從網路觀點對天線技術於多用戶排程及分時雙工/分碼多工無線系統之研究

(1)

國立交通大學

電信工程學系

博士論文

從網路觀點對天線技術於多用戶排程及

分時雙工/分碼多工無線系統之研究

A Network Perspective Investigation of MIMO

Antenna Techniques in Multiuser Scheduling and

TDD/CDMA Systems

研究生：陳瓊璋

指導教授：王蒞君博士

(2)

從網路觀點對天線技術於多用戶排程及

分時雙工/分碼多工無線系統之研究

A Network Perspective Investigation of MIMO

Antenna Techniques in Multiuser Scheduling and

TDD/CDMA Systems

研究生: 陳瓊璋 Student: Chiung-Jang Chen

指導教授: 王蒞君博士 Advisor: Dr. Li-Chun Wang

國立交通大學

電信工程學系

博士論文

A Dissertation

Submitted to Institute of Communication Engineering

College of Electrical Engineering and Computer Science

National Chiao Tung University

in Partial Fulfillment of the Requirements

for the Degree of Doctor of Philosophy

in

Communication Engineering

Hsinchu, Taiwan

(3)

從網路觀點對天線技術於多用戶排程及

分時雙工/分碼多工無線系統之研究

研究生: 陳瓊璋指導教授: 王蒞君博士

國立交通大學

電信工程學系

摘要

應用天線技術所帶來的高容量增益，雖然可以讓使用者在不增加頻

寬的環境下，提高無線通道的傳輸速率，但在實際應用該技術的同時，卻

也可能面臨兩項主要的議題－硬體實現複雜度，以及可能的鏈路可靠度降

低。針對上述兩項議題，傳統研究多採用實體層技術的解決方案，然而在

本論文中，則嘗試從網路觀點來提供另一種解決問題的方法－開發並利用

網路中已經存在的資源(例如用戶分集)，幾乎不需要額外增加成本，就可

以提供更多解決問題的維度。在本論文中，除了印證利用網路觀點所得到

的用戶分集，可以解決應用天線技術過程可能面臨的兩項議題外，更重要

的是，若應用這整個網路觀點的設計概念，於其他通信系統的設計上，將

有機會增加該系統的運作效能。

首先，我們考慮以分集為基礎(diversity-based)的天線技術在多用戶排

程系統的應用，在這裡我們推導出一個可以同時整合用戶維度、天線維

度、以及通道特性維度的公式，利用這個系統容量公式，可以說明多用戶

排程技術的特性，以及解釋排程技術與天線分集之間的交互作用。分析結

果顯示用戶分集是類似天線系統中的選擇分集，因此如果在一個多用戶通

信系統中已經存在了很大的用戶分集增益，這時若使用以分集為基礎的天

線技術，該技術所帶來的額外好處將會受限。另外，我們也說明了在多用

戶排程系統中，若使用時空塊狀編碼(space time block code)天線技術，甚

(4)

至可能導致系統容量損失。

其次，我們考慮以多工為基礎(multiplexing-based)的天線技術在多用

戶排程系統的應用。近來的研究顯示，使用以多工為基礎的天線技術，可

能會犧牲使用以分集為基礎的天線技術所帶來的好處－鏈路穩定度，因

此，我們提出了一個「最弱子通道優先」(strongest-weakest-normalized-

subchannel-first, SWNSF)排程演算法，淬取出多用戶系統中的用戶分集來

補償天線分集的不足，也就可以解決上述鏈路穩定度不足的缺點。接著，

我們嘗試在一個使用 SWNSF 排程演算法的多用戶排程系統中，使用一個

簡單的零強制(zero-forcing)接收器，當用戶數趨近無窮大時，我們證明零

強制接收器可以是一個最佳接收器，換句話說，應用以多工為基礎的天線

技術於多用戶排程系統時，可以用低成本的方式，同時增加系統容量與鏈

路穩定度。

最後，我們考慮波束合成天線技術在分時雙工/分碼多工系統的應用，

我們建議，使用波束合成天線技術，來解決基地站與基地站之間的強干擾

問題，利用鄰近不同基地台之間的合作，我們介紹一個同時上鏈路接收與

下鏈路發射的波束合成機制，來降低在分時雙工/分碼多工系統中出現的

基地站對基地站之間的強干擾，分析結果顯示，利用我們建議的波束合成

機制，可以在低成本的狀態下，有效消除基地站對基地站之間的強干擾問

題。

(5)

A Network Perspective Investigation of MIMO

Antenna Techniques in Multiuser Scheduling

and TDD/CDMA Systems

Student: Chiung-Jang Chen

Advisor: Dr. Li-Chun Wang

Department of Communication Engineering,

National Chiao Tung University, Taiwan

Abstract

The huge capacity gain offered by the multiple-input multiple-output (MIMO) antenna tech-nique can achieve higher data rates on wireless channels without sacrificing bandwidth ef-ficiency. The primary challenge to apply the MIMO technique lies in the implementation complexity and the possible side effect of reliability performance degradation. Traditional efforts to resolve these issues associated with the MIMO technique are mostly based on the physical layer treatment. Differently, this dissertation presents a network perspective ap-proach to revisit the MIMO technique with an aim to provide an alternative settlement. The main advantage of the network perspective approach is a full exploitation of existing dimensions in the network, e.g. the inherent multiuser diversity, with a negligible cost. More important than the results presented in this dissertation, however, is the hope that the net-work perspective methodology here will provide an innovative and flexible design paradigm for wireless systems.

First, the category of fading mitigation based (or diversity-based) MIMO antenna schemes is investigated for the multiuser scheduling system. A unified capacity formula connecting multiple domains of users, antennas and fading characteristics is derived. The analytical capacity formula is powerful in extracting the essence of multiuser scheduling and inter-preting the interplay of antenna diversity and multiuser scheduling. Our analysis indicates that the scheduling technique intrinsically delivers multiuser diversity with an analogy of selection diversity in the multiple antenna system. As a result, we show that the addi-tional capacity gain with the use of fading mitigation based antenna techniques is somewhat

(6)

limited because user population has contributed a large order of selection diversity in the multiuser scheduling system. It is also demonstrated that employing the space time block code (STBC) methods for the multiuser scheduling system may even cause a capacity loss due to the reduced amount of fading gain but without the supplement of array gain.

Second, the category of throughput enhancement based (or multiplexing-based) MIMO antenna schemes is studied for the multiuser scheduling system. Motivated by a recent result of diversity-multiplexing tradeoff in the MIMO system, we propose using the multiuser di-versity to replenish the didi-versity-deficient spatial multiplexing MIMO system. Particularly, we develop a novel strongest-weakest-normalized-subchannel-first (SWNSF) scheduling al-gorithm, requiring only scalar feedback, to enhance the degraded reliability performance of the MIMO system. Our analysis and results indicate that the SWNSF scheduling can sig-nificantly increase the coverage of the multiuser MIMO system while improving the system capacity. Furthermore, we consider a simple spatial multiplexing MIMO system with the zero-forcing receiver. Somewhat surprisingly, from a multiuser scheduling network perspec-tive, we prove that the zero-forcing receiver can be asymptotically optimal in the multiuser scheduling system as the number of users increases to infinity. In other words, the mar-riage of multiplexing-based MIMO antenna schemes and multiuser scheduling techniques can achieve an elegant cross-layer synergy of providing further capacity and reliability per-formance improvements in a cost-effective manner.

Finally, the type of interference suppression based antenna schemes (or antenna beam-forming) is investigated for the time division duplex/code division multiple access (TDD/ CDMA) system. We propose using antenna beamforming techniques to resolve the opposite direction interference problem in the TDD/CDMA system. Through exploiting the multiple antennas of adjacent base stations from the network point of view, we introduce a simulta-neous downlink transmit beamforming and uplink receive beamforming scheme to alleviate the impact of the opposite direction interference. Our analysis and results demonstrate that the proposed antenna beamforming scheme can effectively suppress the opposite direction interference with an economical implementation cost.

(7)

Acknowledgement

Foremost, I would like to express my sincere gratitude to my advisor, Dr. Li-Chun Wang, for providing me insights into important research problems, encouragement and support throughout my Ph.D. studies. This work could not have been done without his advice, guidance and comments.

Special thanks go to my mates at Wireless Network Lab. in NCTU and my colleagues at Chunghwa Telecommunication Labs. in Chung-Li for their kind help in many aspects.

Finally, I am deeply indebted to my family whose love and understanding have accom-panied me through this long journey.

(8)

List of Figures

1.1 Chapter organization of the dissertation. . . 8 2.1 The link variation induced by the SISO system and the SIMO system with

the fading mitigation based antenna technique under Rayleigh fading. . . 10 2.2 A generic fading mitigation based antenna scheme over the (Nt, Nr) MIMO

system, where wt= [ ˜w1,· · · , ˜wNt]

T _{and w}

r = [w1,· · · , wNr]

T _{are the transmit}

and receive antenna weights designed for delivering diversity. . . 11 2.3 A generic spatial multiplexing MIMO system, where the dotted lines represent

inter-subchannel interference. . . 12 2.4 A beamforming pattern that receives a signal from a speciﬁc location and

attenuate signals from other locations. . . 14 2.5 A downlink multiuser scheduling system with TDMA protocol. . . 16 2.6 Opposite direction interference in the TDD/CDMA system. . . 20 3.1 Impact of Nakagami-m channel fading on the capacity of the multiuser

schedul-ing system. . . 29 3.2 Comparison of the achievable system capacity with joint multiuser scheduling

and various diversity-based antenna schemes. . . 37 3.3 Impact of Nakagami-m channel fading on the attainable system capacity with

joint multiuser scheduling and antenna diversity schemes. . . 38 3.4 Comparison of the achievable scheduling gain with diﬀerent values of mean

(14)

4.1 Comparison of the cell radius of the multiuser MIMO system using the RR and SWNSF scheduling. . . 51 4.2 The CDFs of λk,i and ˜λk,i for K = 30 and N = 3. . . . 54 4.3 Capacity improvement with the SWNSF scheduling for the users in diﬀerent

user groups. . . 58 4.4 Average capacity gain resulting from the SWNSF scheduling with diﬀerent

numbers of users in the system. . . 59 5.1 Achievable throughput for the zero-forcing receiver in a single-user MIMO

system. . . 64 5.2 Outage probability for the zero-forcing receiver in a single-user MIMO system. 65 5.3 The CDFs of the output SNR subject to various scheduling algorithms in the

multiuser MIMO system with N = 3, K = 16 and ρ = 0 dB. . . . 78 5.4 Achievable throughput for the zero-forcing receiver in the multiuser MIMO

system with various scheduling algorithms. . . 80 5.5 Outage probability for the zero-forcing receiver in the multiuser MIMO system

with various scheduling algorithms. . . 81 5.6 The impact of the number of antennas on the scheduling gain deﬁned by

˜

Czf/Czf. . . 82

5.7 Comparison of the achievable throughput for the zero-forcing receiver and the optimal receiver operating in the same multiuser MIMO system. . . 83 6.1 An example to illustrate the interference scenario in the TDD/CDMA system,

where Bod = {2, 4, 6} represents the set of the neighboring cells generating the opposite direction interference and Bsd = {1, 3, 5} represents the cells generating the same direction interference. . . 86 6.2 A receiver block diagram with antenna beamformers. . . 90 6.3 An illustrative example for a TDD/CDMA system with Scheme I, whereBod=

(15)

6.4 An illustrative example for a TDD/CDMA system with Scheme II, where

Bod={2, 4, 6}. . . . 96

6.5 An illustrative example for a TDD/CDMA system with Scheme III, where the beam pattern in the center cell is for the uplink reception and those in the neighboring cells Bod={2, 4, 6} are for the downlink transmission. . . . . 99 6.6 Uplink performance comparison of Schemes I and II with diﬀerent numbers

of antenna elements (denoted as N in the ﬁgure), where the number of cells generating the opposite direction interference is equal to three. . . 102 6.7 Uplink performance comparison of Schemes I and II with diﬀerent numbers

of cells generating the opposite direction interference (denoted as B in the ﬁgure), where the number of antenna elements is equal to nine. . . 103 6.8 Performance improvements by implementing downlink transmit beamformer

in the surrounding base stations, where the number of cells generating the op-posite direction interference equal to six and the number of antenna elements equal to nine. . . 104 6.9 Performance comparison of four beamforming schemes with diﬀerent numbers

of cells generating the opposite direction interference, where an antenna array with nine elements is deployed at base stations. . . 105

(16)

List of Tables

3.1 Capacity summary of the multiuser scheduling system with various fading mitigation based antenna schemes, where Nmin = min(Nt, Nr). . . 35 4.1 Coverage and capacity enhancements with the SWNSF scheduling for K =

20, 50 and N = 1, 2, 3 . . . . 60 6.1 Simulation parameters for the TDD/CDMA system. . . 101

(17)

Notations

• |a| : magnitude of the scalar a • In : n × n identity matrix

• AH _{: complex conjugate transpose of the matrix A}

• A−1 _{: inverse of A}

• AT _{: transpose of A}

• [A]ij : (i, j)-th entry of A

• A† _{: pseudoinverse of A}

• kAkF : Frobenius norm of A

• tr(A) : trace of A

• det(A) : determinant of A • E[ · ] : expectation operation

• CN (0, 1) : complex circular symmetric Gaussian random variable • Er(x) : exponential integer function of order r, defined as Er(x) =

R∞

1 e−xtt−rdt

• Γ(x) : gamma function, defined as Γ(x) =R₀∞tx−1e−tdt

• Γ(a, x) : incomplete gamma function, defined as Γ(a, x) =R_x∞ta−1e−tdt

• ˜Γ(a, x) : normalized incomplete gamma function, defined as eΓ(a, x) = _Γ(a)1 R₀xta−1e−tdt

• Γa(x) : complex multivariate gamma function, defined as Γa(x) = πa(a−1)/2Qai=1Γ(x −

i+ 1)

(18)

• X1:L ≤ · · · ≤ XL:L: ordered statistics of a sample of size L from the random variable

X

• φX(ω): Laplace transform of the random variable X

• fX(x) : PDF of the random variable X

• fX1,··· ,XN(x1,· · · , xN) : joint PDF of the random variables X1· · · , XN

• FX(x) : CDF of the random variable X

• FX1,··· ,XN(x1,· · · , xN) : joint CDF of the random variables X1· · · , XN

• dxe : the smallest integer greater or equal to x • min(x1, x2) : minimum of x1 and x2

(19)

Abbreviations

• AF : amount of fading

• CDF : cumulative distribution function • CDMA : code division multiple access • CSI : channel side information

• DCA : dynamic channel assignment • DOA : direction of arrival

• EGC : equal gain combining • EGT : equal gain transmission • FDD : frequency division duplex

• IID : independent and identically distributed • LCMV : linearly constrained minimum variance • MAC : media access control

• MIMO : multiple input multiple output • MISO : multiple input single output • MMSE : minimum mean square error

• MVDR : minimum variance distortionless response • MRC : maximum ratio combining

• MRT : maximum ratio transmission • SC : selective combining

(20)

• ST : selective transmission • STBC : space time block code • SIMO : single input multiple output • SISO : single input single output • SNR : signal-to-noise ratio

• SWNSF : strongest-weakest-normalized-subchannel-first • PDF : probability density function

• RR : round-robin

• SIC : successive interference canceller • SVD : singular value decomposition • TDD : time division duplex

• TDMA : time division multiple access • UCA : uniform circular array

(21)

Symbols

• Nt : number of transmit antennas

• Nr : number of receive antennas

• K : number of users • Ck∗

link : link capacity between the base station and the selected target user k

∗

• hCi : system capacity • ρ : mean SNR

• pk∗ : probability of user k∗ receiving services from the base station

• m : Nakagami fading parameter

• hij : channel gain from the jth transmit antenna to the ith receive antenna

• σ2

n : thermal noise power

• γk : effective output SNR at user k

• Hk : normalized channel matrix between the base station and user k

• λmax : the maximum eigenvalue of HkHH_k

• f : amount of fading gain • a : array gain

• S : selection order

• xk : transmit signal vectors for user k

• yk : receive signal vectors for user k

(22)

• Gk : channel matrix between the base station and user k

• Pt : total transmit power of the base station

• µ : path loss exponent • gk : path loss for user k

• rk : distance away from the base station

• {λk,i}Ni=1 : ordered eigenvalues of HkHHk with λk,N ≥ λk,N −1 ≥ · · · ≥ λk,1≥ 0

• Sk,i : spacing random variable between λk,i and the smallest eigenvalue λk,1

• γth : a predetermined SNR value

• Pout : outage probability

• ˜H_k : channel matrix selected by SWNSF scheduling

• {˜λk,i}Ni=1 : ordered eigenvalues of ˜HkH˜Hk with ˜λk,N ≥ ˜λk,N −1 ≥ · · · ≥ ˜λk,1≥ 0

• {γk,i}Ni=1 : ordered effective SNR at the subchannel output of the spatial multiplexing

MIMO system with γk,N ≥ γk,N −1 ≥ · · · ≥ γk,1 ≥ 0 for user k

• {˜γ_k,i}N

i=1 : ordered effective SNR at the subchannel output of the spatial multiplexing

MIMO system with ˜γk,N ≥ ˜γk,N −1 ≥ · · · ≥ ˜γk,1 ≥ 0 for user k subject to SWNSF

scheduling

• N : number of antennas at both ends of the spatial multiplexing MIMO system • r : cell coverage subject to RR scheduling

• ˜r : cell coverage subject to SWNSF scheduling

• Ck : link capacity for user k subject to RR scheduling

(23)

• β : Euler’s constant ' 0.5772

• Czf : achievable link throughput for the zero-forcing receiver

• Copt : achievable link throughput for the optimal receiver

• ˜Cma zf , ˜C mi zf , ˜C si zf and ˜C sg

zf : achievable link throughput for the zero-forcing receiver

operat-ing in the multiuser MIMO system with the max-max, max-min, spatially-independent and spatially-greedy scheduling policies, respectively

• ˜Pma out, ˜P mi out, ˜P si out and ˜P sg

out : outage probability for the zero-forcing receiver operating in

the multiuser MIMO system with the max-max, max-min, spatially-independent and spatially-greedy scheduling policies, respectively

• {˜γma

n:N}Nn=1, {˜γn:Nmi }n=1N , {˜γn:Nsi }Nn=1 and {˜γ

sg

n:N}Nn=1 : ordered effective SNR at the

sub-channel output of the zero-forcing receiver operating in the multiuser MIMO system with the max-max, max-min, spatially-independent and spatially-greedy scheduling policies, respectively

• Bsd and Bod : the set of the neighboring cells generating the same direction interference

and the opposite direction interference, respectively • Kj : number of active users in cell j ∈ Bsd

• Iod, Isd and Iic : opposite direction interference, same direction interference and

intra-cell interference, respectively

• Pr : the power-controlled level at the base station

• L : processing gain

• r0 and rkj : the distance from mobile kj (j ∈ Bsd) to cell 0 and to cell j, respectively

• ak : array manifold vector for the signal arriving from the target user k

(24)

• ˜ak : array manifold vector for the signal transmitting to the target user k

• ˜bj : array manifold vector for the signal transmitting from cell j ∈ Bod

• xk, xod, xsdand xic: received signal vectors with contributions from the desired user k,

opposite direction interference, same direction interference and intracell interference, respectively

• Φx : covariance matrix of the received signal vector

• Φod, Φsd and Φic : covariance matrix of the received opposite direction interference,

same direction interference and intracell interference, respectively

• Φk : normalized covariance matrix of the received interference plus noise

• ˜Φod: covariance matrix of the received opposite direction interference when the

beam-forming Scheme IV is used

• ˜Φk: normalized covariance matrix of the received interference plus noise when the

beamforming Scheme IV is used

• wbs and wmv : combining weight for the beam-steering and MVDR beamformers,

respectively

• ˜wbs : beamforming weight for the transmit beam-steering beamformer

• ˜K_j : number of active users in cell j ∈ Bod

• B : number of cells generating the opposite direction interference • N : number of antenna elements at the base station for beamforming • ck(·) : spreading code of user k

• T : bit duration

(25)

• τk : propagation delay of user k

• γbs, γmv, ˜γbs and ˜γmv : the resulting bit energy-to-interference density ratio subject to

(26)

Chapter 1 Introduction

As a famous quote in the movie Jurassic Park says, “Life will find a way,” so will an innovation of thinking find a way, too. An ancient wisdom has also taught us how to escape from a deadlock of our lives: step back, take another broader view and we will find a whole different world. This philosophy turns out to be useful for designing practical wireless systems to meet the challenge of achieving higher bandwidth efficiency for the ever-increasing data rate requirement in the future. In response to this philosophy, this dissertation is going to present a story about a beautiful encounter of the multiple-input multiple-out (MIMO) antenna technique with its underlying wireless systems that may provide a different look to view the MIMO technique.

One of key techniques that promise significant capacity improvements for a wireless system is the MIMO antenna technique. This magic basically derives from an exciting the-oretical prediction that the capacity of a MIMO system can be scalable with the number of antennas employed at both ends of the transmitter and receiver [1]. However, the ad-vance of MIMO techniques does not only bring enthusiasm but also skepticism. The main concerns lie in receiver complexity [2] and the accompanying side effect of reliability perfor-mance degradation for the MIMO system [3]. Traditionally, many efforts made in resolving these concerns associated with the MIMO technique are mostly based on the physical layer treatment. This dissertation, however, adopts a network perspective approach to revisit the MIMO technique with an aim to provide an alternative settlement. Specifically, we will

(27)

consider the multiuser scheduling and time division duplex/code division multiple access (TDD/CDMA) systems with MIMO techniques.

Multiuser scheduling systems provide an opportunity for the multiple antennas among different users to cooperate with each other. A consequence of such a cooperation, from a network point of view, is an expansion of available dimensions since the number of exploitable antennas is virtually increased to the sum of the multiple antennas from all users. The additional dimension can be utilized to improve the system performance as well as reduce the implementation complexity of the MIMO receiver. Interestingly, what makes this happen is just a shift of a point of view: from the point-to-point perspective to the point-to-multipoint network perspective. The use of scheduling techniques for a wireless system, suggested in great part by the latest development of information theory, can take advantage of the delay-tolerant data characteristics to deliver a multiuser diversity gain. This system-wide benefit resulting from the scheduling strategy can be illustrated by an analogy of the water-filling principle: pouring the resource to the user with best channel quality. Wireless standards IS-856 [4] and 3GPP R5 [5] (the upgraded systems of cdma2000 and WCDMA) have changed the downlink design into a multiuser scheduling system for supporting efficient and high speed packet data access.

Another example that is used to demonstrate the strength of the network perspec-tive approach is the cooperation of multiple antennas among different adjacent cells in the TDD/CDMA system. The story, however, is somewhat different from the multiuser schedul-ing system. In TDD/CDMA systems, the downlink capacity can be greatly enhanced to support asymmetric traffic by simply allocating different numbers of time slots for uplink and downlink transmissions. Nevertheless, this merit can invoke a severe opposite direction interference problem among adjacent cells of the TDD/CDMA system [6]. The issue of oppo-site direction interference, if not properly tackled, may hinder the TDD/CDMA system from being deployed in a large-scale network with multiple cells. From the network perspective to invite a cooperation between multiple antennas from adjacent cells, this dissertation will

(28)

show that a search for eﬀective and economical antenna techniques to resolve the opposite direction interference issue thus becomes possible.

In summary, this dissertation is an attempt to exploit the possible interaction and coop-eration between the MIMO technique and its underlying wireless system to simultaneously enhance the system performance and reduce implementation cost. The key idea behind the network perspective methodology is to encourage a full exploitation of all the available di-mensions within the whole wireless network. While the dissertation focuses on the multiuser scheduling and TDD/CDMA systems, it is our hope that the network perspective design methodology can be also applied to the other wireless systems in the future.

1.1 Problem and Solution

The main question addressed here is: how to effectively and efficiently use the multiple antenna technique on top of the multiuser scheduling and TDD/CDMA systems? The research of multiple antenna techniques has been a fertile area in the history. Nevertheless, most of the studies are based on a point-to-point view without contemplating the interplay between the multiple antenna technique with its underlying communication system. In this dissertation, from a network point of view, we explore the interaction between the multiple antenna technique and the communication system to develop effective strategies for utilizing the MIMO technique on top of the multiuser scheduling and TDD/CDMA systems. Towards this end, the following methodology is adopted:

1. Examine the degrees of freedom that the various multiple antenna techniques can provide.

2. Distill the essences of the multiuser scheduling and TDD/CDMA communication sys-tems.

3. Develop analytical frameworks to evaluate the marriage of diﬀerent types of multiple antenna techniques and the considered communication system.

(29)

4. Suggest strategies to leverage the synergy of combining the multiple antenna techniques with the multiuser scheduling and TDD/CDMA systems.

As will be seen in the rest of the dissertation, there exist interesting and profound connec-tions between the multiple antenna technique and its underlying communication system. Examples and results for illustrating the strength of the network perspective approach will be demonstrated later in the dissertation.

The following addresses the issues of combining the multiple antenna technique with the multiuser scheduling and TDD/CDMA systems.

1.1.1 Fading Mitigation Based Antenna Techniques for Multiuser

Scheduling Systems

In point-to-multipoint communication systems, a special type of diversity, called the mul-tiuser diversity, can be exploited to improve spectral efficiency [7, 8]. This kind of diversity can be explained as an analogy of the water-filling principle across multiple users: higher system spectral efficiency can be attained by pouring more resources to the user with better channel quality. A proper scheduling algorithm is the key to extract the multiuser diversity inherent in the multiuser system [9, 10, 11]. Some current industrial standards, such as the IS-856 [4] and the 3GPP R5 [5], have adopted scheduling techniques to enhance spectral efficiency for delay-insensitive data services.

Generally speaking, scheduling is a media access control (MAC) layer technique to de-liver multiuser diversity gain by taking advantage of independent channel variations among user population. By contrast, antenna diversity is a physical layer approach to offer reli-able transmissions with the major goal of mitigating channel fading. In a multiuser system where a channel-aware scheduling algorithm arranges transmissions based on the link qual-ity of multiple users, various antenna diversqual-ity schemes may provide different effective link statistics and ultimately may lead to distinct capacity results with the effect of scheduling. Thus, the cross-layer interaction between antenna diversity and multiuser scheduling is not

(30)

straightforward and requires careful investigations. An analytical framework will be estab-lished to reveal insights into the interlay of antenna diversity and multiuser scheduling in Chapter 3.

1.1.2 Throughput Enhancement Based Antenna Techniques for

Multiuser Scheduling Systems

Using the spatial multiplexing technique to transmit signals over a MIMO channel has attracted great attentions because of its capability to deliver remarkable capacity gain [1, 12, 13]. However, recent studies have revealed that the large capacity benefit result-ing from the spatial multiplexresult-ing MIMO system may come at the price of link reliability degradation when the prior channel information is not available at the transmitter [3, 14]. In [3], the authors derived an optimal tradeoff curve for the maximum achievable diversity gain of an open-loop MIMO system given any realized multiplexing gain. In [14], the au-thors quantitatively evaluated some practical diversity-based and multiplexing-based MIMO schemes. Their numerical results indicated that it is difficult to simultaneously accomplish both diversity gain and multiplexing gain in an open-loop MIMO system. Due to the tradeoff of antenna multiplexing gain against antenna diversity gain, applying the spatial multiplex-ing scheme to transmit data over the MIMO channel may lead to smaller coverage areas subject to the same total transmit power and link reliability requirement. How to pursue high throughput with the spatial multiplexing MIMO scheme while maintaining satisfactory link reliability remains an open research issue. In Chapter 4, we will propose a scheduling scheme, called the strongest-weakest-normalized-subchannel-first (SWNSF) scheduling, to replenish the diversity-deficient MIMO multiplexing system with multiuser diversity.

1.1.3 Throughput Enhancement Based Antenna Techniques for

Multiuser Scheduling Systems with Zero-Forcing Receivers

One popular approach to achieve the promised remarkable capacity gain of the MIMO sys-tem is the spatial-multiplexing method by sending parallel data streams across Nt multiple transmit antennas [2]. In order to decode the spatially multiplexed signals over the MIMO

(31)

system, various receiver architectures such as the zero-forcing receiver, the MMSE receiver and the successive interference canceller (SIC) have been introduced in the literature [13, 15]. Generally speaking, the performance enhancement from one type of MIMO receivers to an-other usually comes at the price of higher implementation cost. For example, the zero-forcing receiver is known to suﬀer from the eﬀect of noise enhancement despite its simplicity. The SIC-based receiver, on the other hand, can possibly achieve the full capacity gain of the MIMO system with much higher complexity [16]. In contrast with those pure physical layer approaches, in Chapter 5 we will leverage the cross-layer cooperation between the simple zero-forcing receiver and scheduling technique to realize the full theoretical capacity of the MIMO channel.

1.1.4 Interference Suppression Based Antenna Techniques for

TDD/CDMA Systems

One of the key advantages for the TDD system is the capability to deliver asymmetric traffic services by allocating different numbers of uplink and downlink time slots. However, in a TDD/CDMA system, asymmetric traffic may result in severe opposite direction interference because downlink transmitted signals from neighboring base stations may interfere with the uplink received signals of the home cell. In the literature, there are two research direc-tions to avoid the opposite direction interference. The first research direction is from the perspective of channel assignment techniques, such as [17, 18]. Another research direction to alleviate the impact of the opposite direction interference in TDD/CDMA systems is to apply advanced antenna techniques [19, 20, 21]. Compared with other categories of smart antenna technology, beamforming is known for its capability of suppressing strong interfer-ence [22, 23]. In addition, beamforming can easily exploit the reciprocity of TDD channels to leverage the benefit of joint downlink and uplink beamforming. In Chapter 6, we will investigate the effect of beamforming technique to resolve the opposite direction interference in TDD/CDMA systems.

(32)

1.2 Dissertation Outline

This dissertation deals with the topics of combining the multiple antenna technique with the multiuser scheduling and TDD/CDMA systems. To be explicit, the multiple antenna techniques are categorized according to their design purposes into three families:

• Fading mitigation based antenna techniques.

• Throughput enhancement based antenna techniques. • Interference suppression based antenna techniques.

Chapter 2 provides an overview of multiple antenna techniques that fall into the three cate-gories and a brief introduction of the multiuser scheduling and TDD/CDMA communication systems. Fig. 1.1. illustrates the organization of the remaining chapters of the dissertation. Chapter 3 investigates the interplay between the fading mitigation based antenna tech-nique and the multiuser scheduling system. A uniﬁed capacity formula that connects the ingredients of fading characteristics, multiuser scheduling gain and antenna diversity gain is derived here. Through the uniﬁed capacity analysis, the interaction between multiuser scheduling techniques and various diversity-based antenna schemes will be unveiled.

Chapter 4 discusses the marriage of the throughput enhancement based antenna tech-nique and the multiuser scheduling system. From the diversity-multiplexing compensation point of view, it is shown that the scheduling technique can signiﬁcantly improve the de-graded reliability performance of the multiplexing-based MIMO system while further en-hancing the throughput performance. The outcome of cell coverage extension and system capacity improvement is shown in this chapter to substantiate this beneﬁt.

Chapter 5 revisits the topic of combining the throughput enhancement based antenna technique with the multiuser scheduling system. In contrast with Chapter 4 where the optimal receiver is assumed, this chapter considers the zero-forcing receiver. By showing that the eﬃciency of the zero-forcing can approach that of the optimal receiver in the multiuser

(33)

Fading mitigation based antenna techniques

Throughput enhancement based antenna techniques

Interference suppression based antenna techniques

Multiuser scheduling systems TDD/CDMA systems ] Chapter 3 Chapters 4 and 5 Chapter 6

Figure 1.1: Chapter organization of the dissertation.

MIMO system with many users, this chapter endorses the elegant marriage of the throughput enhancement based antenna technique with the multiuser scheduling system.

Chapter 6 shifts the focus to the TDD/CDMA system. The interference suppression based antenna technique is investigated to resolve the opposite direction interference for the TDD/CDMA system. By exploiting the cooperation among multiple antennas of adjacent base stations, a low-cost and feasible scheme is introduced in this chapter.

Finally, Chapter 7 provides some concluding remarks and suggests topics for future re-search.

(34)

Chapter 2 Background and Literature Survey

2.1 Overview of Multiple Antenna Techniques

Consider a point-to-point link with Nt transmit antennas and Nr receive antennas or an (Nt, Nr) multiple-input multiple-output (MIMO) system. When only multiple antennas are employed at one end, multiple-input single-output (MISO) and single-input multiple-output (SIMO) are used to denote the (Nt, 1) and (1, Nr) MIMO systems, respectively. Likewise, for Nt = Nr = 1 the MIMO reduces to the single-input single-output (SISO) system. In the literature, dozens of antenna techniques have been developed to exploit the additional spatial (antenna) domain. Based on the design purpose, the multiple antenna technique can be classiﬁed into three categories: fading mitigation based antenna techniques, throughput enhancement based antenna techniques and interference suppression based antenna tech-niques.

2.1.1 Fading Mitigation Based Antenna Techniques

This family of fading mitigation based antenna techniques are basically designed for compen-sating again channel fading so as to provide stable channel variations. The central principle behind this category of antenna schemes is to produce independent replicas of the desired signal over the MIMO channel so that the receiver can utilize the multiple faded copies to restore the original signal with higher reliability. In order to guarantee high degrees of in-dependence, the fading mitigation based antenna technique generally requires the multiple

(35)

0 10 20 30 40 50 60 70 80 90 −30 −25 −20 −15 −10 −5 0 5 10 15 Time

Singal power strength (dB)

SISO N

r = 6

N

r = 3

Figure 2.1: The link variation induced by the SISO system and the SIMO system with the fading mitigation based antenna technique under Rayleigh fading.

antennas be widely spaced.

What mostly characterizes the fading mitigation based antenna technique is its capability of delivering antenna diversity gain to mitigate channel fading. For simplicity, we shall also call the fading mitigation based antenna technique as the diversity-based antenna technique in this dissertation. Fig. 2.1 illustrates the induced link variation by using the fading mitigation based antenna technique under Rayleigh fading. Compared with the SISO system, one can see that the link ﬂuctuation is more damped for the SIMO system with the diversity-based antenna scheme.

An operational deﬁnition for the antenna diversity gain can be made as follows. Assuming that an uncoded data signal is sent through the MIMO system and Pe is the resulting error probability at the receiver, the antenna diversity gain D can be measured by [3]

D =− lim

ρ→∞

log (Pe(ρ))

log ρ , (2.1)

(36)

˜ w1 6 × ˜ wNt 6 × -x × w1 6 × wNr 6 y - R

Figure 2.2: A generic fading mitigation based antenna scheme over the (Nt, Nr) MIMO system, where wt = [ ˜w1,· · · , ˜wNt]

T _{and w}

r = [w1,· · · , wNr]

T _{are the transmit and receive} antenna weights designed for delivering diversity.

antenna diversity gain corresponds to the slope of the error probability versus SNR curve on a log-log scale at the high SNR regime. When each antenna pair between the transmitter and receiver encounters independent fading, the maximum diversity gain that can be extracted from an (Nt, Nr) MIMO system is Dmax = NtNr.

Figure 2.2 illustrates a generic fading mitigation based antenna scheme for the MIMO system. With the properly designed antenna weights at the transmitter and/or receiver, some representative antenna schemes capable of delivering full antenna diversity gain are introduced as follows.

• For the SIMO system, receive methods such as selective combining (SC), equal gain

combining (EGC) and maximum ratio combining (MRC) are commonly used to provide diversity gain [24, 25, 26].

• For the MISO system, transmit methods like selective transmission (ST), equal gain

transmission (EGT) and maximum ratio transmission (MRT) can be also utilized to yield diversity gain [13, 27]. Compared with the receive methods, the implementation of transmit methods generally requires the prior channel knowledge at the transmitter. To overcome this weakness, the space time block code (STBC) method was introduced by [28, 29]. By putting diﬀerent levels of correlation across the multiple transmit

(37)

x1 xNt Spatial m ultiplexing -· -· -· x3x2x1 y1 yNr Receiv er -· -· -· y3y2y1

-Figure 2.3: A generic spatial multiplexing MIMO system, where the dotted lines represent inter-subchannel interference.

antennas as well as several time slots, the STBC method can achieve full antenna diversity gain without the requirement of prior channel side information (CSI).

• For the MIMO system, both transmit and receive diversity can be obtained by using the

hybrid schemes such as ST/MRC [30], EGT/EGC [31] and MRT/MRC [32, 33]. The STBC method can be also applied to the MIMO system with proper generalizations [34, 35].

2.1.2 Throughput Enhancement Based Antenna Techniques

A different line of thinking to design antenna schemes focuses on exploiting the spatial dimensions to enhance the transmission throughput. The development of this category of antenna schemes is mostly inspired by the pioneering works [1, 12, 36] where a theoretically remarkable capacity benefit from the MIMO system was predicted under the Rayleigh fading channel. Since that, a lot of efforts have been made to realize the promised capacity benefit by devising practical antenna techniques.

One popular approach to fulﬁll the promised capacity gain of the MIMO system is the spatial multiplexing method [2]. As shown in Fig. 2.3, the spatial multiplexing MIMO sys-tem sends parallel data streams across Nt transmit antennas. In case the channel exhibits

(38)

rich scattering and the receiver has Nr multiple antennas satisfying Nr ≥ Nt, the spatially multiplexed signals can be successfully restored at the receiver, thereby providing an capac-ity gain or multiplexing gain approximately Nt times the SISO system [37]. However, the successful recovery of the spatially multiplexed signals calls for eﬀective inter-subcahnnel interference cancellation mechanisms.

Various receiver architectures such as the zero-forcing (ZF) receiver, the minimum-mean-square-error (MMSE) receiver and the successive interference canceller (SIC) have been introduced to recover the spatially multiplexed signals in the literature [13, 15]. Generally speaking, the performance improvement from one type of the spatial multiplexing MIMO receiver to another usually comes at the price of higher implementation cost. For example, the zero-forcing receiver is known to suﬀer from the eﬀect of noise enhancement despite its simplicity. By comparison, the SIC-based receiver can possibly achieve the full capacity gain of the MIMO system with much higher complexity [16]. Other receiver designs that try to reduce complexity while maintaining high performance can be also found, for example, in [38].

The multiplexing gain of an (Nt, Nr) MIMO system can be deﬁned as [3]

R = lim

ρ→∞

log (C(ρ))

log ρ , (2.2)

where C(ρ) is the ergodic capacity of the MIMO system operating at the SNR condition of ρ. Accordingly, it is followed from [1] that the maximum multiplexing gain that can be achieved from an (Nt, Nr) MIMO system is Rmax = min(Nt, Nr). Since this category of antenna techniques feature the desirable linear growth of multiplexing gain, we will also call the throughput enhancement based antenna technique as the multiplexing-based an-tenna technique in the dissertation. Using the definition of (2.1) and (2.2), [3] theoretically demonstrated a fundamental tradeoff between the achievable diversity gain and multiplexing gain in an open-loop MIMO system. Similarly, [14] quantitatively evaluated several existing MIMO antenna schemes and concluded that it is difficult to simultaneously accomplish both diversity gain and multiplexing gain. Based on this understanding, some recent works such

(39)

desired signal

interference interference

Figure 2.4: A beamforming pattern that receives a signal from a speciﬁc location and atten-uate signals from other locations.

as [39] began to seek new coding schemes that provides ﬂexible adjustment between the maximum achievable diversity gain and the maximum achievable multiplexing gain over the MIMO system.

2.1.3 Interference Suppression Based Antenna Techniques

Another aspect of using multiple antennas is to lay stress on its capability of rejecting strong interference. By exploiting the angular resolution provided by the antenna arrays, the interference suppression based technique can effectively cancel interference radiating from certain directions [40, 41, 22]. In order to ensure good resolvability in the angular domain, this category of antenna schemes generally requires the antenna arrays to be densely spaced [42]. In addition, antenna topology is relevant for the antenna arrays to create different spatial signatures that can result in different angular selectivity.

While other antenna techniques such as [43] can have the good capability of cancelling interference, we restrict to the antenna beamforming technique (also known as adaptive antenna array technique) as the interference suppression based technique is referred in this dissertation. The term “beamforming” derives from the fact that early spatial ﬁlters were

(40)

designed to form pencil beams to receive a signal from a speciﬁc location and attenuate signals from other locations [40]. As shown in Fig. 2.4, the beamforming pattern is commonly used to describe the angular resolvability of an antenna array. Some antenna beamforming schemes designed for reception or transmission are introduced as follows.

• For the SIMO system, receive beamforming techniques such as the steering

beam-former, minimum variance distortionless response (MVDR) beamformer and linearly constrained minimum variance (LCMV) beamformer can be used to perform spatial ﬁltering [23, 44]. Generally speaking, in the (1, Nr) SIMO system the degrees of free-dom that can be used to cancel the undesired interference signals is (Nr− L), where

L is the number of constrains imposed on the optimization criterion in deriving the

receive beamforming weight [40].

• For the MISO system, transmit beamforming schemes such as the transmit

beam-steering beamformer can be used to concentrate the transmission energy towards the desired direction [42, 45]. Transmit beamforming differs from receive beamforming in that the desired signals from multiple users are only coupled by the transmission power, but also by the shape of the radiation pattern [46]. This poses a more diffi-cult challenge for optimal transmit beamforming designs since the transmit power and beamforming weight for all users have to be jointly considered. Various optimal beam-forming algorithms based on different optimization criteria can be found in [46, 47].

2.2 Introduction of Multiuser Scheduling Systems

Industrial standards such as IS-856 [4] and 3GPP R5 [5] are multiuser scheduling systems supporting high speed downlink packet data access. In this dissertation, we adopt IS-856 as a reference model to investigate the characteristics of the multiuser scheduling system.

(41)

user 1

User 2

scheduled data in time slot

user K

User 5 User 3 User 8 User 1

: downlink signal

: uplink feedback

Figure 2.5: A downlink multiuser scheduling system with TDMA protocol.

2.2.1 Scheduling Technique and Multiuser Diversity

Consider a multiuser scheduling system with the base station serving K downlink users as shown in Fig. 2.5. The basic operations of the multiuser scheduling system supporting high speed downlink packet data access are described as follows [48].

• The time-division-multiple-access (TDMA) protocol is adopted for the base station to

service one user at a time slot.

• Each user measures and tracks its channel condition via the downlink common pilot

signal and reports back to the base station through the uplink feedback channel.

• With the channel information from the feedback of all users, the base station determines

to service one target user according to certain scheduling policies.

• Once the target user is selected, the base station uses the rate control adaptive

modu-lation to transmit as many information bits as possible to the target user with its full transmit power.

By taking advantage of delay-tolerant data characteristics, the scheduling technique can extract the multiuser diversity from such the multiuser system to improve spectral eﬃciency [7]. The multiuser diversity gain can be explained as an analogy of the water-ﬁlling principle

(42)

across multiple users: pouring more resources to the user with better channel quality [8]. While allocating more resources to the user with better channel quality, the system may immediately face another key issue – how to schedule the transmissions for other users whose channel qualities are poor. Therefore, one of the major challenges in designing wireless scheduling algorithms is to achieve higher total system throughput but without sacriﬁcing fairness to individual users too much.

A literature survey regarding the wireless scheduling technique is discussed as follows. [49, 50] modified the fair scheduling policies used in traditional wireline systems to the wireless. However, their scheduling algorithms assume a simple two-state on-off channel model, which may not be able to fully capture all the wireless channel characteristics. In [9, 10, 51], several wireless scheduling algorithms were evaluated by Monte Carlo simulations for practical radio channels. In [52] and [53], it was shown that one-by-one time division scheduling scheme is better than the code division scheme from the standpoints of higher energy efficiency and better received signal quality, respectively.

2.2.2 Scheduling for Multiuser MIMO Systems

The multiuser scheduling system in Fig. 2.5 can be extended to the multiuser MIMO system. A multiuser MIMO system consists of the base station employed with Nt transmit antennas and K multiple users employed with Nr receive antennas each. Accordingly, each link between the base station and individual user constitutes an (Nt, Nr) MIMO system.

Generally speaking, scheduling is a media access control (MAC) layer technique to deliver multiuser diversity by exploiting independent channel variations among user population. By contrast, the multiple antenna technique is a physical layer approach to improve the performance of wireless links. A probe into the cross-layer interaction between the multiple antenna technique and multiuser scheduling has recently attracted attentions in the research community. In particular, a literature survey associated with the topic of scheduling for multiuser MIMO systems is discussed as follows.

(43)

• A number of works that used the fading mitigation based antenna techniques for the

multiuser scheduling system can be found, for example, in [11, 51] and [54]-[58]. In [11], Viswanath and Tse proposed an opportunistic transmission scheme to increase the capacity of the multiuser scheduling system requiring only limited feedback. In [51], the authors conducted computer simulations to evaluate the capacity of the multiuser MIMO system with the ST and STBC antenna schemes. In [54], a semi-analytical result of spectral eﬃciency with the MRC antenna scheme and K-order multiuser diversity was derived. In [55], the impact of multiuser scheduling on the STBC methods was discussed in terms of the receive SNR distribution. Also, information theoretical treatments for the multiuser MIMO system with antenna diversity were provided in [56, 57, 58].

• Relatively, fewer works have taken advantage of scheduling techniques to improve the

performance of the multiuser MIMO system with multiplexing-based antenna schemes, such as [59, 60]. Through simulation, the authors in [59] showed that multiuser di-versity can provide additional capacity gain for the multiplexing-based MIMO system with linear receivers. In [60], the authors harnessed multiuser diversity to enhance the capacity of the multiplexing-based MIMO systems using a random beamforming technique.

• When the base station is allowed to simultaneously transmit multiple beams to

dif-ferent users (that is, not restricted to the TDMA protocol), both the temporal and spatial (antenna) domains can be exploited by scheduling to provide higher selection diversity order [61]-[64]. Because the signals to be transmitted for the multiple users are interfered with each other, the spatial-temporal domain scheduling is usually com-bined with additional pre-transmit signal processing techniques such as dirty-paper coding to achieve better performance [65]-[67].

In Chapter 3, we will introduce a uniﬁed analytical framework to investigate the capacity of the multiuser MIMO system with various diversity-based antenna schemes. In Chapters

(44)

4 and 5, we will discuss the topic of combining the multiplexing-based antenna scheme with the multiuser scheduling system.

2.3 Introduction of TDD/CDMA Systems

Time division duplex (TDD) is a duplex scheme where both uplink and downlink traffic takes place in the same unpaired frequency band. By allocating different numbers of slots for uplink and downlink, the TDD system can support asymmetric traffic services with great flexibility [68, 69]. Another merit of the TDD system over the frequency division duplex (FDD) system is the channel reciprocity that can be exploited to implement, for example, efficient open-loop power control [68, 70], pre-equalization [71, 72] and pre-rake diversity technique [73, 74]. With these advantages, modern cellular standards such as [75] have incorporated the TDD mode into their system designs.

2.3.1 Opposite Direction Interference

Despite the advantage of flexibly supporting traffic asymmetry in the TDD system, the TDD/CDMA system may pose a severe opposite direction interference problem across cells due to the universal frequency reuse of CDMA [6]. Figure 2.6 illustrates the typical in-terference scenario in the TDD/CDMA system. Assume that cells A and B in the figure have different rates of traffic asymmetry and allocate time slots independently according to their own traffic requirements. During a particular time slot to, one can find that the uplink received signals at cell A may suffer strong interference from the downlink transmitted sig-nals of the neighboring cell B. In this dissertation, we call this kind of base stations to base stations interference the opposite direction interference because the desired signal is in the

uplink direction, while the interference is from the downlink direction.

On the other hand, in time slot ts of Fig. 2.6, the uplink transmissions from the users in cell B will interfere with the uplink signals of cell A. We call this kind of mobile stations to base stations interference the same direction interference. The same direction interference also occurs in FDD/CDMA systems. Many previous works, such as [76, 77], have analyzed

(45)

Cell A Cell B

(1) : Opposite direction interference from cell B to cell A (2) : Same direction interference from cell B to cell A

(1) (2)

downlink uplink

uplink

U : uplink time slot D : downlink time slot

D D D U U U D D D D U U (1) (2) Cell A Cell B o

t

s

Figure 2.6: Opposite direction interference in the TDD/CDMA system.

the impact of the same direction interference. Thanks to power control mechanisms and other techniques, the impact of the same direction interference can be effectively managed in FDD/CDMA systems. However, the opposite direction interference, which is unique in TDD/CDMA systems, is substantially different from the same direction interference. First, it is difficult to coordinate many base stations throughout the entire service area to perform downlink power control simultaneously. Moreover, since the transmitter power of a base station is much higher than that of a mobile station, the opposite direction interference introduced by the neighboring base stations will severely degrade the quality of uplink signals transmitted from a mobile station [6, 78].

Usually, to avoid the opposite direction interference in TDD/CDMA systems, one can use different frequency carriers among adjacent cells. Obviously, this approach sacrifices frequency reuse efficiency. To use the same frequency carriers in every TDD/CDMA cell, one possible solution to avoid the opposite direction interference is to restrict all the neighboring

(46)

cells to adopting the same slot allocation pattern [79], i.e., all the assignments for either uplink or downlink transmissions in every time slot are the same. However, this approach implies that all cells will be forced to adopt the same rate of traﬃc asymmetry in the entire system, which is obviously not a very practical restriction. The key to relax this restriction is to overcome the opposite direction interference in the TDD/CDMA system. In the literature, there are two research directions to avoid the opposite direction interference:

• The ﬁrst research direction is from the perspective of channel assignment techniques,

such as [17, 18]. In [17], Haas and McLaughlin proposed a dynamic channel assign-ment algorithm to reduce the occurrence of the opposite direction interference due to asymmetric traﬃc. However, the authors in [18] concluded that it may be diﬃcult to achieve the optimal time slot allocation in an environment with multiple TDD/CDMA cells.

• Another research direction to alleviate the impact of the opposite direction interference

in TDD/CDMA systems is to apply the multiple antenna technique [19, 20, 21]. The authors in [19] and those in [20] proposed to adopt sector antennas combined with time slot allocation methods to suppress the opposite direction interference for the TDD/CDMA system and for the TDD/TDMA system, respectively. Furthermore, it was shown in [21] that the diversity-based SC and MRC antenna schemes are not feasible to resolve the opposite direction interference issue for the TDD/CDMA system. In Chapter 6, we will investigate the eﬀect of using the antenna beamforming technique to resolve the opposite direction interference in the TDD/CDMA system.

(47)

Chapter 3 Fading Mitigation Based Antenna

Techniques for Multiuser Scheduling

Systems

In this chapter, we develop an analytical framework to study the interaction between the fading mitigation based (or diversity-based) antenna technique and the multiuser scheduling system. We consider a multiuser scheduling system as described in Section 2.2. Under the generalized Nakagami fading channel model, we will derive a uniﬁed capacity formula applicable for the multiuser scheduling system with a number of fading mitigation based antenna schemes, including

• Selective transmission/selective combining (ST/SC), standing for that the selective

transmission and selective combining schemes are utilized at the transmitter and the receiver, respectively.

• Maximum ratio transmission/maximum ratio combining (MRT/MRC). • Selective transmission/maximum ratio combining (ST/MRC).

• Space-time block codes (STBC).

With a further change of parameters, the derived capacity formula can be versatile to inter-pret the interplay of antenna diversity and multiuser scheduling within the multiuser MIMO network.

(48)

3.1 Channel Model

To investigate the impact of channel fading on the multiuser scheduling system, this chapter considers the generalized Nakagami fading channel model. Consider a multiuser scheduling system with a base station serving K downlink users as shown in Fig. 2.5. To begin with, we assume that the base station and each user have only one single antenna. Let hk be the channel gain between the base station and user k, and let σ_n2 be the thermal noise power. Accordingly, γk = |hk|2/σn2 denotes the received instantaneous SNR of user k. We assume that each link between the base station and any user is subject to independent Nakagami fading with a common Nakagami fading parameter m. Then, the probability density function (PDF) of the received SNR for user k is [24]

fγk(γ) = m ρk m γm−1 Γ(m)exp −mγ ρk , γ > 0 (3.1)

where ρk is the average received SNR, and Γ(·) is the gamma function deﬁned by

Γ(m) = _∞

0

tm−1e−tdt . (3.2)

When m = 1, the Nakagami fading channel is identical to the Rayleigh fading channel. For

m > 1, a line-of-sight or a specular component exists. As m → ∞, the Nakagami channel

approaches to the AWGN channel.

To ease notation, we denote X ∼ G(p, q) as a gamma distributed random variable with parameters p and q. Then, the PDF of X is given as [80]

fX(x) =

qp Γ(p)x

p−1_e−qx_{, x > 0 .} _(3.3)

Furthermore, the cumulative distribution function (CDF) of X can be expressed by

FX(x) = Γ(p, qx) , (3.4)

where Γ(·, ·) is the normalized incomplete gamma function deﬁned by [82] Γ(a, x) = 1

Γ(a) x

0

從網路觀點對天線技術於多用戶排程及分時雙工/分碼多工無線系統之研究

國 立 交 通 大 學

電信工程學系

博 士 論 文

從網路觀點對天線技術於多用戶排程及

分時雙工/分碼多工無線系統之研究

A Network Perspective Investigation of MIMO

Antenna Techniques in Multiuser Scheduling and

TDD/CDMA Systems

研 究 生： 陳 瓊 璋

指導教授： 王 蒞 君 博士

從網路觀點對天線技術於多用戶排程及

分時雙工/分碼多工無線系統之研究

A Network Perspective Investigation of MIMO

Antenna Techniques in Multiuser Scheduling and

TDD/CDMA Systems

研 究 生: 陳瓊璋 Student: Chiung-Jang Chen

指導教授: 王蒞君 博士 Advisor: Dr. Li-Chun Wang

國 立 交 通 大 學

電 信 工 程 學 系

博 士 論 文

A Dissertation

Submitted to Institute of Communication Engineering

College of Electrical Engineering and Computer Science

National Chiao Tung University

in Partial Fulfillment of the Requirements

for the Degree of Doctor of Philosophy

in

Communication Engineering

Hsinchu, Taiwan

從網路觀點對天線技術於多用戶排程及

分時雙工/分碼多工無線系統之研究

研 究 生: 陳 瓊 璋 指導教授: 王 蒞 君 博士

國 立 交 通 大 學

電 信 工 程 學 系

摘 要

應用天線技術所帶來的高容量增益，雖然可以讓使用者在不增加頻

寬的環境下，提高無線通道的傳輸速率，但在實際應用該技術的同時，卻

也可能面臨兩項主要的議題－硬體實現複雜度，以及可能的鏈路可靠度降

低。針對上述兩項議題，傳統研究多採用實體層技術的解決方案，然而在

本論文中，則嘗試從網路觀點來提供另一種解決問題的方法－開發並利用

網路中已經存在的資源(例如用戶分集)，幾乎不需要額外增加成本，就可

以提供更多解決問題的維度。在本論文中，除了印證利用網路觀點所得到

的用戶分集，可以解決應用天線技術過程可能面臨的兩項議題外，更重要

的是，若應用這整個網路觀點的設計概念，於其他通信系統的設計上，將

有機會增加該系統的運作效能。

首先，我們考慮以分集為基礎(diversity-based)的天線技術在多用戶排

程系統的應用，在這裡我們推導出一個可以同時整合用戶維度、天線維

度、以及通道特性維度的公式，利用這個系統容量公式，可以說明多用戶

排程技術的特性，以及解釋排程技術與天線分集之間的交互作用。分析結

果顯示用戶分集是類似天線系統中的選擇分集，因此如果在一個多用戶通

信系統中已經存在了很大的用戶分集增益，這時若使用以分集為基礎的天

線技術，該技術所帶來的額外好處將會受限。另外，我們也說明了在多用

戶排程系統中，若使用時空塊狀編碼(space time block code)天線技術，甚

至可能導致系統容量損失。

其次，我們考慮以多工為基礎(multiplexing-based)的天線技術在多用

戶排程系統的應用。近來的研究顯示，使用以多工為基礎的天線技術，可

能會犧牲使用以分集為基礎的天線技術所帶來的好處－鏈路穩定度，因

此，我們提出了一個「最弱子通道優先」(strongest-weakest-normalized-

subchannel-first, SWNSF)排程演算法，淬取出多用戶系統中的用戶分集來

補償天線分集的不足，也就可以解決上述鏈路穩定度不足的缺點。接著，

我們嘗試在一個使用 SWNSF 排程演算法的多用戶排程系統中，使用一個

簡單的零強制(zero-forcing)接收器，當用戶數趨近無窮大時，我們證明零

強制接收器可以是一個最佳接收器，換句話說，應用以多工為基礎的天線

技術於多用戶排程系統時，可以用低成本的方式，同時增加系統容量與鏈

路穩定度。

最後，我們考慮波束合成天線技術在分時雙工/分碼多工系統的應用，

我們建議，使用波束合成天線技術，來解決基地站與基地站之間的強干擾

問題，利用鄰近不同基地台之間的合作，我們介紹一個同時上鏈路接收與

下鏈路發射的波束合成機制，來降低在分時雙工/分碼多工系統中出現的

基地站對基地站之間的強干擾，分析結果顯示，利用我們建議的波束合成

機制，可以在低成本的狀態下，有效消除基地站對基地站之間的強干擾問

題。

A Network Perspective Investigation of MIMO

Antenna Techniques in Multiuser Scheduling

and TDD/CDMA Systems

Student: Chiung-Jang Chen

Advisor: Dr. Li-Chun Wang

Department of Communication Engineering,

National Chiao Tung University, Taiwan

國立交通大學

博士論文

研究生：陳瓊璋

指導教授：王蒞君博士

研究生: 陳瓊璋 Student: Chiung-Jang Chen

指導教授: 王蒞君博士 Advisor: Dr. Li-Chun Wang

國立交通大學

電信工程學系

博士論文

研究生: 陳瓊璋指導教授: 王蒞君博士

國立交通大學

電信工程學系

摘要