A handoff flow with different type of service

In this simulation, there is only one MH in the simulation and do the handoff during the simulation. The MH may have a voice flow or a video flow. The traffic loads in both IEEE 802.11e and IEEE 802.16e are heavy. The number of flows in IEEE 802.11e is at least ten flows and the number of flows in the IEEE 802.16e is at least sixty flows.

First we show the result when the flow is a voice flow. The average throughput in IEEE 802.11e and the IEEE 802.16e are showed in the below.

The MH is in the IEEE 802.16e first and starts to move. It moves to the IEEE 802.16e at 9.6, 21.8, 27.8, 44, 53.4, 57.9, 76.9, 84.2, 95.6 second and moves to IEEE

802.11e at 8.3, 17, 24.1, 38.8, 48.1, 56.8, 68.6, 81.2, 88.3 second.

The simulation does not have any traffic first and will start to set up every flow at first 7 second.

In IEEE 802.11e, when a voice flow is set up or set down, the voice’s throughput (Prio_ 0) does not have obvious change. The real time video’s throughput (Prio_ 1) has more obvious change, but is not by reason of the handoff voice flow. The really reason is the other video flow’s start or end at that time and the voice connection only need a little resource and the IEEE 802.11e use the contention process to content the resource. The backoff time is randomly selected. So the throughput is not guaranteed.

The vertical broken line in Figure 3-2 is the time point that the MH to 802.11e network. We only show the time that the throughput has obvious change.

Figure 3-3 to Figure 3-10 are the average throughput of the different types of services in 802.16e. The vertical dashed lines in the figures are the time points that the MH handoff to the 802.16e network. Although the traffic load is heavy, the network can still satisfy every flow’s minimum QoS requirement and the residual resource can be shared out to the flows that the maximum QoS requirement are not satisfied. The voice connection need very small requirement, so the throughput of other services just a little degradation.

0

Figure 3-2 The average throughput of three type of service flows in IEEE 802.11e

0

Figure 3-3 The average throughput of the UGS

0

Figure 3-4 The average throughput of the rtPS type 1

0

Figure 3-5 The average throughput of the rtPS type 2

0

Figure 3-6 The average throughput of the rtPS type 3

0

Figure 3-7 The average throughput of the ertPS type 1

0

Figure 3-8 The average throughput of the ertPS type 2

0

Figure 3-9 The average throughput of the ertPS type 3

0 5 10 15 20 25 30 35 40

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 Time (sec)

Throughput (kbps)

nrtPS-1 nrtPS-2

min. requirement

Figure 3-10 The average throughput of the nrtPS

In Figures 3-11 and 3-12, the average delay in 802.11e is less than that in the 802.16e network. It is because in 802.11e networks, many packets will be dropped after exceeding the retry limit. On the other hand, the packets of the flows in 802.16e network will be queued in the buffer, until the nodes receive the grant from the BS, or queued over the maximum latency time than drop the packet. But in the heavy traffic load environment, the nodes can’t get enough grants to transmit all the packets. So the delay time is almost equal to the maximum latency time.

0

Figure 3-11 The average delay of service flow in IEEE 802.11e

0

Figure 3-12 The average delay of service flow in IEEE 802.16e

Figures 3-13 and 3-14 are the throughput and delay time of the voice in the MH.

Here we can see that when the flow handoff to IEEE 802.16e network or to the IEEE 802.11e network, the effect of the existed flow is not obvious.

The MH is in the IEEE 802.11e network first and starts to move. It moves to the IEEE 802.16e at 14.2, 31.5, 57.8, 66.9, 74.6, 79.7, 96.9 second and moves to IEEE 802.11e at 30.2, 38.5, 61.2, 70.5, 77.6, 86.1 second.

40

Figure 3-13 The average throughput of the HO voice flow

0

Figure 3-14 The average delay of the HO voice flow

The other scenario is a MH with a video flow and may handoff to another network. Like the scenario in the before, the vertical broken lines in Figure 3-15 are the handoff time in 802.11e. And the video flows’ effect is not caused by the handoff video flow, is the start or end of other video flows. The voice flows in the IEEE 802.11e are also not affected because these flows have high priority and will select a little backoff time.

From Figure 3-16 to Figure 3-23 are the average throughput in 802.16e. When the MH is at the IEEE 802.16 network, the average throughput of UGS is not affected, but the other services are a little decrease and still satisfy their minimum QoS requirement. But the decrease is more obvious than the scenario before.

0

Figure 3-15 The average throughput of different service flow in IEEE 802.11e

0 10 20 30 40 50 60

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 Time (sec)

Throughput (kbps)

UGS

UGS requirement

Figure 3-16 The average throughput of the UGS

0 50 100 150 200 250 300 350

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95 100 Time (sec)

Throughput (kbps)

rtPS-1

min. requirement

Figure 3-17 The average throughput of the rtPS type 1

0

Figure 3-18 The average throughput of the rtPS type 2

0

Figure 3-19 The average throughput of the rtPS type 3

0

Figure 3-20 The average throughput of the ertPS type 1

0

Figure 3-21 The average throughput of the ertPS type 2

0

Figure 3-22 The average throughput of the ertPS type 3

0

Figure 3-23 The average throughput of the nrtPS

Figure 3-24 and 3-25 are the average delay. In IEEE 802.11e network, when the MH move to the IEEE 802.11e network, the delay time of priority 0 does not increase, but the flows with priority 1 or priority 2 will increase the delay time a little. The obvious variation is because some flows are stop at that time. In IEEE 802.16e network, the delay time does not change obviously when the MH moves to the IEEE 802.16e or the other flows are stop.

Figure 3-26 and 3-27 are the throughput and delay time of the video in the MH Like the voice flow, the video flow in IEEE 802.16e network is with larger delay time than the IEEE 802.11e network with the same reason.

In the simulation, although the throughput and the delay time in IEEE 802.16e work does not better than IEEE 802.11e network, but the curve does not change obviously. This means that the IEEE 802.16e network can apply a stable environment, and can make sure that the new flow can achieve their QoS requirement.

0

Figure 3-24 The average delay of the service flow in IEEE 802.11e

0

Figure 3-25 The average delay of the service flow in 802.16e

0

Figure 3-26 The average throughput of the HO video flow

0

Figure 3-27 The average delay of the HO video flow