Chapter 2 Max Freedom Last Downlink Scheduling Algorithm
2.5. Concluding Remarks
In this chapter, a max freedom last (MFL) downlink scheduling algorithm is proposed for DSRC networks. The MFL scheduling algorithm can minimize the system handoff rate under a maximum tolerable delay for the service. We also give a proof that the process in the reverse line-up phase can minimize the system handoff rate. Simulation results show that the
MFL scheduling algorithm has better performance in the service failure rate and the system handoff rate than FCFS and EDF. This is because the MFL scheduling algorithm considers the remaining SCH dwell time, the remaining transmission time, and the queueing delay for all OBUs. Therefore the MFL scheduling algorithm can deal with the mobility issue at DSRC networks and reduce the system overhead.
Chapter 3
In the next generation transportation system, the drivers or users in mobile environment expect to get various types of information closely related to their life from anywhere and at anytime. ITS is developed to reach the goal. The service provided by ITS should have time limitation otherwise the information would become useless even if the information data is successfully received by users. The time limitation depends on the effective area of the service data.
Therefore the information data in ITS can be categorized into to three types. The first type is the service data not depending on the location such as music, news, or software program. The second type is the service data depending on the location and required immediately such as the traffic accident information or the road condition. The third type is the service data depending on the location but not required immediately such as travel information. The second type belongs to the urgent service. The size of the information data is small and this kind of data will be transmitted in CCH directly. The first type and the third type are the service supported by SCH. In this chapter, the MFL scheduling algorithm considers only the first and the third type services. The data for these two type services can be classified into different classes according to the effective area. Each class will have
different maximum tolerable delay. We verify the MFL scheduling algorithm proposed in the chapter 2 in this kind of scenario.
The rest of this chapter is organized as follows. At first, the modified system model is described in section 3.2. And the simulation result and concluding remarks are presented in section 3.3 and 3.4, respectively.
3.2. System Model
Figure 3-1 shows the modified structure of the queue selector. The mainly difference is that OBUs which requests different classes of services have various maximum tolerable delay.
When the OBU requests information from the information server, RSU can know the service class of the information data. The service class defines the effective area range of the information type. Besides, each OBU has different velocity. Therefore RSU can determinate the maximum tolerable delay for each OBUs as the input of the MFL algorithm. The maximum tolerable delay equals the effective area range divided by the velocity of the OBU.
In this chapter, we define three classes for the service data: small range, medium range, and large range. The small range and medium range are corresponding to the third type of service and the large range is corresponding to the first type of service in the section 3.1. Although the first type of service is not depending on the location, the user does not hope to wait for too long time.
Therefore the weighting function described in equation (2.11) should be modified as follows:
where T is the maximum tolerable delay for OBU i . The other portions of the system i
model and the algorithm are the same as these in the chapter 2.
Figure 3-1. Modified structure of the Queue Selector
3.3. Simulation Result
3.3.1. Simulation Environment
There are three classes of service data in the simulation: small range, medium range, and large range. We define the value of the effective distance for each class individually. The system parameters are described as follows:
System Parameters Notation Value
The distance between two RSUs d 400 m
The mean arrival rate of the request data OBUs
µ
n variable The parameter of the Pareto distribution α 1.1 The minimal number of MSDU of the request data k 150 The maximal number of MSDU of the request data m 10000The size of each MSDU M 1000 bytes
The minimal velocity of vehicles Vmin 60 km/hr The minimal velocity of vehicles Vmax 120 km/hr
The maximum SCH Time Tsmax 100 ms
CCH Wait Time T w 5 ms
Repeat period of the RST transmission Tr 1050 ms The maximal number of accepted OBUs per RST r 4
Data transmission rate R 18 Mbps
The number of probes between two RSTs p 9 The effective distance for small range class of service 500 m The effective distance for medium range class of service 1 km
The effective distance for large range class of service 5 km Table 3-1. System Parameters with Service Classification
3.3.2. Simulation Result and Conclusion
Figure 3-2 shows the service failure rate versus the mean arrival rate for the three algorithms. It can be seen that MFL still has smaller service failure rate than FCFS and EDF.
Besides, the difference value in service failure rate between the MFL and EDF becomes
larger than the simulation result in chapter 2. It is because that MFL algorithm can trace the variation of the maximum tolerable delay for OBUs.
Figure 3-2. Service Failure Rate
Figure 3-3 shows the system utilization versus the mean arrival rate for the three algorithms. It can be seen that three algorithms can achieve the same system utilization. The result is the same as that in chapter 2.
Figure 3-4 shows the system handoff rate versus the mean arrival rate for three algorithms. It can be seen that MFL algorithm still have better than EDF and FCFS and the trend of the curves is almost the same as it in chapter 2.
Figure 3-4. System Handoff Rate
3.4. Concluding Remark
In this chapter, we classify the service into three classes according to the effective range of the request data. The maximum tolerable delay requirement for each OBU depends on the effective range of the request data and the velocity of vehicle. In simulation result, we can find that MFL scheduling algorithm has better performance in service failure rate compared with the chapter 2. It is because the MFL algorithm can trace the variation of the maximum
tolerable delay for OBUs. And the MFL algorithm also has better performance than EDF and FCFS in system handoff rate. Therefore the MFL scheduling algorithm can be applied to DSRC networks with different maximum tolerable delay for each OBU in ITS.
Chapter 4 Conclusion
In this thesis, a max freedom last (MFL) downlink scheduling algorithm is proposed for DSRC networks. We consider that RSU provides a service that downlink transmits data to OBUs that request from some information server in the Internet. This MFL algorithm can minimize the system handoff rate under the maximum tolerable delay requirement. The algorithm schedules the OBUs according to the remaining SCH dwell time, remaining transmission time, queueing delay, and the maximum tolerable delay for each OBU. These factors should be considered for the ITS services in the mobility environment in order to let the system operate more efficiently.
The MFL algorithm trends to serve the OBUs which can be completely served first in order to minimize the system handoff rate and then serve other OBUs if there is remaining resource in order to make full system utilization. Also, the algorithm increases the priority gradually for the OBUs which are close to the maximum tolerable delay to avoid OBUs failure.
We classify the service into three classes according to the effective range of the request data. The maximum tolerable delay requirement for each OBU can obtain from its effective range of the request data and velocity.
Simulation results show that the MFL algorithm has better performance than FCFS and EDF in the service failure rate and the system handoff rate. And MFL also achieve the same system utilization as FCFS and EDF.
Bibliography
[1] C. H. Park, and D. H. Cho, “An adaptive logical link control for wireless Internet service in ITS,” Vehicular Technology Conference, vol.4, pp.2213 – 2217, Sept. 1999.
[2] T. Munaka, Y. Ito, and S. Kubota, “A study of info-communication technologies required in ITS networks,” Applications and the Internet Workshops, pp.171 – 176, Jan.
2001.
[3] H. S. Oh, C. G. Yae, D. H. Ahn, and H. B. Cho, “5.8 GHz DSRC Packet Communication System for ITS Services,” IEEE Vehicular Technology Conference, pp. 2223-2227, Sept.
1999.
[4] R. O'Connor, “Control Channel Operation in the ITS Radio Service Band,”
HighwayElectronics, Jan. 2003.
[5] “5 GHz Band Dedicated Short Range Communications: MAC Extension and Lower Layer Management,” ASTM-YYYY, http://www.leearmstrong.com/DSRC/DSRC Homeset.htm
[6] J. Zhu and S. Roy, “MAC for dedicated short range communications in intelligent transport system,” IEEE Commun.Mag., vol.41, Issue.12, pp.60 – 67, Dec. 2003.
[7] H. M. Choi, C. S. Yim, and D. G. Oh, “Advanced DSRC system for supporting mobile IP,” Applications and the Internet Workshops, pp.165 -170, Jan. 2001.
[8] IEEE Trial-Use Recommended Practice for Multi-Vendor Access Point Interoperability via an Inter-Access Point Protocol across Distribution Systems Supporting IEEE 802.11 Operation, IEEE Std 802.11F, July 2003.
[9] E. K. Paik, and Y. H. Choi, “Prediction-Based Fast Handoff for Mobile WLANs,”
ICT2003, vol.1, pp.748-753, Feb. 2003.
IP QoS with NeighborCasting,” ICC 2002, vol.5, pp.3245 – 3249, April 2002.
[11] J. C. Chen, and P. Agrawal, “Fast link layer and intra-domain handoffs for mobile Internet,” Computer Software and Applications Conference, pp.325 – 33025-27, Oct.
2000.
[12] R. Qian, and G. Z. Feng, “A minimum-dwelling-time prioritization scheme for handoff in mobile cellular systems,” ICCT '98, vol.1, pp.1 - 5, Oct. 1998.
[13] S. J. Park, H. J. Lee, J. A. Han, and B. G. Kim, “A deadline scheduling queue handoff scheme for low and high mobility users in wireless ATM networks,” IEEE Pacific Rim Conference, vol.1, pp.198 – 201, Aug. 1997.
[14] E. Y. Ha, Y. H. Choi, and C. S. Kim, “A multicast-based handoff for seamless connection in picocellular networks,” IEEE Asia Pacific Conference, pp.167 – 170, Nov.
1996.
[15] T. Munaka, T. Yamamoto, M. Kuroda, and T. Watanabe, “Advanced join mechanism for multicast group management in DSRC-based ITS networks,” Intelligent Transportation Systems, pp.1147 – 1151, Aug. 2001.
Vita
姓名:石皓棠
學歷:
2002~2004 國立交通大學電信工程研究所
1998~2002 國立交通大學電信工程系
1995~1998 國立師範大學附屬高級中學
E-mail: [email protected]