1. Albert- Ludwigs- Universität Freiburg Peer -To – Peer Networks luetooth Scatternet Based on Cube Connected Cycle H. K. Al-Hasani

2. luetooth Scatternet Based on Cube Connected Cycle What is....? –Piconet –Scatternet Other approaches : –TSF and BlueRings –Chains and Loops –Stars –BlueCubes CCC CCC and Scatternet CCC and iCCC What makes CCC different...? Conclusion

3. luetooth Scatternet Based on CCC What is....? Piconet Scatternet http://Tux.crystalxp.net

4. Bluetooth M P S 4/16 Piconet: One Master, seven Slaves Master determines Hopping- frequency. Active Slaves : can communicate. Parked Salves : listen Piconet: One Master, seven Slaves Master determines Hopping- frequency. Active Slaves : can communicate. Parked Salves : listen

5. Bluetooth Scatternet: Two or more Piconets are connected through a bridge. Slave- Slave bridge Scatternet: Two or more Piconets are connected through a bridge. Slave- Slave bridge M P S B 4/16 Piconet: One Master, seven Slaves Master determines Hopping- frequency. Active Slaves : can communicate. Parked Salves : listen Piconet: One Master, seven Slaves Master determines Hopping- frequency. Active Slaves : can communicate. Parked Salves : listen

6. Bluetooth M P S 4/16 Scatternet: Two or more Piconets are connected through a bridge. Master- Slave bridge Scatternet: Two or more Piconets are connected through a bridge. Master- Slave bridge Piconet: One Master, seven Slaves Master determines Hopping- frequency. Active Slaves : can communicate. Parked Salves : listen Piconet: One Master, seven Slaves Master determines Hopping- frequency. Active Slaves : can communicate. Parked Salves : listen

7. Bluetooth M P S 4/16 Master-Master bridge is forbidden Scatternet: Two or more Piconets are connected through a bridge. Master- Slave bridge Scatternet: Two or more Piconets are connected through a bridge. Master- Slave bridge Piconet: One Master, seven Slaves Master determines Hopping- frequency. Active Slaves : can communicate. Parked Salves : listen Piconet: One Master, seven Slaves Master determines Hopping- frequency. Active Slaves : can communicate. Parked Salves : listen

8. luetooth Scatternet Based on CCC Other approaches : TSF and BlueRings Chains and Loops Stars BlueCubes

9. Other approaches : 6/16 S M TSF : Roles assignment; unique path. Nodes in the middle are Master-Slave. Extending the tree = Extending Routing length Time complexity: n-1 TSF : Roles assignment; unique path. Nodes in the middle are Master-Slave. Extending the tree = Extending Routing length Time complexity: n-1

10. Other approaches : BlueRings : Multi path; fault tolerance; no Roles assignment Time complexity: n BlueRings : Multi path; fault tolerance; no Roles assignment Time complexity: n 6/16 S M TSF : Roles assignment; unique path. Nodes in the middle are Master-Slave. Extending the tree = Extending Routing length Time complexity: n-1 TSF : Roles assignment; unique path. Nodes in the middle are Master-Slave. Extending the tree = Extending Routing length Time complexity: n-1

11. Other approaches : Chains and Loops: No Master-Slave bridge, Parked in one and active in another;Time delay. B M M B S M 7/16

12. Other approaches : Star:Node in the middle is bottleneck. Time complexity: n-1 Star:Node in the middle is bottleneck. Time complexity: n-1 M S 7/16 B M M B S M Chains and Loops: No Master-Slave bridge, Parked in one and active in another;Time delay.

13. Other approaches : Star:Node in the middle is bottleneck. Time complexity: n-1 Star:Node in the middle is bottleneck. Time complexity: n-1 M S 7/16 B M M B S M Collisions = Retransmission = Power consuming Chains and Loops: No Master-Slave bridge, Parked in one and active in another;Time delay.

14. Other approaches : BlueCubes: start with ring and end up with cube - # Piconets is controlled - Roles assignment - No Master- Slave link - Multi disjoint path - Scatternet of the same degree (dimension) can connect. Time complexity: log 2 n BlueCubes: start with ring and end up with cube - # Piconets is controlled - Roles assignment - No Master- Slave link - Multi disjoint path - Scatternet of the same degree (dimension) can connect. Time complexity: log 2 n M S M M B M 8/16

15. luetooth Scatternet Based on Cube Connected Cycle CCC CCC and Scatternet CCC and iCCC What makes CCC different...? http://wikimedia.org

16. Cube Connected Cycle CCC: n-dimensional cube Vertex are replaced by cycles Each cycle has n nodes CCC has n.2 n node X is cyclic index (integer n-1>=X>=0) Y is cubic index (binary Y<= 2 n -1) CCC: n-dimensional cube Vertex are replaced by cycles Each cycle has n nodes CCC has n.2 n node X is cyclic index (integer n-1>=X>=0) Y is cubic index (binary Y<= 2 n -1) node (x,y) cyclic neighbors (x ±1,y) Cubic neighbors (x, y 2 x ) 10/16

17. Cube Connected Cycle CCC: Cyclic index and cubic index Local cycles and primary nodes Outside and Inside leaf sets CCC: Cyclic index and cubic index Local cycles and primary nodes Outside and Inside leaf sets 0 0 3 3 2 2 5 5 1 1 4 4 11/16

18. Cube Connected Cycle 3 3 5 5 0 0 4 4 1 1 2 2 Node ID(1,011) Routing table cubical neighbour: (0,---) cyclic neighbour: (0,101) cyclic neighbour: (0, 001) half smaller, half larger Inside Leaf Set (0,011)(2,011) Outside Leaf Set (1,100)(2,010) 12/16

19. CCC and Scatternet 13/16 CCC: CCC has n.2 n Piconets Every node is a Master Master communicate through bridges min CCC = 5. n.2 n-1 n >=3, max CCC = 13. n. 2 n-1 n >=3 CCC: CCC has n.2 n Piconets Every node is a Master Master communicate through bridges min CCC = 5. n.2 n-1 n >=3, max CCC = 13. n. 2 n-1 n >=3

20. CCC and Scatternet 13/16 * 4-dimentional cube * CCC: CCC has n.2 n Piconets Every node is a Master Master communicate through bridges min CCC = 5. n.2 n-1 n >=3, max CCC = 13. n. 2 n-1 n >=3 CCC: CCC has n.2 n Piconets Every node is a Master Master communicate through bridges min CCC = 5. n.2 n-1 n >=3, max CCC = 13. n. 2 n-1 n >=3

21. CCC and Scatternet 13/16 * 4-dimentional cube * CCC: CCC has n.2 n Piconets Every node is a Master Master communicate through bridges min CCC = 5. n.2 n-1 n >=3, max CCC = 13. n. 2 n-1 n >=3 CCC: CCC has n.2 n Piconets Every node is a Master Master communicate through bridges min CCC = 5. n.2 n-1 n >=3, max CCC = 13. n. 2 n-1 n >=3

22. CCC and iCCC Extending CCC is expensive iCCC: Intermediate CCC Reconstructed CCC has (n+1).2 n Piconets instead of (n+1).2 n+1 Local transmission iCCC: Intermediate CCC Reconstructed CCC has (n+1).2 n Piconets instead of (n+1).2 n+1 Local transmission 14/16 New Master (x,y) x=n

23. What makes CCC different...? If CCC with iCCC are combined: Efficient communication Fast lookup O(n) Broadcast and unicast Dynamic system Fixed routing table Bounded number of reconstruction Roles assignment Efficient communication Fast lookup O(n) Broadcast and unicast Dynamic system Fixed routing table Bounded number of reconstruction Roles assignment 15/16

24. luetooth Scatternet Based on CCC Conclusion Expensive and complicated to reality. Thank you 16/16 S M Routing length B M M B S M Time delay M S Bottleneck M S M M B M Multicasting Expensive

25. References Cycloid: A constant-degree and lookup-efficient P2P overlay network Haiying Shen, Cheng-Zhong Xu, and Guihai Chen Cube Connected Cycles Based Bluetooth Scatternet Formation Marcin Bienkowski1,, Andr´e Brinkmann2, Miroslaw Korzeniowski1,, and Orhan Orhan1 Routing Strategy for Bluetooth Scatternet Christophe Lafon, and Tariq S. Durrani Bluetooth scatternet formation Søren Debois, IT University of Copenhagen Energy-Efficient Bluetooth Scatternet Formation Based on Device and Link Characteristics Canan PAMUK On Efficient topologies for Bluetooth Scatternets Department of Information Engineering University of Padova, ITALY Daniele Miorandi, Arianna Trainito, Andrea Zanella BlueCube: Constructing a hypercube parallel computing and communication environment over Bluetooth radio systems Chao-Tsun Chang Introduction to Bluetooth Technology Lecture notes by Jeffrey Lai, http://www.ensc.sfu.ca Introduction to Wireless and Mobile Systems Dharma Parkash Agrawal, Qing – An Zeng Ad Hok Wireless Networks, architecture and protocols * * Cayley DHTs A Group-Theoretic Framework for Analyzing DHTs Based on Cayley Graphs Changtao Qu, Wolfgang Nejdl, Matthias Kriesell Wireless ad hoc networkingThe art of networking without networking Magnus Frodigh, Per Johansson and Peter Larsson http://bluetooth.com/bluetooth/ http://www.palowireless.com/bluetooth/