We developed a simulation program in C to simulate and evaluate the performance of proposed sleep scheme. In this simulation program, a frame is set to 5ms, and the size of resource limit per frame of an MS is fixed, we can set this value as an input parameter beforehand.
There are lots of connections between a BS and an MS, each connection has its own QoS parameters. The program takes the delay constraints, packet inter-arrival time, and packet size of these connections, then derives the availability/unavilability interval of MS and PSCs according to the proposed algorithm. For each connection, its packet will arrive per inter-arrival time, but MS can only transmit/receive packets on the PSCs’ listening window. So most of the packets cannot be transmitted immediately.
The sleep rate (sleep time of the MS / total time) and the statistic of packet delay will be reported. Changing input connections and resource limits can cause different simulation results. In the simulations, we define a connection set as the minimum unit of input connections. There are 8 real-time connections with different QoS parameters in this set as shown in Table.5.1, each connection has random direction (UPLINK/DOWNLINK) and random start frame. Every time we add the number of connections between BS and MS, these eight connections will be added together.
The following are some simulation cases, each has different meaning. Most of simu-lations are compare to PS scheme in reference [9], that’s because the assumption, input and environment of this work is more similar to ours.
Fig.5.1 fixes the resource limit of an MS to 5000 bytes/frame, and increases the
Table 5.1: A connection set.
Type Delay Constraint Data arrival interval Data Size
(ms) (ms) (bytes)
UGS 135 45 160
UGS 105 35 30
UGS 180 60 160
ERT-VR for VoIP services 60 20 60
ERT-VR for VoIP services 90 30 24
ERT-VR for VoIP services 75 25 20
RT-VR 200 30 375
Figure 5.1: (Sleep Time/Total Time) of an MSS while the resource limit is fixed
0
Maximum Resource in a frame (Bytes)
"MPSS Scheme"
"MPSS with Extended Method"
"PS Scheme"
Figure 5.2: (Sleep Time/Total Time) of an MSS while the number of connections is fixed
number of connection sets between BS and MS step by step, to observe the percentage of sleep time of an MS in total time. According to the simulation results, although at first our scheme has slightly less sleep time, but with more and more connections added, our scheme has better sleep rate apparently. That’s because the advantages of using multiple PSCs was revealed, the situation that MS is awake but no packet arrives is rarely appeared.
Differ to Fig.5.1, Fig.5.2 fixes the number of connection sets to eight sets, and changes the resource limit of an MS. When MS has less resource limit, our scheme get better performance. In other words, resources are allocated well when resource is not plenty.
The maximum packet delay is shown on Fig.5.3. The circumstance of this simulation
4
Figure 5.3: Average packet delay (con-nection 1)
Figure 5.4: Jitter of packets (connection 1)
is the same with Fig.5.1. Actually, different connection’s packet has different delay bound, we take one of the connection (connection 1) in the connection set to describe the value will be less or equal to its delay constraint. All the packet delay of this type of connections will be recorded and the maximum one in a long time will be reported. This result shows even if the traffic load is heavy, no packets delay broke the delay constraint. Noted that the connection sets cannot be added without limit, in case of traffic load more than all the resource of an MS, some connections will be denied during admission control.
Fig.5.4 shows the jitter of the connections. Like above, these values are belong to connection 1. The jitter is defined as standard deviation of all packet delays. We can see the curve raised gently with increasing number of connections, but the value of jitter is in the reasonable range all the time. In this figure, the compared method has less jitter because it allocates more resources for packets and MS has longer awake time, arrived packets can be transmitted quickly.
0.5
"Final Sleep Window = 5 frames (Proposed Scheme)"
"Final Sleep Window = 32 frames"
"Final Sleep Window = 64 frames"
"Final Sleep Window = 128 frames"
Figure 5.5: (Sleep Time/Total Time) of an MSS while the data rate of type I connections is λ
"Final Sleep Window = 5 frames (Proposed Scheme)"
"Final Sleep Window = 32 frames"
"Final Sleep Window = 64 frames"
"Final Sleep Window = 128 frames"
Figure 5.6: Response time of the packets of type I connections
Chapter 6. Conclusions
We proposed a sleep scheduling scheme MPSS which conform to the sleep mode mechanism of IEEE 802.16e standard. In this scheme, the connections’ maximum delay constraints, packet inter-arrival time, and packet sizes are considered to determine the parameters of PSCs. The data of a connection can be transmitted during the listening window of corresponding PSCs. Multiple type II PSCs will be conducted for real-time connections, and one type I PSC will be in charge of non-real-time connections. The problem of mismatch between uplink scheduling and listening frames of PSCs is handled.
Besides, an extended method to save more energy but doesn’t guarantee the packet delay constraint is also introduced.
When MS enters sleep mode, we proved the delay of each packet will not larger than its maximum delay constraint, and present the performance and average analysis. The simulation result also shown that our scheme can save more energy of an MS than previous work, and the delay of each packet was not larger than its delay constraint.
This scheme considered the operations of sleep mode in the standard, and doesn’t take heavy computation effort, that is to say it is a practical strategy to save energy of an MS. In the future, we will consider multiple MSs to dynamically adjust the resource limitation in resource reservation procedure instead of fixing the resource of an MS.
Bibliography
[1] IEEE Std. 802.162004. IEEE Standard for Local and Metropolitan Area Networks -Part 16: Air Interface for Fixed Broadband Wireless Access Systems. Oct. 2004.
[2] IEEE Std. 802.16e 2005. Part 16: Air Interface for Fixed and Mobile Broadband Wireless Access Systems - Amendment 2: Physical and Medium Access Control Layers for Combined Fixed and Mobile Operation in Licensed Bands. Feb. 2006.
[3] Y. Xiao. Energy Saving Mechanism in the IEEE 802.16e Wireless MAN. IEEE Communications Letters, 9(7):595–597, July 2005.
[4] Y. Zhang and M. Fujise. Energy Management in the IEEE 802.16e MAC. IEEE Communications Letters, 10(4):311–313, April 2006.
[5] L. Kong and H.-K. Tsang. Performance Study of Power Saving Classes of Type I and II in IEEE 802.16e. In Proc. of the 31st IEEE Conference on Local Computer Networks, pages 20–27, Nov. 2006.
[6] J. Shi, G. Fang, Y. Sun, J. Zhou, Z. Li, and E. Dutkiewicz. Improving Mobile Station Energy Efficiency in IEEE 802.16e WMAN by Burst Scheduling. In IEEE GLOBE-COM, Nov. 2006.
[7] F. Xu, W. Zhong, and Z. Zhou. A Novel Adaptive Energy Saving Mode in IEEE 802.16e System. In Proc. of Military Communications Conference (MILCOM2006), Oct. 2006.
[8] T.-C. Chen, Y.-Y. Chen, and J.-C. Chen. An Efficient Energy Saving Mechanism for IEEE 802.16e Wireless MANs. IEEE Transanctions on Wireless Communications, to appear.
[9] Y.-L. Chen and S.-L. Tsao. Energy-efficient Sleep-mode Operations for Broadband Wireless Access Systems. In IEEE VTC 2006-Fall, 2006.