Adjusting ATIM Window Size Dynamically

According to [11], the ATIM window has a great effect on energy efficiency and performance. If the ATIM window is too large, nodes will not get chance to transmit data and increase power consumption. On the contrary, if the ATIM window is too small, only a few nodes can send ATIMs successfully. In PEMP, the ATIM window size is also adjusted dynamically based on network conditions [11].

Chapter 5

Simulation and Discussion

5.1 Simulation Model

For evaluation, we used ns-2 with the CMU wireless extension [19]. Simulation parameters are showed in Table 3 [15][20][22]. Nodes are randomly placed in an area of 1000 square meters. The transmission rate of each node is 2 Mbits/sec. The transmission range is 250 meters. The routing protocol is DSR (Dynamic Source Routing) [21]. The length of a beacon interval is 100 ms. The number of flow is a half of the number of nodes [20]. We set the upper limit of the number of beacon interval passed (up-bc) as three. Each node generated variable-rate traffic according to the exponential on-off traffic model. The packet size is randomly selected between 256 and 1024 bytes. We use the same energy model as in [15][22]. The power consumption for switching between active and sleep is negligible and not considered here. Nodes do not run out of energy during the simulation. We have three performance metrics: power consumption (J/sec/node), aggregate throughput (Kbytes/sec) and average end to end delay (msec). We study the performance when the network has 20 or 40 nodes. We simulated PEMP, DPSM and PSM. We define PSM(T) as power saving mode, and T represents the size of an ATIM window. In PSM simulations, we changed the size of an ATIM window size between 5 ms and 30 ms.

Note that nodes in PSM continue to stay active after finishing transmissions. But for fair comparison in our experiments PSM will allow a node to switch to sleep state after finishing its transmissions.

22

Table 3: Simulation parameters.

Area 1000 m × 1000 m

Bandwidth 2 Mbps

Range 250 m

Routing protocol DSR

Beacon interval 100 ms

Number of nodes 20, 40

Simulation configuration

Up-bc 3

Traffic rate Exponential Traffic

configuration Packet size 256 ~ 1024 bytes Transmit 420 + 1.9 × frame size(μJ)

Receive 330 + 0.42 × frame size(μJ)

Idle 808 mw

Energy model

Sleep 27 mw

5.2 Simulation Results and Discussion

Fig. 7 and Fig. 8 show the power consumption (J/sec/node) under 20 and 40 nodes, respectively. When the number of nodes increases, the power consumption becomes large. Power consumption in PSM is approximately linear increasing with T (ATIM window size) increasing from 5 ms to 30 ms. That is, if the ATIM is longer in PSM, nodes will consume more power consumption according to Fig. 7 and Fig. 8.

Comparing to PSM and DPSM, PEMP can decrease unnecessary idle time in active state by information exchange and QoS methods. The power consumption of PEMP is 20 % less than DPSM by decreasing the idle time in active state.

20 nodes

Fig. 7: Power consumption comparison under in 20 nodes.

24

Fig. 8: Power consumption comparison under 40 nodes.

Fig. 9 and Fig. 10 illustrate the aggregate throughput (Kbytes/sec) among PEMP, DPSM and PSM under 20 and 40 nodes, respectively. If the ATIM is too small in PSM, time is inadequate to announce ATIM. In our simulation result, the aggregate throughput degrades with the ATIM window size decreasing. PSM with ATIM window size of 5 ms may suffer severe degradation in throughput. In Fig. 9 and Fig.

10, we observe that for PSM the ATIM window of about 20 ms achieve the best throughput. DPSM can achieve higher throughput by choosing a suitable ATIM window. We observe the aggregate throughput of PEMP is 0.5 % less than that of DPSM.

20 nodes

Fig. 9: Aggregate throughput comparison under 20 nodes.

40 nodes

Fig. 10: Aggregate throughput comparison under 40 nodes.

The average end to end delay is shown in Fig. 11 and Fig. 12 under 20 and 40 nodes, respectively. The average end to end delay is computed by summarizing the end to end delay of all the connection flow and averaging it. We can see that smaller ATIM size cause large end to end delay. This is because using a small ATIM window

26

size there is not sufficient time to announce the ATIM. When a node fails to send its ATIM frame in the current ATIM window, it should retransmit ATIM frames in the next ATIM window, resulting in long end to end delay. PEMP has 6% longer end to end delay than DPSM.

Fig. 11: End to end delay comparison under 20 nodes.

40 nodes

Fig. 12: End to end delay comparison under 40 nodes.

Chapter 6

Conclusions and Future Work

6.1 Concluding Remarks

We have proposed a power efficient MAC protocol (PEMP) for multihop ad hoc networks by decreasing the idle time in active state. The overall power consumption can be reduced in multihop ad hoc networks by information exchange and QoS methods. In PEMP, nodes with larger buffered data should transmit data later and nodes with smaller buffered data should transmit first. In addition, starvation avoidance is also addressed by raising a node’s transmission priority if necessary. As multihop ad hoc networks are getting popular, it is important to have a power efficient MAC protocol for extending the battery life of wireless nodes. Simulation results have shown that the proposed PEMP can achieve 20% less power consumption with penalties of 6% longer end to end delay than DPSM.

6.2 Future Work

In the IEEE 802.11, a node can be in one of two power management modes, active mode and power saving mode. In active mode, a node can achieve higher network throughput but consume more energy. In power saving mode, a node can reduce power consumption but network throughput will degrade. On demand power management (ODPM) [10] addressed a design space between active mode and power saving mode for power saving. The future work is to integrate ODPM into PEMP to achieve a better tradeoff between power consumption, throughput and delay in multihop ad hoc networks.

28

Bibliography

[1] IEEE Std 802.11b, IEEE Standard for Wireless LAN Medium Access Control (MAC) and Physical Layer Specifications: Higher-Speed Physical Extension in the 5GHz Band, 1999.

[2] IEEE Std 802.11a, IEEE Standard for Wireless LAN Medium Access Control (MAC) and Physical Layer Specifications: Higher-Speed Physical Extension in the 2.4 GHz Band, 1999.

[3] IEEE Draft Std 802.11e, Medium Access Control (MAC) Enhancements for Quality of Service (QoS), D4.4, Jun. 2003

[4] S. Narayanaswamy, V. Kawadia, R. S. Sreenivas, and P. R. Kumar. “Power control in ad hoc networks: theory architecture algorithm and implementation of the COMPOW protocol,” in Proc. European Wireless, Florence, Italy, Feb.

2002, pp.156-163.

[5] Krunz, M. and Muqattash, A. “A power control scheme for MANETs with improved throughput and energy consumption,” in Proc. IEEE WPMC, Oct.

2002, pp. 771-775.

[6] Vikas Kawadia and Kumar, P.R. “Power control and clustering in ad hoc networks,” in Proc. IEEE INFOCOM, Apr. 2003, vol. 3, pp. 459-469.

[7] Shih Liu Wu, Yu Chee Tseng Jang Ping Sheu “Intelligent medium access for mobile ad hoc networks with busy tones and power control,” IEEE journal on Selected Areas in Communications, pp. 1647-1657, Sep. 2001,.

[8] E. S. Jung and N. H. Vaidya, “Mediun access control for ad hoc Networks: a power control MAC protocol for ad hoc networks,” in Proc. ACM MOBICOM, Sep. 2002, pp. 26-46.

[9] J.P. Monks, V. Bharghavan and W. W. Hwu, “A power controlled multiple access protocol for wireless packet networks,” in Proc. IEEE INFOCOM, Apr.

2001, pp. 219-228.

[10] R. Zheng and R. Kravets, “On-demand power management for ad hoc networks,” in Proc. IEEE INFOCOM, Apr. 2003, vol. 1, pp. 481-491.

[11] E. S. Jung and N. H. Vaidya, “An energy efficient MAC protocol for wireless LANs,” in Proc. IEEE INFOCOM, June 2002, vol. 3, pp. 1756-1764.

[12] Wei Ye, John Heidemann, and Deborah Estrin, “An energy efficient MAC protocol for wireless sensor network,” USC/Information Sciences Institute [13] Y.-C. Tseng; C.-S. Hsu and T.-Y. Hsieh, “Power-saving protocols for IEEE

802.11-based multi-hop ad hoc networks,” in Proc. IEEE INFOCOM, June 2002, pp. 200-209.

[14] M. Liu and M. T Liu, “A power-saving scheduling for IEEE 802.11 mobile ad hoc networks,” in Proc. ICCNMC, Oct. 2003, pp. 238-245.

[15] S.-L. Wu and P.-C. Tseng “An energy efficient MAC protocol for IEEE 802.11 WLANs,” in Proc. IEEE CNSR, May 2004, pp. 137-145.

[16] B. Tavli and W. B. Heinzelman, “TRACE: time reservation using adaptive control for energy efficiency,” IEEE Journal on Selected Areas in Communications, pp. 1506-1515, Dec. 2003.

[17] E. S. C. Takahashi, “Application aware scheduling for power management on IEEE 802.11,” in Proc. IEEE IPCCC, Feb. 2000,pp. 274-253.

[18] W. H. Yuen and Ch. W. Sung, “Energy efficiency and network connectivity of mobile ad hoc networks,” in Proc. Distributed Computing Systems, May 2003, pp. 38–45.

30

[19] The CMU Monarch Project., “The CMU monarch project’s wireless and mobility extension to Ns,” [Online]. Available : http://www.monarch.cs.cmu.edu/cmu-ns.html.

[20] E. S. Jung and N. H. Vaidya, “A power saving MAC protocol for wireless networks,” Tech. Rep., July 2002, [Online]. Available:

http://www.crhc.uiuc.edu/~nhv/papers/npsm_tech.pdf

[21] D. B. Johnson and D. A. Maltz, “Dynamic source routing in ad hoc wireless networks” draft-ietf-manet-dsr-04.txt, 2001.

[22] L. M. Feeney. “An energy consumption model for performance analysis of routing protocol for mobile ad hoc networks,” ACM/Kluwer Mobile Networks and Applications, Vol. 6, No. 3, pp. 239-249, 2001.