Dinh Sindhura Tokala and My T Thai nanguyen tdinh sindhura mythai ciseufledu MOBICOM 2011 Overlapping Communities in Dynamic Networks Their Detection and Mobile Applications ID: 1024535
Download Presentation The PPT/PDF document "Nam P. Nguyen, Thang N." is the property of its rightful owner. Permission is granted to download and print the materials on this web site for personal, non-commercial use only, and to display it on your personal computer provided you do not modify the materials and that you retain all copyright notices contained in the materials. By downloading content from our website, you accept the terms of this agreement.
1. Nam P. Nguyen, Thang N. Dinh, Sindhura Tokala and My T. Thai{nanguyen, tdinh, sindhura, mythai}@cise.ufl.eduMOBICOM 2011Overlapping Communities in Dynamic Networks: Their Detection and Mobile Applications
2. MotivationA better understanding of mobile networks in practiceUnderlying structures?Organization of mobile devices?Better solutions for mobile networking problemsForwarding and routing methods in MANETsWorm containment methods in OSNs (on mobile devices)and possibly more …
3. Communities in mobile networksForwarding & Routing on MANETsSensor Reprogramming in WSNsWorm containment in Cellular networksCommunity Structure
4. Community structureNo well-defined concept(s) yetDensely connected inside each communityLess edges/links crossing communities
5. How do communities help in mobile networks?Forwarding & Routing on MANETsSensor Reprogramming in WSNsWorm containment in Cellular networks
6. Community detectionThe detection of network communities is importantQ: A quick and efficient CS detection algorithm?A: An Adaptive CS detection algorithmHowever, …Large and dynamic Mobile networksOverlapping communities
7. An adaptive algorithm::Input networkNetwork changesBasic communitiesPhase 1: Basic CS detection ()Updated communitiesPhase 2: Adaptive CS update ()Our solution:AFOCS: A 2-phase and limited input dependent framework
8. Phase 1: Basic communities detectionBasic communitiesDense parts of the networksCan possibly overlapBases for adaptive CS updateDutiesLocates basic communitiesMerges them if they are highly overlapped
9. Phase 1: Basic communities detectionLocating basic communities: when (C) (C) (C) = 0.9 (C) =0.725Merging: when OS(Ci, Cj) OS(Ci, Cj) = 1.027 = 0.75
10. Phase 1: Basic communities detection
11. Phase 2: Adaptive CS updateUpdate network communities when changes are introducedNetwork changesBasic communitiesUpdated communitiesNeed to handleAdding a node/edgeRemoving a node/edge+ Locally locate new local communities+ Merge them if they highly overlap with current ones
12. Phase 2: Adding a new nodeuuuY(Ct) ≥ t(4) × Y(OPT(u)t)
13. Phase 2: Adding a new edge
14. Phase 2: Removing a nodeIdentify the left-over structure(s) on C\{u}Merge overlapping substructure(s)
15. Phase 2: Removing an edgeIdentify the left-over structure(s) on C\{u,v}Merge overlapping substructure(s)
16. AFOCS: SummaryPhase 1: Basic CS detection ()Network changesPhase 2: Adaptive CS update ()Node/edge insertionsNode/edge removals
17. A community-based forwarding & routing strategy in MANETsChallengesFast and effective forwardingNot introducing too much overhead infoAvailable (non-overlapping) community-based routingsForward messages to the people/devices in the same community as the destination.Our method:Takes into account overlapping CSForwards messages to people/devices sharing more community labels with the destination
18. Experiment set upData: Reality Mining (MIT lab)Contains communication, proximity, location, and activity information (via Bluetooth) from 100 students at MIT in the 2004-2005 academic year500 random message sending requests are generated and distributed in different time pointsControl parametershop-limittime-to-livemax-copies
19. Results+ Competitive Avg. Delivery Ratio and Delivery Time+ Significant improvement on the number of Avg. Duplicate MessagesAvg. Delivery RatioAvg. Delivery TimeAvg. Duplicate Message
20. A community-based worm containment method on OSNsOnline social networks have become more and more popularWorm spreading on OSNsFrom computers computers (traditional method)From mobile devices mobile devices (Smart phones, PDAs, etc)
21. Worm containment methodsAvailable methods (cellular networks)Choosing people/devices from different disjoint communities and send patches to themOur method:Choosing the people/devices in the boundary of the overlap to send patches & have them redistribute the patches
22. Experiment set upDataset: Facebook network []New Orleans region63.7K nodes + 1.5M edges (Avg. degree = 23/5)Friendship and wall-postsWorm propagationFollows “Koobface” spreading modelAlarm thresholdα = 2%, 10% & 20%
23. Results
24. Results+ Better infection rates+ Number of nodes to be patched is greatly reducedα = 2%α = 10%α = 20%
25. SummaryAFOCSA 2-phase adaptive framework to identify and update CS in dynamic networksFast and efficientForwarding & Routing strategy on MANETsCompetitive Avg. Time and Delivery RatioSignificant improvement of number of Avg. Duplicate MessagesWorm containment on OSNsA tighter set of influential people/devicesBetter performance in comparison with other methods.
26. AcknowledgementFundingNSF CAREER Award grant 0953284DTRA YIP grant HDTRA1-09-1-0061DTRA grant HDTRA1-08-10.ShepherdDr. Cecilia Mascolo, University of Cambrigde, UK
27. Q&AThank you for your attention
28. Back-up slidesAdditional slides for questions that may arise in the presentation
29. Choosing
30. AFOCS performance
31. AFOCS performance