1. A High-Throughput MAC Protocol for Wireless Ad Hoc Networks Wanrong Yu, Jiannong Cao, Xingming Zhou, Xiaodong Wang, Keith C. C. Chan, Alvin T. S. Chan, and H.V. Leong School of Computer, National University of Defense Technology, Changsha, China Department of Computing, HongKong Polytechnic University IEEE Transaction on Wireless Communications, Vol. 7, No.1, January 2008

2. Outline Introduction Introduction CTMAC CTMAC Performance evaluation Performance evaluation Conclusions Conclusions

3. Inefficiency of IEEE 802.11

4. Background and related work Effort has been made on increasing the throughput of MANETs Effort has been made on increasing the throughput of MANETs –Adding additional control gap –Transmission power control (TPC)

5. Additional control gap To add control gap between between RTS/CTS and DATA packets for scheduled transmission To add control gap between between RTS/CTS and DATA packets for scheduled transmission –[6]:MACA-P, IEEE PerCom 2003 –[7]:enhancement of MACA-P, BoradNets 2004 Do not consider tolerable interference Do not consider tolerable interference

6. Transmission power control TPC is used per-packet to increase the spatial channel reuse TPC is used per-packet to increase the spatial channel reuse –[5]: A power control MAC protocol for ad hoc networks, ACM/Kluwer Wireless Networks, 2005 –[12]: POWMAC, IEEE J. Select Areas Commun., 2005 Latency in change of transmission power is huge [11], which make TPC-based solutions difficult to use in practice Latency in change of transmission power is huge [11], which make TPC-based solutions difficult to use in practice

7. Goal of the paper To propose a MAC protocol works on To propose a MAC protocol works on –Single channel –Single transceiver –Single transmission power

8. Basic operation of CTMAC Master pair Slave pair

9. CTMAC In CTMAC, every node maintains an Active Neighbor List (ANL) In CTMAC, every node maintains an Active Neighbor List (ANL) For node i, ANL i contains For node i, ANL i contains Address of u Channel gain bewteen i and u Starting time of DATA and ACK To distinguish the transmitter and receiver Maximum tolerable interference of u

10. Handshakes in CTMAC: case 1 RTS-CTS RTS-CTS –Successful in information exchange –The receiver agrees with the sender and reply a normal CTS

11. Handshakes in CTMAC: case 2 RTS-CTS-ATS (adjust-to-send) RTS-CTS-ATS (adjust-to-send) –The slave receiver modify the value of T data and T ack declared by the sender in RTS –The slave receiver includes the new value in CTS –The slave sender has to inform its neighbors the new value by ATS

12. Handshakes in CTMAC: case 3 RTS-NCST-ATS (abort-to-send) RTS-NCST-ATS (abort-to-send) –The slave receiver finds it is implssible to continue the slave transmission –The slave sender informs its neighbor by ATS

13. Minimum reaching power Minimum reaching power Accumulated interference power Accumulated interference power Total future interference that node v can tolerate, P rx is the raching power Total future interference that node v can tolerate, P rx is the raching power Tolerable interference estimation

14. Maximum tolerable interference that each future neighboring node can add Maximum tolerable interference that each future neighboring node can add α is the ratio between the interference caused by nodes outside and inside the transmission range α is the ratio between the interference caused by nodes outside and inside the transmission range –α = 0.5 for two ray models and uniform distributed nodes

15. Concurrent transmisson control RC0: requirement of time RC0: requirement of time –For the slave sender to check if the master transmission’s ACG is long enough for exchanging of control packets RC1: for slave transmitter RC1: for slave transmitter –Check if transmission will collide with any scheduled transmission –For all u in ANL, check if

16. Concurrent transmission control RC2: for slave receiver RC2: for slave receiver –To determine if the accumulated should not violate the slave receiver’s SINR – RC3: ACK transmission RC3: ACK transmission –To postpone the transmission of ACK

17. Parameters in the simulation

18. Line topology

19. Throughput of line topology

20. Random grid topology 800 meter square area 800 meter square area The square is split into n*n small squares The square is split into n*n small squares one node placed in a small square randomly one node placed in a small square randomly m transmission pairs m transmission pairs

21. Throughput (m=2)

22. Throughput (m=3)

23. Throughput (m=4)

24. Cluster topology An area of 400*400 m An area of 400*400 m 16 nodes, split into 4 equal groups 16 nodes, split into 4 equal groups Each group occupying a 100*100 square Each group occupying a 100*100 square The receiver is selected from another cluster with a probability of p The receiver is selected from another cluster with a probability of p

25. Throughput (p=0.25)

26. Throughput (p=0)

27. Random topology 1000*1000 meter area 1000*1000 meter area 100 nodes placed randomly 100 nodes placed randomly M end-to-end flows M end-to-end flows Size of control gap is 640B Size of control gap is 640B

28. Throughput at random topology

29. Conclusion CTMAC CTMAC –Is based on a single transceiver circuitry –Operates over a single channel –Works on single transmission power

30. Thank you!!