FHP Simulation Results and Discussions

In the simulations, the topology of the cellular IEEE 802.11e WLAN systems contains 7 × 7 hexagonal and wrap-around QBSSs . Arbitrary two adjacent QBSSs are assumed to use different physical channels. The radius of coverage of each QAP is 50 meters, and the distance between any two neighboring QAPs is 80 meters. A two-path Rayleigh fading channel model [63] is also considered. The WLAN system and its parameters are based on those in [2, 50, 52], where the SIFS, PIFS, and DIFS are assumed to be 10µs, 20µs, and 40µs, respectively; a beacon interval is 20ms; the duration of CFP is fixed to be 10ms; and a slot time (aSlotTime) of PHY is 9µs. The slot time in the CCP is given as 78µs, which is sufficient for a round of an HO-REQ symbol and its ACK.

Two scenarios with different number of contention-based, prioritized static QS-TAs, besides the HO-REQs, are assumed in the simulations. In scenario 1 (2), there are 5 (4) QSTAs with background access category (AC BK), 5 (4) QSTAs with

best-effort access category (AC BE), 1 (2) QSTA(s) with video access category (AC VI), and 1 (2) QSTA(s) with non-handoff voice access category (AC VO) [52]. All static QSTAs are located randomly and activated in a saturation mode of that their ac-cess transmissions are always activated. The arrival proac-cess of the HO-REQ in each QBSS is assumed to be in Poisson distribution. Each HO-REQ has to seek for a successful transmission under the 100ms system delay bound and the 8 times retry limit (dot11LongRetryLimit). Otherwise, the HO-REQ will be forcedly terminated.

Noticeably, the delay is the time difference between the HO-REQ arrival and its successful access.

In the aspect of the FAM design, we choose the logarithm function, log₁₀k, k=1, 2,..., 10, for the 4-tuple elements of A_X and B_Y contained in (3.23) and (3.24), respectively. By such a way, the ranges of trapezoidal functions are wider when measures of u or η are lower, and thus FAM would be more sensitive to the worse conditions in u and η. The trapezoidal fuzzy set ranges of A_X are set with A_VL=(0, 0, log₁₀2, log₁₀3), A_L=(log₁₀2, log₁₀3, log₁₀4, log₁₀5), A_M=(log₁₀4, log₁₀5, log₁₀6, log107), A_H=(log106, log107, log108, log109), and A_VH=(log108, log109, 1, 1). The same settings are also applied to B_Y. The values of Z for the terms HD, MD, LD, NC, LI, MI, and HI are set with 0.25, 0.5, 0.75, 1, 1.25, 1.5, and 2, respectively.

The FHP will be compared with the conventional enhanced distributed channel access (EDCA) method [52], where both HO-REQs and other packets use EDCA to access in CP, but HO-REQs are given with the highest priority (AC VO).

Fig. 3.9 shows the mean forced termination rate of HO-REQs caused by the overstep of either retry limit (8 times) or system delay bound (100ms). It can be found that the FHP provides an almost zero forced termination rate for HO-REQs, and the performances of FHP are the same in scenarios 1 and 2 . The reasons are that

Figure 3.9: Mean forced termination rate of HO-REQs

the FHP designs a CCP dedicatedly designated for HO-REQs and provides a uniform separation for these HO-REQs access, which can prevent HO-REQs from colliding with each other in CCP; also the access of HO-REQs in CCP of FHP are not affected by the contention-based QSTAs in CP. This is a great advantage for the entire handoff process in cellular WLAN systems because the uncertainty of the media access delay of HO-REQs can be eliminated and WLAN systems can obtain more accurate time for other pro-active handoff procedures. The conventional EDCA method, however, attains a large forced termination rate for HO-REQs, which is usually required rather low, say 5×10⁻³ at least. In scenario 1, its forced termination rate is about 1.2×10⁻² when there is one arrival of the new HO-REQs in average per BI. It further exceeds 6 × 10⁻² when the average number of new HO-REQ arrivals per BI is higher then 3.

Performance in scenario 2 is worse than that in scenario 1. The reasons are that the HO-REQs have to contend with static QSTAs in CP and waste time in backoff, and

Figure 3.10: System throughput of CP and CCP utilization

the more QSTAs with higher priority such as AC VI and AC VO in the system, the worse the forced termination rate of HO-REQs would be.

Fig. 3.10 depicts the system throughput of CP (left vertical axis) and the CCP utilization (right vertical axis), where the system throughput of CP by FHP includes those in CCP and CP, as shown in Fig. 3.6. It can be seen that FHP can still achieve higher system throughput of CP than the conventional EDCA, and the superiority is more significant when the number of high-priority static QSTAs becomes larger (scenario 2). The reason is that FHP designs a dedicated CCP for HO-REQs, which will result in a smaller number of high priority users (only AC VO) in CP and thus less backoff and collisions for other contention-based packets. The mutual influence between HO-REQs and other contention-based packets can be decreased. Also, FAM provides a high CCP utilization for HO-REQs, which is over 0.75. This implies that there are not too many slots wasted in CCP and the time duration of CCP partitioned

from CP is well-controlled by the proposed FAM. Moreover, the system throughput of CP using the conventional EDCA is affected greatly by the number of HO-REQs, and the influence is more deteriorated if more static QSTAs with higher priority exist, as in scenario 2. The phenomenon somewhat justifies the reason we mentioned in this paragraph, which denotes that more high priority QSTAs, including static and handoff, would lead to higher probabilities of backoff and collision in CP.

3.4 Concluding Remarks

In this chapter, an intuitive scheduling and admission control (ISAC) scheme is proposed based on IEEE 802.11e cellular WLAN systems. The ISAC scheme con-siders admission control, based on not only the quality of service (QoS) required by each mobile user, but also the link quality of air interface influenced by fading, noise, and interference. Moreover, a standard-compatible fast handoff protocol (FHP) is proposed for cellular IEEE 802.11e WLAN systems. It consists of a controlled contention period (CCP), which is partitioned from the contention period (CP) for HO-REQs, and a fuzzy adjustment method (FAM) for handoff requests (HO-REQs), which adaptively adjust the proper length of CCP. Simulation results show that the FHP can decrease the forced termination rate of HO-REQs and still enhance the system throughput of CP. This major advantage brought by FHP, which eliminates the uncertainty of media access delay for HO-REQs, is significant. As a result, the WLAN systems can obtain more accurate time estimations for other active pro-cedures in the entire handoff process. This cannot be achieved by using conventional enhanced distributed channel access (EDCA) method in the standards.

Chapter 4

A Fuzzy Q-Learning Admission