4.1 Simulation Scenario
All scenarios have been implemented in the network simulator (NS2, 2.1b7) [17].
In these simulations, there are no hidden stations and the channel is assumed to be error free. The simulation topology is shown in Fig. 4.1. Table 4.1 shows the IEEE 802.11a PHY/MAC parameters used in these simulations and Table 4.2 shows the network parameters selected for the three ACs.
Fig. 4.1 Simulation Topology
SIFS 16µs DIFS 34µs
PHY Rate 54Mbps Minimum Bandwidth 6Mbps Slot Time 9µs PHY Header 24bytes Preamble Length 20µs PCLPHeader Length 4µs
CCATime 4µs RxTxTurnaroundTime 2µs
aCWmin 15 aCWmax 1023
IP Header Size 20bytes UDP Header Size 8bytes Table 4.1 The IEEE 802.11a PHY/MAC parameters
Parameters Audio Video Best-effort Data Priority High Medium Low
CWmin 3 7 15
CWmax 7 15 1023
AIFSN 2 2 7
Packet Size(bytes) 160 1080 1500 Packet Interval(ms) 20 2.16 12 Sending Rate(Kbps) 64 4000 1000
Table 4.2 MAC Parameters for the Three Traffic Categories
4.2 Priority Test
Fig. 4.2 is an example of throughput of different streams to test the priority. In this case, one QAP and only one QSTA are in the BSS and the payload of each stream is 1500 bytes for comparing the priority only without any other factor, like fragmentation. The audio stream tries to transmit 64Kbps, the video stream tries to transmit 10Mbps, and each of the other two data streams tries to transmit 15Mbps. Audio, video and Data_1 start at the same time.
Before 8 seconds, there is enough network capacity, the audio and video achieves their own transmission rates, 64Kbps and 10Mbps respectively, so does Data_1, 15Mbps. After 8 seconds, Data_2 starts to send packets. However, the network capacity is not sufficient, so the two data streams share the remaining sources and the throughput of audio and video still remains their own transmission rates. The example shows that when the network capacity is enough, the lower priority traffic makes use of the extra resources. When the capacity is insufficient, the throughput of higher priority traffic is still maintained while the lower priority traffic shares the remaining sources. The bandwidth shown in Fig. 4.2 is the average bandwidth. From our test, the maximum throughput for 1500 bytes payload is 24.76Mbps which is close to 25Mbps calculated in [18] and indirectly proves the correctness of our implementation model since we have made a lot of changes in NS-2 module.
Fig. 4.2 Priority Test for Different Streams
Fig. 4.3 The Mean Latency for Priority Test
The mean latency is shown in Fig. 4.3. After 8 seconds, the mean latency of audio or video traffic is both lower than 1ms. But, the mean latency of data_1 or data_2 is larger than 100ms which proves the lower priority is suppressed to transmit when the network capacity is not sufficient.
4.3 Simulation Results for DCF and EDCA
The simulation results of bandwidth with DCF and EDCA mechanisms are shown in Fig. 4.4 and Fig. 4.5 respectively. The mean delay is shown in Fig. 4.6 and Fig. 4.7. There are total 15 traffic streams in this scenario, five for each AC, and the parameters are set as Table 4.2. By comparing Fig. 4.4 and Fig. 4.5, which plot the throughput of each traffic stream, we observe that the throughput of video and data are significantly different for DCF and EDCA. The video traffic is well served with EDCA while there are many packets dropped with DCF. The mean delay of video and voice traffic for EDCA is around 1ms, but for DCF, the mean delay of video is larger than 50ms. The mean delay of data traffic for EDCA is far larger than that of real-time traffic, which tells us again that the transmission lower priority is suppressed in order to increase the transmission probability of the higher priority. So, EDCA can achieve QoS requirements for real-time traffic if there is enough network capacity.
However, if we add a new video stream to increase the channel loading, at high loads, EDCA can not perform well as in low loads, as shown in Fig. 4.8. In Fig. 4.9, the latency of video is larger than 100ms, so the transmission of video is almost unpractical at this case. Therefore, we proposed our methods to solve the problem described in Chapter 3.2 and the method is more easily implemented than HCCA, which has the same purpose to improve the performance of the real-time traffic in wireless network.
Fig. 4.4 The Throughput for DCF
Fig. 4.5 The Throughput for EDCA
Fig. 4.6 The Mean Latency for DCF
Fig. 4.7 The Mean Latency for EDCA
Fig. 4.8 The Throughput for EDCA at High Loads
Fig. 4.9 The Mean Latency for EDCA at High Loads
4.4 Simulation Results for the Proposed Method
In this section, we study the performance of our proposed method. Table 4.2 shows the parameters of traffic streams. The SurplusFactor of all real-time traffic is
1 .
1 , ALT[1]=0.7×100ms=70ms , ALT[2]=0.2×100ms=20ms , the beacon interval is 100ms, and the damping factor f is 0.9. At the beginning, there are six audio streams, five video streams and two best-effort data streams in the system. At the time 10s, another two video streams arrive. Fig. 4.10 and Fig 4.12 show that if the procedure of protecting the existing real-time traffic described in Chapter 3.2.1 is not implemented in QAP, the performance of video stream is terrible. But if the protected procedure is implemented, QoS of the existing performance is guaranteed as in Fig.
4.11 and Fig. 4.13. The reason why the new arriving video streams still be transmitted at 0.1Mbps is that TXOPBudget[2] still remains resources at high loads so that the new streams can make use of it to provide more services.
Fig. 4.10 The Throughput of EDCA without Protecting Real-Time Traffic
Fig. 4.11 The Mean Latency of EDCA without Protecting Real-Time Traffic
Fig. 4.12 The Throughput of EDCA with Protecting Real-Time Traffic
Fig. 4.13 The Mean Latency of EDCA with Protecting Real-Time Traffic
Next, we extend the scenario to test the second proposed method described in Chapter 3.2.2. α and β are set equal to 0.8. ∆1 is set to 10, ∆2 is set to15,
∆ is set to 3 0, E[idle]=0.031ms and Threshold is calculated from equation (3.20) dynamically changing with beacon interval. If new three data traffic streams arrival at the time 25s, the performance of video traffic is degraded again, as show in Fig. 4.14 and Fig. 4.15. After the implementation of the second proposed method, the real time traffic is protected again because the procedure lowers the throughput of data traffic after the time 25s as in Fig. 4.16 and Fig. 4.17.
Through the simulation, obviously our proposed methods are workable.
Therefore, with the control of real time traffic and best-effort data traffic, the network system can easily achieve the requirements of QoS.
Fig. 4.14 The Throughput of EDCA without Data Control
Fig. 4.15 The Mean Latency of EDCA without Data Control
Fig. 4.16 The Throughput of EDCA with Data Control
Fig. 4.17 The Mean Latency of EDCA with Data Control
Chapter 5
Conclusion
In thesis, we have given an overview of the IEEE 802.11e and evaluated the performance, in transmitting QoS applications. The block acknowledgment and direct link protocol are not implemented in our simulation model. How to apply these two mechanisms to enhance the QoS and to automatically change ATL[i] with network capacity is the future work.
Through our simulations, we find that although EDCA could provide a service differentiation among different access categories, it is still deficient in QoS guarantee at heavily loaded traffic network conditions under huge amount of best-effort traffic.
In such case, we proposed the schemes of protecting the existing real time traffic and suitably controlling the best-effort data to avoid damaging the performance of the time-bounded traffic. For voice and video streams, QSTAs listen the budget from QAP to determine on accepting or rejecting the new streams. For best-effort data transmission, QAP dynamically control the data parameter based on the traffic condition. Under the implementation of these schemes on QAP and QSTAs, QoS requirements from heavy load traffic in EDCA mode could be better satisfied.
Bibliography
[1] IEEE Std. 802.11, "Wireless LAN Medium Access Control (MAC) and Physical Layer (PHY) Specifications,” 1999.
[2] IEEE Std. 802.11e/D5.1, "Medium Access Control (MAC) Enhancements for Quality of Service (QoS),” October 2003.
[3] D. Chen, D. Gu and J. Zhang, "Supporting Real-Time Traffic with QoS in IEEE 802.11e Based Home Networks," Consumer Communications and Networking Conference (CCNC), pp. 205-209, January 2004
[4] S. Choi, JD Prado, S. Shankar and S. Mangold, "IEEE 802.11e contention-based channel access (EDCF) performance evaluation," in Proc. IEEE ICC'03, vol. 2, pp. 1151-1156, May 2003.
[5] T. S. Ho and K. C. Chen, "Performance evaluation and enhancement of the CSMA/CA MAC protocol for 802.11 wireless LAN’s," in Proc. IEEE PIMRC, Taipei, Taiwan, pp. 392-396, October 1996.
[6] L. Yang, "P-HCCA: A New Scheme for Real-time Traffic with QoS in IEEE 802.11e Based Networks," APAN Network Research Workshop 2004.
[7] Q. Ni, L. Romdhani, and T. Turletti, "A Survey of QoS Enhancements for IEEE 802.11 Wireless LAN," Journal of Wireless Communications and Mobile Computing, vol. 4, no. 5, pp. 547-566, August 2004.
[8] B. Walke, Mobile Radio Networks, 2nd ed., Wiley, 2001.
[9] G. Bianchi, "Performance analysis of the IEEE 802.11 distributed coordination function," IEEE Journal on Selected Areas in Communications, pp. 535-547, March 2000.
[10] H. Zhu and I. Chlamtac, "An Analytical Model for IEEE 802.11e EDCF Differential Services," IEEE ICCCN3, Dallas, October 2003.
[11] J. Gosteau, M. Kamoun, S. Simoens, P. Pellati, "Analytical developments on QoS enhancements provided by IEEE 802.11 EDCA, " ICC 2004, Paris.
[12] L. Romdhani, Qiang Ni, and T. Turletti, "Adaptive EDCF: enhanced service differentiation for IEEE 802.11 wireless ad-hoc networks," IEEE Wireless Communications and Networking, vol. 2, pp. 1373 – 1378, March 2003.
[13] Y. Xiao, H. Li, and S. Choi, "Protection and Guarantee for Voice and Video Traffic in IEEE 802.11e Wireless LANs," Proc. of IEEE INFOCOM 2004.
[14] Y. Xiao and H. Li, "Voice and Video Transmissions with Global Data Parameter Control for the IEEE 802.11e Enhance Distributed Channel Access," IEEE Transactions on Parallel and Distributed Systems, vol. 15, no. 11, pp. 1041-1053, November 2004.
[15] Y. Xiao and H. Li, "Evaluation of distributed admission control for the IEEE 802.11e EDCA," IEEE Communications Magazine, vol. 42, pp. S20 - S24, September 2004.
[16] D. Pong and T. Moors, "Call Admission Control for IEEE 802.11 Contention Access Mechanism," In Procceedings of IEEE Globecom, 2003.
[17] Berkeley Network Simulator ns-2. http://www.isi.edu/nsnam/ns.
[18] T. Cooklev and V. Shah, "Throughput of 802.11e, assuming 802.11a, 802.11b, or 802.11g physical layers," IEEE 802.11 interim meeting, Sept. 2004, document 11-04-1076-00-000e. 3.