協作多點 (CoMP)是一種最近興起的傳輸技術，其主要作用為應付新一代無線通訊系統中的小區間干擾問題。在過去十數年內，研究員研發了 CoMP中一些關鍵的新技術，當中包括 MIMO合作和干擾協調。本論文考慮一個聯合用戶排程和干擾協調的問題。在傳統的研究中，用戶排程和干擾協調通常作為獨立的問題進行研究。可是，從本質上這兩個問題是相互影響的，因此傳統的研究將導致系統性能退化。為此，本論文探討了一個聯合用戶排程和波束形成（JACoB）的問題，這當中採用了一種稱為協同波束形成（CoBF）的干擾協調技術。具體而言，本文把 JACoB問題表達成了一個可支持用戶數最大化的問題，而其中的 CoBF設計將盡可能地配合用戶的需求而改變。 / 本論文有兩個主要的貢獻。第一，本文把 JACoB問題轉換成一個 ℓ₀範數最小化問題。其後本文採用 ℓ₁範數近似法將 JACoB問題近似為一個凸優化問題。第二，本文提出一種新型的分佈 JACoB方法。本文提出的分佈方法是基於塊坐標下降法。該方法不同於傳統的基於次梯度方法的分佈方法，如原始/對偶分解。 / 仿真結果顯示，採用本文提出的 JACoB方法（無論是集中的或是分佈的）所能支持的用戶數量遠超過現有的固定波束形成方法。此外，本文提出的分佈 JACoB方法能達到與集中JACoB方法相近的性能，而且其收斂速度亦是相當快的。 / Coordinated MultiPoint (CoMP) cooperative transmission has recently emerged as a promising technique for mitigating intercell interference in next generation wireless communication systems. Several key techniques for CoMP have been endeavored over the past decades, for example, MIMO cooperation and interference coordination. The present work studies a joint user scheduling and interference coordination problem in the CoMP downlink systems. Conventionally, user scheduling and interference coordination are treated as separate problems. This may result in a degradation of the system performance as the two problems are actually intertwined with each other. As such, this thesis considers a joint admission control and beamforming (JA-CoB) problem which employs a popular interference coordination technique called coordinated beamforming (CoBF). In particular, the JA-CoB problem is stated as a user number maximization problem where the CoBF design can be adapted to the set of selected users. / There are two major contributions in this thesis. Firstly, the JA-CoB problem is cast as an ℓ₀ norm minimization problem and then tackled by the now popularized ℓ₁ approximation technique. Secondly, a novel decentralized JACoB method is developed. The proposed de-centralized method is based on the simple block coordinate descent method, which is different from the conventional approaches which em-ploy subgradient-based method such as dual/primal decomposition. / The simulation results indicate that: i) the proposed centralized method yields a performance close to the optimum JACoB design while the complexity is significantly reduced; ii) employing the proposed JA-CoB methods (either centralized or decentralized) gives a significant gain over a fixed beamformers design in terms of the number of supported users. Moreover, the decentralized JACoB method achieves a performance close to its centralized counterpart, whilst the convergence speed is considerably fast. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Detailed summary in vernacular field only. / Wai, Hoi To. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2012. / Includes bibliographical references (leaves 77-80). / Abstracts also in Chinese. / Abstract --- p.i / Acknowledgement --- p.iv / Chapter 1 --- Introduction --- p.1 / Chapter 1.1 --- Overview of techniques for CoMP --- p.2 / Chapter 1.2 --- Overview of user scheduling algorithms --- p.4 / Chapter 1.3 --- Contributions --- p.6 / Chapter 2 --- The JACoB problem and the related works --- p.8 / Chapter 2.1 --- System model --- p.8 / Chapter 2.2 --- Joint admission control and beamforming (JACoB) --- p.10 / Chapter 2.2.1 --- Coordinated beamformers design --- p.11 / Chapter 2.2.2 --- Semide nite relaxation for the CoBF problem --- p.13 / Chapter 2.3 --- Related works --- p.14 / Chapter 2.3.1 --- Common trend in JACoB - deflation heuristic . --- p.18 / Chapter 2.4 --- Decentralized methods --- p.19 / Chapter 3 --- Centralized JACoB method --- p.21 / Chapter 3.1 --- Step 1 - a new formulation to JACoB --- p.21 / Chapter 3.2 --- Step 2 - ℓ₁ approximation to JACoB --- p.24 / Chapter 3.2.1 --- Properties of the ℓ₁ JACoB problem --- p.26 / Chapter 3.3 --- Proposed JACoB method --- p.28 / Chapter 3.3.1 --- Prescreening procedure --- p.28 / Chapter 4 --- Decentralized JACoB method --- p.31 / Chapter 4.1 --- Block coordinate descent method --- p.32 / Chapter 4.2 --- Smooth approximation to ℓ₁ JACoB --- p.34 / Chapter 4.2.1 --- Empirical iteration complexity of the BCD method --- p.38 / Chapter 4.3 --- Proposed decentralized JACoB method --- p.40 / Chapter 5 --- Simulation results --- p.43 / Chapter 5.1 --- Performance of centralized JACoB methods --- p.44 / Chapter 5.2 --- Performance of decentralized JACoB methods --- p.48 / Chapter 5.3 --- Summary --- p.52 / Chapter 6 --- Conclusions and future directions --- p.53 / Chapter 6.1 --- Future directions --- p.53 / Chapter 6.1.1 --- From a practical point of view --- p.54 / Chapter 6.1.2 --- From a theoretical point of view --- p.54 / Chapter A --- A primal decomposition method for (3.4) --- p.56 / Chapter B --- A projected gradient method for (4.3) --- p.60 / Chapter C --- Proofs --- p.67 / Chapter C.1 --- KKT conditions for (2.6) and (3.5) --- p.67 / Chapter C.2 --- Proof of Proposition 2.1 --- p.68 / Chapter C.3 --- Proof of Proposition 3.3 --- p.69 / Chapter C.4 --- Proof of Proposition 3.2 --- p.69 / Chapter C.5 --- Proof of Proposition 3.5 --- p.71 / Chapter C.6 --- Proof of Fact 4.1 --- p.75 / Bibliography --- p.77
| Identifer | oai:union.ndltd.org:cuhk.edu.hk/oai:cuhk-dr:cuhk_328500 |
| Date | January 2012 |
| Contributors | Wai, Hoi To., Chinese University of Hong Kong Graduate School. Division of Electronic Engineering. |
| Source Sets | The Chinese University of Hong Kong |
| Language | English, Chinese |
| Detected Language | English |
| Type | Text, bibliography |
| Format | electronic resource, electronic resource, remote, 1 online resource (xi, 80 leaves) : ill. (some col.) |
| Rights | Use of this resource is governed by the terms and conditions of the Creative Commons “Attribution-NonCommercial-NoDerivatives 4.0 International” License (http://creativecommons.org/licenses/by-nc-nd/4.0/) |
Page generated in 0.0123 seconds