANET to multiple clusters, and computes a tree for each
cluster and a tree connecting all clusters. Each tree is computed
by GA in some nodes selected in the corresponding cluster.
Through experiments using network simulator, we confirmed that
our method outperforms existing on-demand multicast routing
protocol in some useful objectives.

Published in: Technology
  • Be the first to comment

  • Be the first to like this

(Slides) A Method for Distributed Computaion of Semi-Optimal Multicast Tree in MANET

  1. 1. A Method for Distributed Computation of Semi-Optimal Multicast Tree in MANET Eiichi Takashima, Yoshihiro Murata, Naoki Shibata*, Keiichi Yasumoto, and Minoru Ito. Nara Institute of Science and Technology, *Shiga University
  2. 2. <ul><li>Background </li></ul><ul><li>Proposed method </li></ul><ul><li>Evaluation experiment </li></ul><ul><li>Conclusion </li></ul>Outline of this presentation
  3. 3. Background <ul><li>Video streaming - one of the most important application in mobile ad-hoc network (MANET) </li></ul><ul><li>Objective : Delivering video to many nodes in MANET </li></ul><ul><ul><li>Using a multicast tree </li></ul></ul><ul><ul><li>Satisfying QoS constraints </li></ul></ul><ul><ul><ul><li>Bandwidth </li></ul></ul></ul><ul><ul><ul><li>Delay </li></ul></ul></ul><ul><ul><li>Optimized for any given objective </li></ul></ul><ul><ul><ul><li>Power consumption (Mobile nodes are operated on battery) </li></ul></ul></ul><ul><ul><ul><li>Maximizing number of receiver nodes </li></ul></ul></ul>
  4. 4. Background <ul><li>Optimizing multicast tree on MANET </li></ul><ul><ul><li>A hard task - an NP-hard problem </li></ul></ul><ul><ul><li>Dynamic network topology </li></ul></ul><ul><ul><li>Limited capabilities of mobile terminals </li></ul></ul><ul><ul><ul><li>Computation </li></ul></ul></ul><ul><ul><ul><li>Communication </li></ul></ul></ul>
  5. 5. Existing studies <ul><li>P. Sinha, et al. [1] </li></ul><ul><ul><li>Distributed algorithm </li></ul></ul><ul><ul><li>Good scalability </li></ul></ul><ul><ul><li>No handling of multiple QoS constraints </li></ul></ul><ul><ul><li>No optimization for a particular objective </li></ul></ul><ul><li>Li Layuan, et al.[2] </li></ul><ul><ul><li>Centralized algorithm </li></ul></ul><ul><ul><li>Optimizes any objective with multiple QoS constraints </li></ul></ul><ul><ul><li>Poor scalability </li></ul></ul><ul><ul><ul><li>Cost of gathering topology information </li></ul></ul></ul><ul><ul><ul><li>Centralized computation </li></ul></ul></ul>[2] Li Layuan and Li Chunlin, &quot;QoS Multicast Routing in Networks with Uncertain Parameters&quot;, APWeb, (2003). [1] P. Sinha and R. Sivakumar and V. Bharghavan, &quot;MCEDAR: Multicast core extraction distributed ad-hoc routing&quot;, WCNC(1999),
  6. 6. <ul><li>Background </li></ul><ul><li>Proposed method </li></ul><ul><li>Evaluation experiment </li></ul><ul><li>Conclusion </li></ul>Outline of this presentation
  7. 7. Goal of this research <ul><li>Constructing multicast tree </li></ul><ul><ul><li>Satisfying all given QoS constraints </li></ul></ul><ul><ul><li>Optimizing a given objective </li></ul></ul><ul><ul><ul><li>total power consumption </li></ul></ul></ul><ul><ul><ul><li>tree stability </li></ul></ul></ul><ul><li>Good scalability </li></ul><ul><ul><li>Distributed computation </li></ul></ul>
  8. 8. Our Approach <ul><li>We use GA (Genetic Algorithm) to construct semi-optimal multicast tree </li></ul><ul><li>To realize distributed computation </li></ul><ul><ul><li>we compute multicast tree on several nodes in parallel using GA </li></ul></ul><ul><ul><li>Each node solves a sub-tree for the whole multicast tree </li></ul></ul><ul><li>We divide MANET into multiple clusters </li></ul><ul><li>Advantage of using GA </li></ul><ul><ul><li>Quick computation using results of previous computation </li></ul></ul><ul><ul><li>Especially when topology change is small </li></ul></ul>
  9. 9. Hierarchical computation <ul><li>Two tier computation : “ local trees ” and “ global tree ” </li></ul><ul><ul><li>A local tree connects nodes in a cluster </li></ul></ul><ul><ul><li>The global tree connects clusters </li></ul></ul>cluster Global Tree node Local tree
  10. 10. Target Environment & Assumption <ul><li>Service </li></ul><ul><ul><li>deliver small video (or audio ) data from a sender node to multiple receiver nodes in MANET </li></ul></ul><ul><ul><li>requirement : transmission rate B , tolerable end-to-end delay D </li></ul></ul><ul><li>MAC protocol of wireless communication </li></ul><ul><ul><li>IEEE 802.11 </li></ul></ul><ul><li>Mobile nodes </li></ul><ul><ul><li>move at speed of 4 Km/hour (pedestrian) </li></ul></ul><ul><ul><li>can measure available bandwidth and delay to neighboring nodes </li></ul></ul><ul><ul><li>can estimate approximate distances to neighboring nodes by strength of radio wave signals </li></ul></ul>
  11. 11. Problem Definition <ul><li>Input: </li></ul><ul><ul><li>topology info: G=(V,E ) , where V is set of nodes, E is set of links </li></ul></ul><ul><ul><li>sender node: s  V </li></ul></ul><ul><ul><li>receiver nodes: R={r 1 ,…r m }  V </li></ul></ul><ul><li>Output: </li></ul><ul><ul><li>Multicast tree: T=(V’,E’) , where V’  V, E’  E </li></ul></ul><ul><li>Constraints: </li></ul><ul><ul><li>each link e  E has available bandwidth no less than B </li></ul></ul><ul><ul><li>total delay of each path in T is no more than D </li></ul></ul><ul><li>Objective : </li></ul><ul><ul><li>maximize stability of T (links are connected for longer time) </li></ul></ul><ul><ul><li>maximize service availability (more nodes can receive video) </li></ul></ul><ul><ul><li>minimize total power consumption </li></ul></ul><ul><ul><li>etc </li></ul></ul>
  12. 12. Typical Objective Functions <ul><li>Our method solves problem for intra-cluster and inter-cluster separately  use different functions </li></ul><ul><li>Global Tree T’: maximize F G </li></ul><ul><ul><li>F G =  NumberOfReceivers ( T’) </li></ul></ul><ul><ul><li>  NumberOf DelayViolation(T’) </li></ul></ul><ul><ul><li>+  Stability (T’) </li></ul></ul><ul><li>Local Tree T’’: maximize F L </li></ul><ul><ul><li>F L = NumberOfReceivers (T’’) +   S tability(T’’) </li></ul></ul><ul><ul><li>*  are coefficients. </li></ul></ul>Term for power consumption can also be added service availability Tree stability service availability Tree stability
  13. 13. Procedure Phase1: Cluster division Cluster division Gathering topology info in each cluster Gathering topology info between clusters Computation of global tree Cluster re-division Computation of local tree Inter cluster e e e e e S Intra cluster Cluster head: responsible to local tree construction Top cluster head: responsible to global tree construction e e e e e S
  14. 14. Phase2: Gathering Local Topology Info Cluster division Gathering topology info in each cluster Gathering topology info between clusters Cluster re-division Inter cluster Intra cluster (1) Cluster head floods request msg in its cluster e e e e e S Computation of global tree Computation of local tree e e e e e S
  15. 15. Phase2: Gathering local topology Info Cluster Division Gathering topology info in each cluster Gathering topology info between clusters Cluster re-division Inter cluster Intra cluster (1) Cluster head floods request msg in its cluster (2) Each node received the message sends back a message with its ID and link state info including B/W and delay to neighboring nodes. e e e e e S Computation of global tree Computation of local tree e e e e e S
  16. 16. Phase3: Gathering global topology info Cluster Division Gathering topology info in each cluster Gathering topology info between clusters Cluster re-division Inter cluster e e e e e S (1) Each cluster head measures QoS info on paths to cluster heads of adjacent clusters. (2) Each cluster head sends the info to the top cluster head. Intra cluster Computation of global tree Computation of local tree e e e e e S
  17. 17. Phase4: Computation of global tree Inter cluster Intra cluster e e e e e S Cluster Division Gathering topology info in each cluster Gathering topology info between clusters Cluster re-division (1) Top cluster head (and some nodes) computes global tree by using island model GA. Computation of global tree Computation of local tree e e e e e S
  18. 18. Phase4: Computation of global tree Inter cluster Intra cluster Cluster Division Gathering topology info in each cluster Gathering topology info between clusters Cluster re-division (1) Top cluster head (and some nodes) computes global tree by using island model GA. (2) Information of global tree is sent to each cluster head in the tree. Computation of global tree Computation of local tree e e e e e S e e e e e S
  19. 19. Phase5: Computation of local tree Cluster Division Gathering topology info in each cluster Gathering topology info between clusters Cluster re-division Inter cluster e e e e e S Intra cluster Each cluster head computes local tree which can be grafted to global tree Computation of global tree Computation of local tree e e e e e S
  20. 20. Phase5: Computation of local tree Inter cluster Intra cluster Cluster Division Gathering topology info in each cluster Gathering topology info between clusters Cluster re-division The island model GA is used for computation of local tree Computation of global tree Computation of local tree e e e e e S e e e e e S
  21. 21. Phase5: Computation of local tree Cluster Division Gathering topology info in each cluster Gathering topology info between clusters Cluster re-division Inter cluster e e e e e S Intra cluster Computation of global tree Computation of local tree The info of local tree is sent to each node in the tree e e e e e S
  22. 22. Phase5: Computation of local tree Cluster Division Gathering topology info in each cluster Gathering topology info between clusters Cluster re-division Inter cluster e e e e e S Intra cluster Computation of global tree Computation of local tree The semi-optimal multicast tree has been constructed among nodes. e e e e e S
  23. 23. Phase6: Cluster re-division Cluster Division Gathering topology info in each cluster Gathering topology info between clusters Cluster re-division Inter cluster e e e e e S Intra cluster Computation of global tree Computation of local tree After a while, MANET is clustered again and procedure from phase2 is repeated to reflect change of topology. e e e e e S
  24. 24. <ul><li>Background </li></ul><ul><li>Proposed method </li></ul><ul><li>Evaluation </li></ul><ul><li>Conclusion </li></ul>Outline of this presentation
  25. 25. Evaluation <ul><li>Criteria </li></ul><ul><ul><li>Advantage of GA for computing multicast tree </li></ul></ul><ul><ul><li>Feasibility in practical environment </li></ul></ul><ul><ul><li>Superiority to existing method </li></ul></ul>
  26. 26. Advantage of the proposed algorithm <ul><li>Objective is to investigate </li></ul><ul><ul><li>scalability against number of nodes </li></ul></ul><ul><ul><li>efficiency of re-computation when topology changes </li></ul></ul><ul><li>Experimental Configuration </li></ul><ul><ul><li>Mobility model of nodes </li></ul></ul><ul><ul><ul><li>Random way point, 4 Km/hour </li></ul></ul></ul><ul><ul><li>PC (laptop) for executing algorithm </li></ul></ul><ul><ul><ul><li>CPU Intel(R) Pentium(R) M processor 1500MHz , Windows XP , cygwin 1.5.18 , gcc version 3.4.4. </li></ul></ul></ul>
  27. 27. Result of (re)computation time of tree <ul><li>Computation time : </li></ul><ul><ul><li>6 sec for 800 nodes </li></ul></ul><ul><ul><li>1 sec for 100 nodes </li></ul></ul><ul><li>Re-computation time </li></ul><ul><ul><li>shortened to 60% </li></ul></ul>Seconds ■ Computation time ― approximation of computation time ■ Re-computation time ― approximation of recomputation time Number of nodes sufficient
  28. 28. Feasibility in practical environment <ul><li>Evaluated the following points with 1000 nodes on 30 clusters (33 nodes per cluster) </li></ul><ul><ul><li>Computation cost </li></ul></ul><ul><ul><li>Required bandwidth for MANET </li></ul></ul><ul><li>Experimental result </li></ul><ul><ul><li>Computation time : 0.04 second </li></ul></ul><ul><ul><li>Needed bandwidth : 6.3K bps </li></ul></ul><ul><li>Proposed method is feasible in practical environment. </li></ul>
  29. 29. Superiority to existing method <ul><li>Investigated performance of our method </li></ul><ul><li>Show superiority to existing method </li></ul><ul><li>Index : transition of packet arrival rate as time progresses </li></ul><ul><li>Experimental configuration </li></ul><ul><ul><li>Area size 3000 m ×3000 m </li></ul></ul><ul><ul><li>Number of nodes 1000 </li></ul></ul><ul><ul><li>Simulator   GTNetS </li></ul></ul><ul><ul><li>Radio Range 160m </li></ul></ul><ul><ul><li>MAC layer protocol IEEE802.11 (Max. 2Mbps) </li></ul></ul><ul><ul><li>Max of Speed 4 Km/hour </li></ul></ul><ul><ul><li>Mobility model random waypoint </li></ul></ul>
  30. 30. Comparison with existing method <ul><li>AQM (on-demand multicast routing method)[3] </li></ul><ul><li>Proposed method </li></ul><ul><ul><li>Optimized for communication stability </li></ul></ul><ul><ul><li>Optimized for the number of receivers </li></ul></ul><ul><ul><li>Optimized for power consumption </li></ul></ul>[3]K. B¨ur and C. Ersoy. Ad Hoc Quality of Service Multicast Routing. Computer Communications , 29(1):136–148, December 2005. Power saving stability number of receivers Yes No Yes Power-saving No No Yes #. of receivers No Yes Yes Stability
  31. 31. Transition of packet arrival rate The proposed method is superior to AQM in terms of packet arrival rate second AQM Stability #. of receivers Power-saving 公
  32. 32. Conclusion <ul><li>We proposed a new multicast routing method for MANET. </li></ul><ul><ul><li>To construct the semi-optimal multicast tree satisfying several QoS constraints for any given objective </li></ul></ul><ul><li>We show that the proposed method is feasible in practical environment. </li></ul>
  33. 33. <ul><li>The End </li></ul>
  34. 34. Result of power consumption Unit : Watt-second
  35. 35. Power consumption <ul><li>Compared item </li></ul><ul><ul><li>Transmission power consumption in 20 seconds </li></ul></ul><ul><ul><ul><li>20 seconds : reconstruction interval of multicast tree </li></ul></ul></ul>

×