無線網路之媒體接取控制與排程機制

國 立 交 通 大 學

電機工程學系

博 士 論 文

無線網路之媒體接取控制與排程機制

Medium Access Control and Scheduling

Schemes for Wireless Networks

研究生：顏志明

指導教授：張仲儒 博士

無線網路之媒體接取控制與排程機制

學生：顏志明 指導教授：張仲儒

國立交通大學電機工程學系

摘要

題。它可利用不同的仲裁訊框間隔 (Arbitration Inter Frame Space) 來區分不同的服 務以提升服務品質，但是效能並不是很好。為了要在無線區域網路保證多媒體服務

的服務品質，提出一個適應性p-persistentr (APP) 基礎的媒體接取控制法。APP媒體 接取控制法依照不同的服務種類給予不同的允諾機率來服務多媒體使用者並利用 傳送的允諾機率來區分使用者。當使用者有較大的延遲封包他會有較高的允諾機 率。從分析與模擬結果中証實，APP可以降低使用者間的延遲變異達15%，並降低 高優先權服務的失敗率。 在IEEE 802.16 都會型網路中，基地台要服務大量的使用者與提供不同種類的 服務型態，所以需要排程機制提升系統效能，並兼顧使用者的服務品質需求。我們 在IEEE 802.16 上 鏈 路 提 出 了 一 個 基 於 動 態 優 先 次 序 的 資 源 分 配 (dynamic

降低複雜度的功效式排程機制 (utility-based throughput maximization and complexity reduction (U_TMCR) scheduling)。

DPRA機制對於急迫性較高的服務，我們給予較高的優先次序值(Priority value)，使具有較高優先次序的使用者能優先被分配系統資源做傳輸。我們也會根 據每一種服務在不同時間的急迫性，動態調整其優先次序。我們提出的DPRA機制 會在子通道(subchannel), 調變方法(modulation order), 以及能量(power)三方面找尋 最佳化的資源分配方法，並且對同一個使用者做一致性分配(consistent allocation)。 由模擬結果顯示，我們提出的方法可以達到傳輸速率最佳化以及QoS的滿足，並且 能減少標頭傳輸(transmission overhead)以及降低運算複雜度。 U_TMCR程機制不只在保證服務品質 (QoS) 的情形下最大化系統效能，同時 也降低計算複雜度，並針對多媒體使用者做通道配置、天線選擇與決定調變方法。 U_TMCR 機制根據通道品質和使用者的服務品質需求，為每個使用者設計效用函 數 (utility function)，將排程問題轉成考慮系統限制對整個系統最佳化效用的問題。 U_TMCR 機制也提出一個低複雜度演算法來解決所提出的最佳化問題。由模擬驗 證，U_TMCR可以提升8%的系統效能並降低6.25%~29.2%的計算複雜度。

Medium Access Control and Scheduling

Schemes for Wireless Networks

Student: Chih-Ming Yen Advisor: Chung-Ju Chang

Institute of Electrical Engineering

National Chiao Tung University

Abstract

To guarantee the quality of service (QoS) in wireless network is an important issue. Currently, the different networks are optimized for different services, but it becomes urgent to provide varied service in wireless networks. To address this problem, there need to re-design a medium access control (MAC) or scheduling scheme to satisfy the QoS of multimedia service and to enhance the network utilization.

In the WLAN, the goal of the medium access control (MAC) protocol is to deal with the contention of stations. It uses the different arbitration inter frame space to differentiate the services to promote the service quality, but the QoS satisfaction is not good enough. In order to support multimedia services in the WLAN, an adaptive

p-persistent-based (APP) MAC scheme for IEEE 802.11 WLAN is investigated. The

APP MAC scheme can further differentiate priorities of access categories by the initial permission probabilities and adaptively adjust permission probabilities to transmission stations according to its transmission state. Numerical and simulation results show that the APP MAC scheme can reduce the dropping probability of high priority service and effectively reduce the delay variance by 15%.

802.16 WiMAX system. Therefore, it needs to elaborately design the scheduling scheme to enhance QoS satisfaction with high system efficiency. In the IEEE 802.16 system, a dynamic priority resource allocation (DPRA) scheme for uplink and a utility-based throughput maximization and complexity reduction (U_TMCR) scheduling scheme for downlink are investigated.

The DPRA scheme dynamically gives priority values to difference services based on the urgency degrees and allocates system radio resources according to the priority values. It can maximize the system throughput and satisfy differentiated QoS requirements. Also, the DPRA scheme performs consistent allocation to conform the uplink frame structure of IEEE 802.16, to fulfill QoS requirement, and to reduce the computational complexity. Simulation results show that the proposed DPRA scheme performs very close to the optimal method, which is by exhaustive search, in system throughput; and it outperforms the conventional EFS algorithm [39] in the performance measures such as system throughput, rtPS packet dropping rate, ratio of unsatisfied nrtPS, and average transmission rate of BE.

The goals of the U_TMCR scheme are not only to maximize system throughput under QoS guarantee but also to reduce computational complexity. Based on channel quality and QoS requirements of each user, the U_TMCR scheme designs a utility function for every user and formulates the scheduling into an optimization problem of overall system utility function subject to system constraints. It also contains a heuristic TMCR algorithm to efficiently solve the optimization problem. Simulation results show that the U_TMCR scheme can improve the system throughput by 8% and reduce the computational complexity by 6.25%~29.2%.

Acknowledgments

I would like to express my sincere gratitude to my advisor, Dr. Chung-Ju Chang, for his great care and guidance at research and the methodology of schedule throughout my school time.

I am also indebted to all friends in BCN Lab. Thank you all for your kind help. This dissertation is dedicated to my family for their understanding and encouragement.

Contents

Chinese Abstract...i

Abstract ... iii

Acknowledgments...v

Contents...vi

List of Figures ... viii

List of Tables ...ix

Chapter 1 ...1 1.1 Motivation ...1 1.2 Literature Survey...3 1.3 Dissertation Organization...7 Chapter 2 ...9 2.2 System Models ...13

2.3 The Adaptive P-Persistent (APP) MAC Scheme ...14

2.4 Analysis ...16

2.4.1 System Throughput ...20

2.4.2 Delay ...21

2.4.3 The Optimal Value of P0...22

2.5 Numerical and Simulation Results...22

2.5.1 Data Only Environment ...22

2.5.2 Multimedia Service Environment ...28

2.6 Concluding Remarks ...34

Chapter 3 ...36

3.1 Introduction ...36

3.3 Dynamic Priority-based Resource Allocation Scheme ...45

3.3.1 Problem Formulation...45

3.3.2 DPRA Scheme...47

3.4 Simulation Results...54

3.4.2 Source Model and QoS Requirements ...54 3.4.3 Performance Evaluation ...56 3.5 Concluding Remarks ...63 Chapter 4 ...64 4.1 Introduction ...64 4.2 System Model...68 4.2.1 System Assumptions ...70

4.3 Utility-based TMCR Scheduling Scheme ...73

4.3.1 Utility Function ...73 4.3.2 Problem Formulation...75 4.3.3 Heuristic TMCR Algorithm ...76 4.4 Simulation Results...80 4.5 Concluding Remarks ...89 Chapter 5 ...91 Bibliography...96 Vita ...103

List of Figures

Figure 2.1 State transition diagrams for the APP MAC scheme ...18

Figure 2.2 Collision probabilities of APP, BEB, and DDFC...24

Figure 2.3 System throughputs of APP, BEB, and DDFC...25

Figure 2.4 Mean delays of APP, BEB, and DDFC ...26

Figure 2.5 Delay variances of APP, BEB, and DDFC...27

Figure 2.6 performance of APP with optimal * 0 P and BEB with Wopt...28

Figure 2.8 System throughput ...32

Figure 2.9 (a) Mean delay and (b) delay variance of low priority packet...33

Figure 3.1 Flow chart of the DPRA scheme...53

Figure 3.2 System Throughput...58

Figure 3.3 Voice Packet Dropping Rate ...59

Figure 3.4 Video Packet Dropping Rate...59

Figure 3.5 Ratio of Unsatisfied HTTP Users ...60

Figure 3.6 FTP Average Transmission Rate ...61

Figure 3.7 Average Number of Iterations ...62

Figure 4.1. System configuration of the downlink MIMO-OFDMA system...69

Figure 4.2. Flow chart of the heuristic TMCR algorithm ...77

Figure 4.3 System Throughput ...84

Figure 4.4 (a) Voice packet dropping rate; (b) Video packet dropping rate ...85

Figure 4.5 (a) Mean delay of voice packet; (b) Mean delay of video packet...86

Figure 4.6 Guaranteed ratio of HTTP packets...87

List of Tables

Table 2.1 Parameter Settings for a WLAN Environment...23

Table 3.1 System-Level Parameters ...54

Table 3.2 QoS Requirements of each service type...55

Table 4.1 System-Level Parameters ...81

Chapter 1

Introduction

1.1 Motivation

The success of second-generation (2G) mobile system in the pass decades gives rise to the development of third-generation (3G) mobile system. For example, global system for mobile communication (GSM ) and IS-95 from 2G system designed to carry speech and low-bit-rate data has been upgraded to provide higher data-rate services on their 3G version. Besides, the range of wireless system, such as general packet radio service (GPRS), IMT-2000, Bluetooth, and IEEE 802.11 wireless local area networks (WLANs) have as well developed from 2G to 3G. All those system were separately designed to meet the needs of a variety of service types, data rates, and users. Having just one system can never be sufficient enough to be substituted for all other technologies since every single system has its own merits and relative drawbacks.

Mobile cellular networks have progressed to 3G within two decades due to the advancement of wireless technologies and the emergence of multimedia data services. People have pictured the prospects of wireless networks as the integration of lots of existing and newly developed wireless access networks such as WLAN, IEEE 802.16

wireless metropolitan area networks (WMANs), GPRS, and universal mobile telecommunications system (UMTS). The tendency towards packet-switched technologies, and the increasing use and acceptance of the internet protocol (IP) illustrates that an IP core network is going to connect different wireless access networks together. Therefore, based on that conception, several overlapping IP-based wireless access network domains will be the future of all-IP wireless networks.

The future wireless networks have some evident features listed as the following. Firstly, all-IP based hybrid networks which allow users to use any system at anytime and anywhere are the future of wireless networks. Multiple wireless networks provide users carrying an integrated terminal with a wide range of applications. Secondly, not only telecommunications services but also data and multimedia services are provided by the future wireless systems. High-data-rate services with good system reliability will be provided in order to support multimedia services, and, simultaneously, the cost of a low per-bit transmission will stay maintained. Thirdly, the future network will provide personalized service. It is anticipated that users in widely different locations, occupations, and economic classes will use the services while the future wireless services are launched. For the reason of catering the needs of diverse users, personal and customized services should be designed. Finally, facilities for integrated services will also be provided by the future wireless systems. Users can use multiple services at a time from any service provider.

There are two types of existing wireless systems: IP-based and non-IP-based. IP-based systems are usually optimized for data services (e.g., IEEE 802.11 WLAN). As for non- IP-based systems, many of which are highly optimized for voice delivery (e.g., GSM, cdma2000, and UMTS). In order to integrate the multimedia service, one major

particularly considering time-sensitive or multimedia applications, the quality of service (QoS) guarantee for end-to-end services needs to be stated.

1.2 Literature Survey

In personal area network (PAN), high transmission rate and low design complexity in medium access control (MAC) protocol are the benefits that wireless local area networks (WLAN) possess. Hot spot cells and indoor environments widely apply it for diverse applications. Due to channel sharing, the MAC protocol is the key role to determine the efficiency and performance of the WLAN. The Abramson proposed an elegant MAC protocol, called ALOHA [1]. In ALOHA, high collision probability result form stations allowed transmitting immediately upon receiving data from upper layers. To decrease the collision probability, carrier sense multiple access (CSMA) scheme [2] requires stations to transmit until the medium becomes idle. When a station detects the channel is idle, it can transmit with a probability of 1 or p (0 < p < 1). The former is called 1-persistent CSMA and the latter is p-persistent CSMA. In the IEEE 802.11 [3], it adapt the CA scheme of DCF further reduces frame collision probability by requiring each backlogged station to perform binary exponential backoff (BEB) after the medium becomes idle. In BEB, if a station successfully transmits a frame, its contention window will be reset to an initial value. However, if the transmission fails, the window size is doubled. Nevertheless, owing to the situation that collided stations would have smaller probability to access the medium than new stations, it generally entails larger delay variance on stations and results in unfairness.

The important and challenging issue is to divide the channel fairly among stations. Therefore, the design of efficient MAC protocols with high-throughput performance, as well as a high degree of fairness performance, has been a major focus in WLAN

research areas [4], [5], [49]-[51]. There have been many studies of backoff algorithms [6]-[13], [16], [45], but all of them did not address the problem of larger delay variance.

For real-time applications, higher delay variance leads to larger amount of dropped packets due to excess delay. For non-real-time data applications, on the other hand, higher delay variance usually causes larger requirement of storage buffer or more probability of buffer overflow. Hence, in order to reduce the delay variance, the radio resource over WLAN interface is necessarily to be fairly shared by an effective MAC protocol. The fairness problem of MAC in WLAN that some algorithms solve was proposed [14], [15]. However, channel throughput is decreased because of the high collision probabilities.

Dynamic contention window (CW) schemes [16]-[18], different maximum packet length scheme [18], and various interframe space (IFS) schemes [18]-[20] are usually adopted to design the priority differentiation in order to support multimedia services for the IEEE 802.11e [52] WLAN. However, owing to the backoff scheme, large delay variance in the same access category (AC) would still be arisen by these solutions. Obviously, larger probability of quality-of-service (QoS) violation of multimedia traffic is brought about by higher delay variance because of excess delay.

In WMANs, the problem of future wireless communication is resolved by multiple-input multiple-output (MIMO) based orthogonal frequency division multiplexing (OFDM) because it helps achieve high system capacity and provide transmit/receiver diversity for reliable communication link. Downlink resource management for multiuser OFDM (MU-OFDM) systems has recently been investigated [21]-[25], in which topics were emphasized on transmission power allocation, subcarrier allocation, bit allocation, or adaptive modulation and coding (AMC). The goal of the

fairness, or guarantee QoS requirements.

Many papers investigated the downlink resource allocation [21]-[32] but few papers probed into the uplink resource allocation. Both downlink and uplink perform the resource allocation primarily through the base station (BS). The power distribution over the selected set of subcarriers for every user is included in the algorithm in [35] so that it minimizes the total power being used. In [36], a greedy subcarrier allocation algorithm, based on a marginal rate function, and an iterative water-filling power allocation algorithm were proposed. A practical algorithm and the optimization problem were presented in [37]. All of them can nearly reach an optimal solution but they did not focus more on the QoS requirement. The power saving in IEEE 802.16 OFDMA systems via an efficient uplink resource allocation was shown in [38]. While guaranteeing BER, it minimizes the required transmissions power through adaptively adjusting the modulation and coding scheme. However, they don’t attend to their differentiated QoS requirements and multiple services. [39] exhibited an efficient and fair scheduling (EFS) algorithm for each time slot in IEEE 802.16 OFDMA/TDD system. A fixed priority scheme which gives priorities to service traffic according to their QoS requirements is applied to design the EFS algorithm.

The bandwidth is allocated according to channel quality and queue state of the traffic for SS with real-time and non-real-time polling services in [40]. From the previously mentioned works, people either omitted or simplified the QoS requirements and fairness issues. The QoS requirement usually refers to a predefined weight which corresponds to the fixed priority scheme or a minimum required transmission data rate. However, the design of radio resource allocation for practical applications should include the delay bound and the packet dropping rate, regarded as essential QoS requirements to provide multimedia real-time traffic. In addition, buffer conditions of

different traffic types and realistic traffic models should be taken into account.

For MIMO-OFDM systems, the exponential increases in the computational complexity on radio resource scheduling for downlink multiuser is proportional to the number of subcarrier, multiuser, transmitting antenna, and receiving antenna. The multiuser scheduling algorithms for system throughput maximization with reduced complexity in a downlink MIMO/OFDMA system were proposed [26]-[29]. They decoupled the multiuser scheduling problem into frequency and spatial domains. Multiple parallel independent single-user MIMO channels are decomposed from the multiuser downlink MIMO channel by the preprocessing scheme. However, the number of transmitting antennas restrains the number of simultaneously transmitted users. Computational complexity of the scheduling algorithm is still too high and the QoS requirements and user demand were not considered in the scheduling algorithm.

A fixed priority algorithm was proposed [30] in relation to the QoS requirement in MU-MIMO-OFDMA system. The non real-time traffic’s rate of transmission is too low to fulfill the requirement rate while the real-time traffic can be provided in time at low traffic intensity. A dynamic priority scheduling scheme was proposed in [32]. With that scheme, not only is high priority given for urgent users but also the priority of users is dynamically adjusted frame by frame. However, the ARRA does not give the clear differentiation of real-time service from non-real time one but depends on the time to expiration while adjusting the priority.

This is also an important issue: the tradeoff between system performance and computational complexity. Optimal solution [41] can be achieved by the greedy algorithm which performs symbol by symbol allocation, but it causes high computational complexity. The symbol-by-symbol allocation algorithm costs high

defined in IEEE 802.16 for downlink and uplink, respectively. Moreover, most resource allocation algorithms are not only designed for downlink but also claimed to be compatible with uplink. Even so, both the uplink frame structure (UL-MAP) and downlink frame structure (DL-MAP) have different definitions in IEEE 802.16 specifications [42]. Therefore, to meet its individual frame structures, a design of an efficient and feasible resource allocation algorithm for either downlink or uplink is particularly needed.

1.3 Dissertation Organization

In this dissertation, the radio resource allocation schemes in wireless network are studied, and the Qos guarantee radio resource management schemes are investigated in both PAN and WMAN.

For IEEE 802.11 WLAN to provide low delay variance, an adaptive p-persistent-based (APP) medium access control (MAC) scheme [46]-[48] is presented in Chapter 2. Permission probabilities of transmission for stations being incurred with different packet delays can be differentiated through the APP MAC scheme and it is designed as a function of the numbers of retransmissions and re-backoffs so that stations with larger packet delay can have higher permission probability. Moreover, the scheme is modeled by a Markov-chain and successfully analyzed, in which the system throughput and delay are derived from. For multimedia services, the APP MAC scheme adaptively gives transmission stations which are in different access category and with various waiting delay differentiated permission probabilities.

For uplinks in IEEE 802.16 wireless communication systems, a dynamic priority resource allocation (DPRA) scheme [33]-[34] is proposed in Chapter 4. The DPRA scheme dynamically gives four types of service traffic based on their urgency degrees

priority values and allocates system radio resources according to their priority values. It can satisfy differentiated QoS requirements and maximize the system throughput. Also, in order for packets of users to conform the uplink frame structure of IEEE 802.16 to fulfill QoS requirement and reduce the computational complexity, the DPRA scheme performs consistent allocation.

For downlink multiuser MIMO-OFDMA systems, a utility-based throughput maximization and complexity reduction (U_TMCR) scheduling scheme [43]-[44] is proposed in Chapter 3. The U_TMCR scheme allocates subchannels, antenna sequence, and modulation order to multimedia users with goals not only to reduce computational complexity but also to maximize system throughput under QoS guarantee. Based on each user’s channel quality and QoS requirements, both a utility function for every user is designed and the scheduling is formulated into an optimization problem of overall system utility function subject to system constraints by the U_TMCR scheme. It also contains a heuristic TMCR algorithm for efficiently solving the optimization problem.

Finally, the conclusive statements and future research topics are addressed in Chapter 5.

Chapter 2

Analysis of an Adaptive P-Persistent

MAC Scheme for WLAN Providing

Delay Fairness

2.1 Introduction

In the IEEE 802.11, the fundamental mechanism to access the medium is called distributed coordination function (DCF), which is based on carrier sense multiple access with collision avoidance (CSMA/CA) protocol. Retransmissions of collided packets are managed by binary exponential backoff (BEB) rules. The IEEE standard also defines an optional point coordination function (PCF), which is a centralized MAC protocol to support collision free and time bounded services. Both the DCF and PCF can operate concurrently with the same basic service set (BSS) to alternate contention and contention-free periods.

state firstly. The station transmits only if the channel is idle for a period of time equal to a DCF inter frame space (DIFS). Otherwise, the station persists to monitor the channel until the measured idle period equals a DIFS. Additionally, the DCF also adopts a BEB scheme to avoid the occurrence of packet collision.

Traditional MAC scheme accompanying with the BEB algorithm is one of the most widely used scheme for data transmission, because of its simplicity and high channel utilization. However, the fairness of the BEB algorithm is very poor in some cases. For example: considering a WLAN with n stations using the DCF mode to access channel, and the stations always have packets to transmit. When one station transmits successfully, it will decrease the size of the contention window to the size of the initial contention window (W0). Before it transmits a next packet, it has to uniformly choose a

backoff counter in the backoff interval, (0, W0-1). At that instance, other stations which

had experienced collided transmission have a larger backoff interval. As a result, a station with new packet in queue has higher probability to access channel than other waiting stations; that is, unfairness occurs.

Some algorithms to solve the fairness problem of MAC in WLAN were proposed [14], [15]. A multiplicative increase linear decrease (MILD) scheme was proposed in MACAW protocol for WLAN [14]. In the MILD scheme, the contention window of a collided station was increased by multiplying an amount of 1.5, while the contention window of a successful station is decreased by one step. Here, the step was defined as the transmission time of a packet. In the MACAW protocol, the current backoff interval information was included in each transmitted packet, and also a backoff interval copy

mechanism implemented in each station copied the backoff intervals of the overheard

it incurs a new problem. Consider the same example as above. When a station successfully transmits a packet with large contention window, other stations waiting to transmit packets change their backoff interval to the large contention window because of the backoff interval copy mechanism applied in MILD scheme. This algorithm works well when many stations happen to transmit at the same time because the probability of collision decreases. But, it results in long channel idle time and decreases channel utilization if only few stations contend for the wireless channel.

Haas and Deng proposed a new MAC scheme [45] named sensing backoff algorithm (SBA). The SBA is an optimized version of MILD algorithm in slotted ALOHA networks. In the SBA, upon collision, stations multiplied theirs contention windows by α (α>1). The backoff intervals of the transmitting station and the receiving station, after each successful transmission, were multiplied by θ (θ<1). The contention windows of all active stations sensing a successful transmission were decreased by β steps. Once the value of α was chosen, the optimization parameters for SBA can be accordingly determined. The SBA guarantees that the successful transmission probabilities of other stations are the same as that of the previously successful station; that is, the stations have the same transmission probability regardless of the number of retransmission. Unfortunately, SBA does not resolve the problem of large delay variance among stations.

Yamada, Morikawa, and Aoyama proposed a decentralized delay fluctuation control (DDFC) MAC mechanism [15], where the contention window is changed according the packet waiting time. The larger the packet waiting time is, the smaller the contention window will be. The DDFC in nature lessens variance of waiting time from enqueueing to successful transmission. Unfortunately, the channel utilization in DDFC is still low due to the small contention windows and high collision probabilities.

To support multimedia services for the IEEE 802.11e WLAN, dynamic contention window (CW) schemes [16]-[18], different maximum packet length scheme [18], and various interframe space (IFS) schemes [18]-[20] are usually adopted to design the priority differentiation. However, these solutions would still cause large delay variance in the same access category (AC) because of the backoff scheme. Noticeably, higher delay variance results in larger probability of quality-of-service (QoS) violation of multimedia traffic due to excess delay.

This chapter proposes and analyzes an adaptive p-persistent-based (APP) MAC scheme for the IEEE 802.11 WLAN proposed in [46], [47]. The APP MAC scheme, installed in a station, dynamically adjusts the permission probability of transmission for the station itself, and sets the permission probability as a function of the numbers of retransmissions and re-backoffs. The station with longer packet delay, implying larger numbers of retransmissions and re-backoffs, is given higher permission probability. Therefore, the packet delay variance of station for each access can be decreased and the WLAN can provide good delay fairness for stations in each access. The Markov-chain model [20], [49]-[51], is adopted to analyze the proposed APP MAC scheme. The performance measures such as collision probability, system throughput, and mean delay are successfully obtained. Numerical and simulation results show that the APP MAC scheme can effectively reduce the delay variance and thus achieve the delay fairness. The collision probability is decreased and the system throughput is enhanced, compared to conventional schemes. Moreover, discrepancy between numerical and simulation results is provided to corroborate the analyses. These results reveal that the analyses are quite accurate.

initial permission probabilities to various ACs to further differentiate their priorities. Moreover, it adaptively adjusts the permission probability of stations in each AC according to their respective waiting delays to reduce the delay variance of stations within the same AC.

The rest of the chapter 2 is organized as follows. Section 2.2 describes the system model, and section 2.3 introduces the APP MAC scheme. The mathematical analysis of the APP MAC scheme is given in section 2.4. Section 2.5 illustrates the performance comparisons of the APP MAC scheme and other conventional methods, such as BEB MAC and DDFC MAC, by numerical and simulation results. Finally, concluding remarks are given in section 2.6.

2.2 System Models

The IEEE 802.11 distributed coordination function (DCF) adopts the CSMA/CA protocol to support asynchronous data transfer. The station can start to transmit only if the medium is sensed idle for a time interval equal to DCF interframe space (DIFS). Otherwise, the transmission is deferred and the BEB algorithm is invoked. In the BEB algorithm, the station chooses a backoff counter from contention window (W), before transmitting. At the first transmission attempt, W is set to the initial contention window,

W0; otherwise, W depends on the number of transmissions failed for the packet. The

backoff counter is decremented by one at the end of each slot time, σ , as long as the medium is sensed idle, and suspended otherwise. It will be reactivated when the medium is again sensed idle for a period longer than DIFS. When the backoff counter reaches to zero, the station transmits immediately. A collision will occur when two or more stations transmit simultaneously. This kind of scheme is called 1-persistent.

response to its origination station to denote that the transmitted packet has been successfully received. If the acknowledge packet is not received, it assumes that the transmission has been corrupted. For an unsuccessful transmission, W is doubled until it reaches to the maximum value of the contention window, Wmax. For a successful

transmission, if the station still has packets queued for transmission, it enters a new backoff procedure.

In the APP MAC scheme, its backoff procedure is similar to that of the traditional CSMA/CA MAC scheme with BEB backoff algorithm, except when the backoff counter of a station in a backoff stage decreases to zero. At this instant, the station with the APP MAC scheme may transmit packet with a permission probability P or enter into a re-backoff procedure with a probability (1− P). Here, the re-backoff procedure is defined as the process of that the station will remain at the same backoff stage with the same contention window. Noticeably, if P is equal to one, the APP MAC scheme turns to the CSMA/CA MAC scheme with BEB algorithm.

2.3 The Adaptive P-Persistent (APP) MAC Scheme

The adaptive p-persistent (APP) MAC scheme is based on the CSMA/CA protocol with a novel APP transmission algorithm. In which, the value of the permission probability P is adaptively adjusted, according to the state of its packet transmission, which is a function of the number of retransmissions (backoff stages), denoted by RT, and the number of re-backoffs, denoted by RB. It is because RT and RB can be regarded as measures of delay time of packet transmission. If a station enters into the re-backoff procedure one time, the value of RB will be added one until up to RBmax, where RBmax is

increased anymore. If a station suffers a collision, the value of RT will be added one until up to BSmax and the value of RB will be set to zero, where BSmax is the maximum

number of backoff stage. When the value of RT is equal to BSmax and the station collides

again, the station will remain with the value of RT equal to BSmax. If a station achieves a

successful transmission, values of both RT and RB will be set to zero. Consequently, the APP MAC scheme can make a station obtain a higher permission probability P at the same backoff stage if the station has a larger RB; it will make a station obtain a lower permission probability P if the station is in the state with a smaller RT.

More in details, for a station with the APP algorithm, RT and RB are initially zero, and P is assigned to be P0 which is the initial permission probability chosen for the first

transmission of a ready packet. Afterwards, P will be adaptively adjusted according to the function designed by

max max 1 , 0 , 0 . 1 0 0 max max P P + BS RB BS + RB ⎡ ⎤ − + ∗_{⎢ ⎥} ≤ ≤ ≤ ≤ ⎣ ⎦ RB P = RT RT RB (2.1)

The philosophy behind Eq. (2.1) is that a station having larger RT and RB should be promoted to have a larger permission probability P. Also, it is expected that the average waiting time spent at any RB for a given RT would be less than that spent at (RT+1) and RB = 0. Therefore, it is reasonable that P is increased by (1-P0)/BSmax if one

more retransmission and (1-P0)/[BSmax*(1+RBmax)] if one more re-backoff procedure.The

pseudo-code for the APP algorithm in this APP MAC scheme is shown below. [The APP Algorithm]

if (RT = 0 and RB = 0) {

P = P0

Else {

P = P0 + (1-P0) / BSmax * [RT + RB / (RBmax +1)]

}

2.4 Analysis

For any station with the APP MAC scheme, define s(m), r(m), and b(m) to be random processes of the backoff stage, the number of re-backoff, and the value of backoff counter, at time m, respectively, where 0 ≦ s(m) ≦ BSmax, 0 ≦ r(m) ≦ RBmax,

and 0 ≦ b(m)≦ Wi− 1, Wi=2iW0, Wi is the contention window W of the ith backoff stage.

Also, define (s(m), r(m), b(m)) as the state of system. Assume that there are n contending stations in the system, and each station is operated in a saturation condition, denoting it always has a ready packet to transmit. The discrete-time observation points are embedded at the end of each slot time, which follows the medium if sensed idle longer than DIFS interval. The three-dimensional random process {(s(m), r(m), b(m))} is a discrete-time Markov chain under the assumptions that both the collision probability and the packet transmission probability of a station are indifferent to its backoff procedure [49]. The collision probability of a station, denoted by pc, is the probability of that a

station transmits and at least one of the other n− 1 stations transmits; the transmission probability of a station, denoted by p_τ, is the probability of that a station transmits at a randomly selected time slot. It is intuitive that this assumption would be more accurate as long as W0 and n get larger. Under this assumption, pc is supposed to be a constant

value. We can obtain the state transition diagram for a station shown in Fig. 2.1 and state transition probabilities given by

max max , 1 max max 1, max max , {( , , ) | ( , , 1)} 1, 0 , 0 , 0 2, 1 {( , , ) | ( , 1,0)} (1 ) , 0 ,1 , 0 1, 1 {( ,0, ) | ( 1, ,0)} , 1 , 0 , 0 1, {(0,0, ) | ( , ,0)} i i j i i i j c i i i P i j k i j k i BS j RB k W P i j k i j P i BS j RB k W W P i k i j P p i BS j RB k W W P k i j P − − + = ≤ ≤ ≤ ≤ ≤ ≤ − − = − ≤ ≤ ≤ ≤ ≤ ≤ − − = ≤ ≤ ≤ ≤ ≤ ≤ − = max max max max max 0 max max ,0 1 (1 ) , 0 , 0 , 0 1, 1 {( ,0, ) | ( ,0,0)} , 0 1, j c i BS c BS BS p i BS j RB k W W P BS k BS P p k W W ⎧ ⎪ ⎪ ⎪ ⎪ ⎪⎪ ⎨ ⎪ ⎪ − ≤ ≤ ≤ ≤ ≤ ≤ − ⎪ ⎪ ⎪ = ≤ ≤ − ⎪⎩ (2.2) (2.3) (2.4) (2.5) (2.6)

where P{(i, j, k)|(i’, j’, k’)}=Prob{(s(m) = i, r(m) = j, b(m) = k)|(s(m− 1) = i’, r(m − 1) =

j’, b(m− 1)=k’)}, and Pi,j is the permission probability P at state (i, j, 0). Eq. (2.2)

describes the fact that the backoff counter is decremented by 1 at the beginning of each slot time. Eq. (2.3) accounts for the situation that the station re-backoffs again. Eq. (2.4) indicates the case that an unsuccessful retransmission occurs at backoff stage i− 1 thus the backoff stage is increased and the new backoff counter is uniformly chosen in the range (0,Wi− 1). Eq. (2.5) denotes what a successful packet transmission happens, thus a

new packet starts with backoff stage 0 and the initial backoff counter is randomly chosen in the range (0,W0− 1). Finally, Eq. (2.6) stands for that RT is not increased in

subsequent packet transmissions, when the backoff stage reaches the value BSmax.

Define limm→∞(s(m), r(m), b(m)) as the system state at steady state. Let bi,j,k=limm→∞Prob{(s(m), r(m), b(m))=(i, j, k)} be the steady-state probability of the state

0, 0, 0 1, 0, 0 0, 1, 0 1, 1, 0 0,0 0 1 P W − ( ) 0,1 0 1 c P p W − 0,1 1 c P p W 0,0 0 c P p W 0,0 1 c P p W ( ) 1,0 0 1 c P p W − 1,0 1 1 P W − ( ) 1,1 0 1 c P p W − BSmax, 0, 0 BSmax, 1, 0 0, RBmax, 0 1, RBmax, 0 BSmax, RBmax, 0 ( ) max 0, 0 1 RB c P p W − ( ) max,0 0 1 BS c P p W − max max ,0 BS c BS P p W 0, RBmax, 1 1, RBmax, 1 BSmax, RBmax,1 0, RBmax, W0-1 1, RBmax, W1-1 BSmax, 1, 1 BSmax, 0, 1 ( ) max 1, 0 1 RB c P p W − max max max , BS RB c BS P p W max max ,1 BS c RT P p W max max ,0 1 BS BS P W − max max max , 1 BS RB BS P W − max max max , 1 BS RB BS P W − max 1, 1 1 P_RB W − max 0, 0 1 P_RB W − max 0, 0 1 PRB W − max 1, 1 1 PRB W − 1,0 1 1 P W − 1 1 1 1 1 max max, max, BS -1 BS RB W max max, 1, BS -1 BS W max max, 0, BS -1 BS W

max max max max 0 0,0, , , ,0 0,0, 1 0 0 0 0 0,0, 1 , , ,0 0 0 0 , 1 , , , , 1 , 1,0 max max , 1 , , 1 , 1,0 1 , 0 2, 1 , 1 , 0 1,1 , 0 2, 1 , i BS RB c k i j i j k i j BS RB c W i j i j i j i j i j k i j k i j i i i j i j W i j i p b P b b k W W p b P b W P b b b i BS j RB k W W P b b W + = = − = = − + − − − − − = + ≤ ≤ − − = − = + ≤ ≤ − ≤ ≤ ≤ ≤ − − =

∑ ∑

∑ ∑

max max max max max max

max max , , , , ,0 , 1 , 1,0 , , 1 0 1, 1 1, 1 (1 ) (1 ) , i RB k i RB i RB i RB i RB i RB k i i BS j RB b P b P b b W − − + ≤ ≤ − ≤ ≤ − ⎡ ⎤ = _⎣ − + − _⎦+ max

max max max max max

max max , , 1 , , ,0 , 1 , 1,0 max ,0, 1, 1, ,0 ,0, 1 max 0 ,0, 1 0 1, 0 2, 1 (1 ) (1 ) , 0 1, , 1 1, 0 2, i i BS i RB W i RB i RB i RB i RB i RB c i k i j i j i k i j i c i W i BS k W b P b P b i BS W p b P b b i BS k W W p b − − − − − + = − ≤ ≤ − ≤ ≤ − ⎡ ⎤ = _⎣ − + − _⎦ ≤ ≤ − = + ≤ ≤ − ≤ ≤ − =

∑

max max max max max max max

max max

max max max max max 1, 1, ,0 max 0 ,0, ,0, 1 1, 1, ,0 ,0 ,0,0 0 ,0, 1 1, 1, ,0 0 , 1 1, , 0 2, BS RB i j i j j i RB c BS k BS k BS j BS j BS BS BS j BS RB c BS W BS j BS j B j BS P b i BS W p b b P b P b k W W p b P b P W − − = + − − = − − − = ≤ ≤ − ⎡ ⎤ = + _⎢ + _⎥ ≤ ≤ − ⎣ ⎦ = +

∑

∑

∑

Via algebraic manipulation of Eq. (2.7), we can obtain

max , , , ,0 max max 1 , ,0 , ,0,0 max max 0 1 1 ,0,0 , , 0,0,0 max 0 0 1 , 0 1, 0 , 0 1, (1 ) , 0 1, 1 , (1 ) , 1 , i i j k i j i i j i j i r i r RB i r i c m r m s r m s W k b b i BS j RB k W W b P b i BS j RB b p P P b i BS − = − − = = =− ⎧ ₌ − _{≤ ≤ − ≤ ≤ ≤ ≤ −} ⎪ ⎪ ⎪⎪ _{= − ≤ ≤ − ≤ ≤} ⎨ ⎪ ⎪ _{⎛ ⎞} ⎪ = _⎜ − _⎟ ≤ ≤ ⎪ ⎝ ⎠ ⎩

∏

∑

∏

∏

where P_{i, 1−} is set to be zero. Also from Eq. (2.8), bi,j,k can be obtained in terms of b0,0,0,

permission probability Pi,j, and collision probability pc, by

max 1 1 1 0 , , , , , 0,0,0 0 1 1 1 0 2 (1 ) (1 ) , 2 RB i j i r i j k i i h c m r m s r h m s W k b P p P P b W − − − = =− =− =− ⎡ ⎤ − = − _⎢ − _⎥ ⎣

∑

∏

∏

∏

where P-1,j is defined to be 1. By using the normalization condition for stationary state

probabilities, the b0,0,0 can be yielded as

max max max

0,0,0 _{1 1 1 1} 0 , , , 0 0 0 0 1 1 0 1 1 . 2 (1 ) (1 ) 2 i BS RB W i j i RB r i h c m r m s i i j k h m r s b W k P p P P W − − − − = = = =− =− = =− = ⎡ ⎤ − _{− −} ⎢ ⎥ ⎣ ⎦

∑ ∑ ∑

∏

∏

∑

∏

Afterwards, the transmission probability of a station, p_τ, can be derived as

max max , , ,0 0 0 BS RB i j i j i j p_τ P b = = =

∑ ∑

max max 1 1 max 1

, , , , 0,0,0 0 0 1 1 0 1 (1 ) (1 ) , BS RB j i RB r i j i n c m r m s i j n m r s P P p P P b − − − = = =− =− = =− ⎧ ⎡ ⎤⎫ ⎪ ⎪ = _⎨ − _⎢ − _⎥⎬ ⎪ ⎣ ⎦⎪ ⎩ ⎭

∑ ∑

∏

∏

∑

∏

and the collision probability of station, pc, is given by

1

1 (1 ) .n c

p _{= − −} p_τ − _(2.12)

2.4.1 System Throughput

For the derivation of system throughput, we consider that the time span is partitioned into three categories: the idle slot time, denoted by T_σ , the successful transmission time, denoted by T , and the collision time, denoted by _s T . Proportionally, _c

the idle slot time would be with a portion of (1− Ptr), the successful transmission time

would be with a portion PtrPs, and the collision time would be with a portion of Ptr

(1− Ps). The Ptr is the probability of that at least one transmission occurs in a slot time,

and it is given by

1 (1 ) .n tr

P = − − p_τ (2.13)

The Ps is the probability of that a successful transmission occurs, conditioned on the fact

1 (1 ) . n s tr np p P P τ − τ − = (2.14)

Therefore, for a successful transmission of a packet in time Ts, the system throughput,

denoted by S, can be obtained by

(

)

where the denominator denotes the average time interval taken for this successful transmission, and B is the average payload size of a packet.

Values of Ts and Tc are given by, if the basic access mechanism is adopted,

, ,

s t

c t

T H B SIFS ACK DIFS

T H B DIFS δ δ δ = + + + + + + ⎧ ⎨ = + + + ⎩ (2.16)

where H is the time required to transmit PHY and MAC frame headers; B is the _t

average time that a payload is transmitted; SIFS is the duration of SIFS; δ is the propagation delay; ACK is the time required to transmit the acknowledgement packet; and DIFS is the duration of DIFS. They are given by, if the RTS/CTS access mechanism is used, , . s t c T RTS SIFS CTS SIFS

H B SIFS ACK DIFS

T RTS DIFS δ δ δ δ δ = + + + + + ⎧ ⎪ _{+ + + + + + +} ⎨ ⎪ = + + ⎩ (2.17)

Note that collision is assumed to be occurred at RTS frame transmitted.

2.4.2 Delay

As those described for Eq. (2.15), the average time interval taken for a successful transmission of a packet is (1− Ptr)T_σ+ PtrPsTs+ Ptr (1− Ps)T , and its probability is P_c trPs.

If the n contending stations are identical, the average delay of a station, denoted by TD,

(

)

2.4.3 The Optimal Value of P

In WLAN, the number of stations n is not a directly controllable variable. The way to achieve optimal performance is to employ adaptive techniques to tune the value of W0

based on an estimated value of n [49]. Bianchi stated in [49] that the maximum system throughput can be achieved if the optimal initial contention window in BEB, denoted by

Wopt, is given by

2

opt c

W ≈n T σ . (2.19)

In contrast, the initial contention window of the APP MAC scheme since is equivalent to

W0/P0, the optimal value of P0, denoted by P , can be obtained by ₀*

* 0 0 opt P =W W 0 2 c . W n T σ ≈ (2.20)

2.5 Numerical and Simulation Results

2.5.1 Data Only Environment

Table 2.1 lists system parameters of a considered WLAN environment and values of PHY-related parameters, which are referred to specifications of IEEE 802.11 [3]. In the simulations, we compare the APP scheme with the BEB and the DDFC [15] schemes. In the BEB scheme, two initial contention windows, W0=16 and W0=32, are assumed. In

beginning of packet contention, not as the primary usage defined in [15]. In the following figures, results of APP are shown by numerical and/or simulation, while results of BEB and DDFC are given by simulation.

Table 2.1 Parameter Settings for a WLAN Environment

Slot Time, σ 20 µs DIFS 60 µs SIFS 10 µs Propagation Delay 1 µs Bit Rate 11 Mbps PHY Overhead 192 µs

MAC Header 28 byte

ACK Length 14 byte

Data Packet Payload, B 1028 byte

Max Backoff Stage, BSmax 4

Initial Contention Window, W0 16

Transmission Retry Limit ∞

Figure 2.2 illustrates the collision probability pc of the APP, BEB, and DDFC MAC

schemes. It reveals that APP with P0 =1/4 achieves an improvement of collision

probability by 40% (38.8%) over DDFC (BEB with W0=16), when the number of

stations is 8. The reason is that the proposed APP MAC scheme assigns every packet a permission probability P. When two stations count to zero simultaneously, the collision probability of APP is equal to P2_{. Thus, APP has smallest collision probability; and the}

smaller the P0 is, the lower the collision probability would be. This phenomenon is

equivalent to making the initial contention window larger. The figure also exhibits that the discrepancy between numerical and simulation results is less than 3.5%, thus this corroborates the collision probability analysis.

0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5 2 3 4 5 6 7 8 9 10 11 12 13 14 15

The Number of Stations

Collision Probabil ity pc BEB, W0 =16 (simulation) BEB, W0 =32 (simulation) DDFC (simulation) APP, P0 =1/2 (numerical) APP, P0 =1/2 (simulation) APP, P0 =1/4 (numerical) APP, P0 =1/4 (simulation) APP, P0 =1/16 (numerical) APP, P0 =1/16 (simulation)

Figure 2.2 Collision probabilities of APP, BEB, and DDFC

Figure 2.3 depicts the system throughputs of the APP, BEB, and DDFC MAC schemes. It can be seen that the throughput increases first and then decreases. It is because increasing the number of stations not only raises the channel utilization but also enlarges the packet collision probability as shown in Fig. 2.2, so the throughput increases first and decreases due to high collision probability. Also, APP with P0 =1/4

achieves an improvement of throughput by 7% (6.5%) over DDFC (BEB with W0=16)

when the number of stations is 8. The reason is that APP can reduce the collision probability and increase the transmission efficiency consequently. It can also be found that the smaller P0 will cause a lower system throughput when fewer stations are in the

system. It is because the smaller P0 is equal to making a larger initial contention window.

justifies the validity of the throughout analysis. 5 5.2 5.4 5.6 5.8 6 6.2 6.4 6.6 6.8 7 2 3 4 5 6 7 8 9 10 11 12 13 14 15

The Number of Stations

System Throughput (Mbps) BEB, W0 =16 (simulation) BEB, W0 =32 (simulation) DDFC (simulation) APP, P0 =1/2 (numerical) APP, P0 =1/2 (simulation) APP, P0 =1/4 (numerical) APP, P0 =1/4 (simulation) APP, P0 =1/16 (numerical) APP, P0 =1/16 (simulation)

Figure 2.3 System throughputs of APP, BEB, and DDFC

Figure 2.4 shows the mean delays of the APP, BEB, and DDFC MAC schemes. It indicates that the APP with P0 =1/4 achieves an improvement of mean delay by 6.6%

(6.1%) over DDFC (BEB with W0=16), when the number of stations is 8. It is because

the APP enhances the channel utilization. It can also be found that the smaller P0 has a

larger delay time when there are fewer stations in the system but a smaller delay time when there are more stations in the system. Also, the difference between numerical and simulation results is less than 3.23%, and this substantiates the delay analysis.

Figure 2.5 shows delay variances of the APP, BEB, and DDFC MAC schemes versus the number of stations by simulations. It can be found that the APP MAC scheme possesses the lowest delay variation, while the BEB MAC scheme (BEB with W0=16)

the highest. For example, the APP with P0 =1/4 achieves improvement of delay variation

over DDFC (BEB with W0 =16) by 76.4% (79.4%), at the number of stations is 8. Also,

the smaller the P0 is, the more the improvement of delay variation would be. The reason

is the proposed APP scheme adaptively determines the permission probability of transmission according to a function of the number of retransmission (RT) and the number of re-backoff (RB). The APP scheme lets the ready packet with the longest delay time transmit first and delays the new packet, this makes the delay time of packet be close to the mean value.

0 0.005 0.01 0.015 0.02 0.025 2 3 4 5 6 7 8 9 10 11 12 13 14 15

The Number of Stations

Mean Delay (Sec) BEB, W0 =16 (simulation) BEB, W0 =32 (simulation) DDFC (simulation) APP, P0 =1/2 (numerical) APP, P0 =1/2 (simulation) APP, P0 =1/4 (numerical) APP, P0 =1/4 (simulation) APP, P0 =1/16 (numerical) APP, P0 =1/16 (simulation)

Figure 2.4 Mean delays of APP, BEB, and DDFC

Besides, making P0 smaller is equivalent to making the W0 larger, thus lower

Figs. 2.2 – 2.4). It is because the APP scheme is not actually increase the size of W0, but

provides another dimension (permission probability P) to avoid collision and makes the transmission efficiency, and thus the APP scheme has the smallest mean delay and highest system throughput.

0.0000 0.0005 0.0010 0.0015 0.0020 0.0025 0.0030 2 3 4 5 6 7 8 9 10 11 12 13 14 15

The Number of Stations

Delay V arianc e BEB, W0 =16 (simulation) BEB, W0 =32 (simulation) DDFC (simulation) APP, P0 =1/2 (simulation) APP, P0 =1/4 (simulation) APP, P0 =1/16 (simulation)

Figure 2.5 Delay variances of APP, BEB, and DDFC

Figure 2.6 shows the system throughput and delay variance of APP with optimal

* 0

P and BEB with Wopt given in [6] by simulations, where the BEB operates with Wopt to

obtain the maximum system throughput and the APP uses the optimal P0 with fixed W0.

It can be found that APP with optimal * 0

P loses the system throughput by 1.3% but

gains an improvement of delay variation by 15%, compared to BEB with Wopt. This

shows that the APP MAC scheme can achieve maximum system throughput and support good delay fairness.

Syste m Thr out hput (Mbps) De lay Var iance (ms 2 ) * 0 0 P =P * 0 0 P =P

Figure 2.6 performance of APP with optimal * 0

P and BEB with Wopt

2.5.2 Multimedia Service Environment

In the simulations, the multimedia WLAN considers three kinds of ACs: high, medium, and low priorities. High (low) priority AC is for voice (data) service, and medium priority AC is for multimedia message service (MMS). Packets generated from high priority AC stations are modeled in an on-off behavior; medium and low priority AC stations are assumed to be in the saturation mode. The packet payload size of high (medium, low) priority AC is 59 (528, 1028) bytes. The value of BSmax (RBmax) is 5 (5).

Also, parameters of the WLAN are set as follows: slot time = 20 µs, DIFS for high (medium, low) priority AC = 60 (80, 80)µs, SIFS=10 µs, propagation delay = 1 µs, bit rate = 11 Mbps, PHY overhead = 192 µs, MAC header = 28 bytes, and ACK length = 14 bytes. Values of PHY-related parameters are referred to specifications of IEEE 802.11e

number of high priority AC stations is altered to indicate various traffic load conditions. The BEB in [52] and the priority backoff algorithm (PBA) in [17] are selected for comparison. In PBA, each station computes the average quantity, in unit of bytes, of successful transmission data of the system. When a station has packet to transmit, it calculates CW based on the average system quantity and its priority. If the quantity of successful transmission data of the station itself is higher (smaller) than the average system quantity, the station should choose a larger (smaller) CW to let other station (itself) have higher possibility to access the channel, otherwise it uses the same CW to select backoff counter.

The P0 (CWmin) for high, medium, and low priority AC stations in the APP MAC

scheme is assumed to be 1/2 (8), 1/16 (24), and 1/32 (32), respectively. The CWmin of all

priorities in PBA is set to be 16. The BEB with CWmin equal to 8, 24, and 32 (16, 24, and

32) for high, medium, and low priority AC stations, respectively, is called BEB-I (BEB-II). Define the delay time of a voice packet as the time elapsed between the instant of the packet generation and the instant of the packet reception. A voice packet will be dropped if its delay time is larger than 40 ms. Also, the QoS requirement of voice service is defined as the voice packet dropping probability, which is set to be 3%.

Fig. 2.7 depicts (a) dropping probability, (b) mean delay, and (c) delay variance of voice packets in APP, BEB and PBA versus the number of high priority AC stations. It can be found that the voice packet dropping probabilities of the APP and BEB-I schemes are much smaller than those of the BEB-II and PBA schemes. Also, under the QoS requirement of voice service, APP can accommodate more than 20 voice stations, while BEB-I, BEB-II and PBA can have 18, 7 and 0 voice stations, respectively. The APP performs even better than the BEB-I. The reasons are that the APP further differentiates priorities of ACs by the initial assignment of P0, and gives voice service stations a

largest P0 to have a highest priority. Thus, the APP has the least mean delay, which is

shown in Fig. 2.7 (b). Moreover, the APP has both the capability of adaptive adjustment of permission probability and the effect of re-backoff procedure. Thus the APP achieves

Figure 2.7 (a) Dropping probability (b) mean delay and (c) delay variance of voice packets

the station’s transmission delay approaching to the mean value, and it has the smallest delay variance, which is given in Fig. 2.7 (c). On the other hand, the BEB-II cannot

than APP and BEB-I. Therefore, the increasing of the number of high priority stations would enlarge the collision probability of system. This causes BEB-II has higher mean delay, delay variance, and dropping probability of voice packets. The PBA changes

CWmin of high priority stations without considering the number of high priority stations

and the various payload size of different priority. In this simulation scenario, the payload size of voice (high priority) packet is much smaller than that of medium and low priority packets, thus the quantity of successful transmission data of high priority station is less than that of the average system. This leads the high priority stations to change their

CWmin to a small one and then results in a high collision probability. The phenomenon

would make PBA have the highest mean delay, delay variance, and dropping probability of voice packets.

Figure 2.8 shows the system throughput versus the number of high priority stations. It can be seen that APP performs the best and BEB-I performs the worst. When the number of high priority stations is 15, APP achieves an improvement of system throughput over BEB-I, BEB-II, and PBA by 24.1%, 9.9%, and 16.4%, respectively. The reasons are that the APP owns P0 to differentiate the priority, which can reduce collision

probability among stations of different priorities; the APP adaptively adjusts the permission probabilities, which can decrease collision probability among stations in the same AC. Consequently, APP enlarges the channel utilization and enhances the system throughput. Noticeably, when the number of high priority stations is larger than 18, the system throughputs of PBA and BEB-II are a little bit higher than that of APP. That is because APP devotes most of the channel bandwidth to sustain the voice QoS requirement, while PBA and BEB-II violate the voice QoS requirement, which is illustrated in Fig. 2.7 (a).

1.7 2.36 3.02 3.68 4.34 5 5 10 15 20

The Number of Voice Stations

S yste m Thr oughput (M bps) APP BEB-II PBA BEB-I

Figure 2.8 System throughput

Figure 2.9 presents the (a) mean delay and (b) delay variance of low priority packets versus the number of high priority stations. It can be found that the APP scheme has the smallest mean delay and delay variance of low priority packet. When the number of high priority stations is 15, the APP achieves by 21.6% (83.5%), 9.6% (78.3%), and 11.1% (16.9%) improvement of mean delay (delay variance) of low priority packet over the BEB-I, BEB-II, and PBA, respectively. Also, these two delay measures for medium priority packets with APP, BEB-I, BEB-II, and PBA have almost the same results as those of low priority packets, which are not shown here. The reason is that P0 in APP

provides another dimension to avoid collision and makes the transmission efficiency, and thus APP has the smallest mean delay for medium and low priority packets. Also, both the adaptive adjustment of permission probability and re-backoff procedure of APP for the medium and low priority stations work well, therefore their delay variance is the smallest. On the other hand, the BEB-I differentiates priority more greatly by setting a smaller CWmin for voice stations than the BEB-II. This makes voice stations of BEB-I

with BEB-I cannot access the channel more probabilistically and have mean 50 100 150 200 250 5 10 15 20

The Number of Voice Stations

Mean Delay (ms) BEB-I PBA BEB-II APP (a) 0 0.08 0.16 0.24 0.32 0.4 5 10 15 20

The Number of Voice Stations

Delay V arianc e BEB-I BEB-II PBA APP (b)

Figure 2.9 (a) Mean delay and (b) delay variance of low priority packet

delay and delay variance higher than those with BEB-II. In PBA, the payload sizes of medium and low priority packets are large, thus the quantity of successful transmission data of medium and low priority stations are larger than system average quantity. These medium and low priority stations would change CWmin up to maximal contention

window to reduce the collision probability of medium and low priority stations. Therefore, their delay and delay variance are smaller than those of BEB-I and BEB-II.

2.6 Concluding Remarks

This chapter proposed and analyzed an adaptive p-persistent (APP) MAC scheme for IEEE 802.11 WLAN to achieve fairness in the sense of low delay variance. The APP MAC scheme resolves the fairness problem at each access of stations by adaptively determining the permission probability of station according to the state of packet transmission of the station. It differentiates the permission probabilities of stations with various waiting delay, and assigns a higher priority (probability) to stations with larger packet delay. The chapter analyzes the APP MAC scheme by Markov-chain model and successfully obtains the collision probability, the system throughput, and the mean delay. Results show that the discrepancy between the numerical results and the simulation results is very small, and the analyses are quite correct. Besides, the APP MAC scheme can effectively reduce the delay variance and enhance the system throughput.

The initial permission probability P0 is an important design parameter in the APP

MAC scheme. It can be determined by considering the system design objective which is to reduce the delay variance or enhance the system throughput. Besides, the initial permission probability P0 can be adaptively determined according to the system load.

For example, P0 could be set to be 1/16 (1/2) when the system is in heavy (light) load.

For multimedia environment, the APP MAC scheme can differentiate stations with various AC of services in multimedia WLAN by setting different initial permission probabilities. Also, it dynamically determines the permission probability of a station in the same AC, according to its transmission state, to reduce the delay variance of the

performance of multimedia WLAN; it effectively improves the capacity of high priority stations, reduces the mean delay, enhances the mean throughput, and achieves lower delay variance, compared to conventional algorithms.

In realistic implementation, the number of rebackoffs (RB) and the number of retransmissions (RT) are statistical data recorded by stations. The current CW of a station can indicate RT, thus only a register is needed in the station to store the value of RB. Also, the value of P0 (RBmax) for an AC would be set larger (smaller) if the AC is with

Chapter 3

Dynamic Priority Resource Allocation

for Uplinks in IEEE 802.16 Wireless

Communication Systems

3.1 Introduction

Orthogonal frequency division multiplexing (OFDM) has been proposed as a promising technique for future multimedia wireless communication systems due to its ability to mitigate frequency selective fading, intersymbol interference (ISI) and its flexibility for adaptive modulation on each subcarrier. Orthogonal frequency division multiple access (OFDMA) has been adopted for IEEE 802.16 broadband wireless access (BWA) system. Although the medium access control (MAC) signaling has been well defined in the IEEE 802.16 specifications [42], resource management and scheduling are still remained as open issues. Since the wireless channel condition varies with time, adaptive resource allocation has been viewed as one of the key technologies to provide

service (QoS) requirements should be taken into account when developing an efficient resource allocation algorithm. Therefore, an effective resource allocation scheme is required to exploit frequency diversity, multiuser diversity, time diversity, and QoS requirement diversity so that the overall system resource can be efficiently utilized and QoS requirement can be guaranteed.

Subcarrier, bit, and power allocation algorithms for multiuser OFDMA systems to maximize the overall data rate or minimize the total transmitted power under some constraints have been studied in many literatures. Wong et al. [21] proposed a Lagrangian-based algorithm to minimize the total transmission power consumption under user’s QoS requirements, which were defined by a specified data transmission rate and bit error rate (BER). However, a high computational complexity renders it impractical. To reduce the complexity, Zhang and Letaief [25] proposed a near optimum dynamic multiuser subcarrier-and-bit allocation algorithm to maximize the overall spectral efficiency.

Many papers considered the downlink resource allocation [24], [25], [28], but a few papers investigated the uplink resource allocation. Resource allocation of both downlink and uplink is primarily performed by the base station (BS). Das and Mandyam [35] considered the uplink transmission of the OFDMA system and developed an efficient algorithm for subcarrier and bit allocation of each user. The algorithm includs the power distribution over the selected set of subcarriers for every user so that the total used power is minimized. Kim, Han, and Kim proposed a joint subcarrier and power allocation scheme for uplink OFDMA systems to maximize the rate-sum capacity based on Shannon capacity formula [36], where a greedy subcarrier allocation algorithm, based on a marginal rate function, and an iterative water-filling power allocation algorithm were proposed. The scheme was shown to achieve a near optimal solution. Jang and Lee