N UMERICAL R ESULT

Subframe Allocation: Static vs. Dynamic

The first-phase of TPP is advantageous in utilizing the bandwidth when the load of the uplink and downlink are different, as Fig. 7 proves. The FTP traffic load of the downlink is three times of the uplink, and in Fig. 7(a) the downlink utilization is bound to 50% because of the static subframe allocation. However, by stealing the unused uplink slot columns for the downlink, TPP improves the overall link utilization from 75% to 96%. The comparison between static and dynamic allocation of other traffic load ratios is shown in Fig. 7(c).

0

Bandiwidth utilization (%) Total Downlink Uplink

(a) Static subframe allocation. UL:DL = 1:1.

0

(b) Dynamic subframe allocation under the traffic load ratio 3:1 for downlink and uplink.

Ratio (Downlink/Total) Static Dynamic

50% 80% 84%

75% 71% 95%

100% 59% 90%

(c) Different traffic load ratio for static and dynamic allocation.

Fig. 7 Effectiveness of the first phase proportionating.

Effectiveness of the A-Factor

As introduced previously, the A-Factor is used in the second phase to adjust the portion of assigned bandwidth to classes, further contributing to better service differentiation. To understand the effectiveness of employing the A-Factor, we compare it with four schemes which simply use a weight such as R_min, R_max, BRQ, and BRQ–Rmin, for each class. To perform the evaluation, a term named Grant Ratio is defined as the ratio of number of requested slots to the number of allocated ones. A grant ratio larger than 1 means that the service class is allocated more than requested, resulting in bandwidth waste. On the one hand, as presented in Fig. 8, the Grant Ratios of rtPS using Rmin, Rmax and BRQ are about 1.2, implying excessive allocations, while appropriate amounts are provided when using the A-Factor and BRQ–R_min. On the other hand, the nrtPS using Rmax with A-Factor obtains more slots than those in other schemes. In BE, though the one using BRQ–R_min has the highest Grant Ratio, this scheme is not feasible because it tends to favor classes with small Rmin which oftentimes is the BE, and therefore violates the spirit of service differentiation.

0

Fig. 8 Effectiveness of using A-Factor. Four schemes with simple weights are involved for comparison.

Service differentiation

Figure 9(a) and 9(b) display the minimum reserved slots as well as the average

delay for each class under different numbers of SSs. As we can see in Fig. 9(a), the UGS and ertPS sustain the number of reserved slots even when the number of SSs advances 60. For other classes, the system guarantees the differentiated R_min, namely 4:2:1, until the number of SSs exceeds 50. For the average delay depicted in Fig. 9(b), only minor difference is observed among classes initially until the number of SSs reaches 40, rather than 50. This is because not enough additional slots can be allocated but only the minimum requirement is satisfied. Again, the delay of the UGS and ertPS are always kept under 10ms.

0 2 4 6 8 10

10 20 30 40 50 60

(a) The variation of minimum reserved slots and granted slots of each request under each class.

N umber of SSs

Number of allocated slots per reques

UGS ertPS rtPS-min

nrtPS-min BE-min rtPS-total

t

nrtPS-total BE-total

1

(b) Average delay between service classes.

Fig. 9 Service differentiation.

Performance

The performance of the TPP is compared with the Deficit Fair Priority Queue (DFPQ) and Strict Priority (SP) in terms of bandwidth utilization, as depicted in Fig.

10(a). From the figure we can learn that the bandwidth utilizations of the three algorithms increase linearly but start to decrease when hitting a certain level: 85.5%

for TPP, 80.6% for DFPQ and 68.4% for SP. The reason why they are not fully utilized is explored by looking into the average Frame occupation of service classes, as presented in Fig. 10(b). Each class has an unused portion, which occurs during the translation from requested bytes to slots. Since the calculation, namely dividing the requested bytes by slot size, always rounds up, the resulted assignment is often larger than expected. For example as shown in Fig. 10(c), assuming 64 bytes in a slot and the requested size by service flow (SF) #1 is 213 bytes, the number of requested slots is thus 4, causing a 256-213=43 bytes waste. However, the TPP alleviates this effect by reserving minimum required slots first, rather than paying up all requested slots at once for an SF. Take Fig. 10(c) for instance and assume that the number of available slots is nine and the MRTR of each SF is three, TPP breadth-firstly allocates every SF three slots which are slightly insufficient but allocated slots are not wasted;

nonetheless, the DFPQ depth-firstly tries to satisfy all SF’s full requested slots but results in the waste for the first two SFs and the starvation of the third SF having lowest priority. The SP has a largest waste also because of its static allocation.

Besides, the UGS contributes to the relatively massive amount of unused portion than other classes, revealing the drawback of unnecessary slot reservation. Finally, aside the high efficiency in bandwidth consumption, TPP is advantageous in service differentiation. As depicted in Fig. 10(b), the ratio of allocated bandwidth for rtPS, nrtPS and BE is very close to 4:2:1, compared to other two algorithms.

0

(b) Frame occupation under three schemes. Number of SSs is 48.

(c) Example of allocation by TPP and DFPQ. The number of slots which are to be allocated to three service flows is 9.

Fig. 10 Performance comparison with SP and DFPQ.

Chapter 5 Conclusions and Future Works

This work aims at proper bandwidth allocation for 802.16 in order to well utilize the precious wireless link and to support service differentiation. The GPSS is adopted not only to comply with the standard but to provide SSs the flexibility of manipulating the assigned bandwidth. The uplink and downlink bandwidth allocations are considered at the same time so that the allocation can be dynamically adjusted according to the demand of both links.

The Two-Phase Proportionating (TPP) is proposed to achieve the above goals.

Considering the different slot definitions in uplink and downlink, the first phase proportionates the two links according to their accumulated requested sizes. In the second phase, after assigning the minimum reserved slots, the weights of accumulated QoS parameters for service classes are involved to proportionate the remaining slots of the subframes. An adjustment factor, A-Factor, is further adopted to complement the weight parameter and reflect different requested amount of the classes. Finally, each SS obtains its share from all service classes.

The simulation result indicates that the bandwidth utilization increases 20% by applying the first phase proportionating, compared to static allocation. We also show that the A-Factor outperforms other four schemes in preventing from bandwidth waste and differentiating classes. For service differentiation in terms of minimum reserved slots and delay, it is shown that the UGS and rtPS are guaranteed even when a large number of SSs are present.

Given that the service differentiation is carried out in BS, the SSs should also be able to provide similar support in order to meet the QoS requirement of various applications. Therefore, the future work will be focusing on designing a sophisticated