A Decentralized Resource Allocation for Small Cells in Heterogeneous Networks

LE ANH TUAN 경희대학교 대학원 2017 국내박사

Small cell networks are envisioned as a promising solution to enhance the capacity and coverage for indoor and cell edge users with less capital expenditure and maintenance cost in the heterogeneous networks (HetNets). To reap the benefits of small cell network deployment, we target optimization-based approaches to employ the resource allocation problems for small cells with various enabling technologies in the future wireless network. Accordingly, the overall small cell network is modeled by an optimization problem. This problem integrates multiple constraints such as transmit powers, backhaul link capacities, radio resources, interference management, and the users' quality of service (QoS) demands into a unified optimization problem. To solve this problem, we exploit the frameworks of the coalitional game, matching theory, and optimization theory to find sub-optimal solutions in a decentralized manner. In this thesis, we particularly investigate two optimization problems for resource allocation in small cells with the cognitive radio and wireless virtualization technologies. These technologies are identified as some of the key emerging technologies for future wireless networks. At first, we consider a resource allocation in the uplink of the cognitive femtocell network (CFN). The goal is to maximize the uplink sum-rate under constraints of intra-tier and inter-tier interference while maintaining the average delay requirement for femtocell users and protecting the macrocell base station. In the solution section, we aim at developing distributed algorithms in which the CFN implementation is self-organized and self-optimized. To this end, we first propose an autonomous framework, in which the femtocell users self-organize into disjoint groups (DJGs). Then, we examine the coalitional game aspects of the subchannel and power allocations in each DJG. We show that the optimization problem can be formulated as a coalitional game in partition form. This game captures realistic inter-coalition effects that are formed by players who are seeking to cooperate and to form coalitions. By using the recursive core method and optimization theory, we develop a distributed algorithm for the power and channel allocations. We prove that the proposed algorithm always converges to a Nash-stable partition. Then, we study resource allocation for small cell networks with the wireless virtualization considering both the backhaul capacity of the infrastructure provider (InP) and the users' QoS requirements. The optimization problem focuses on the profit gained by a mobile virtual network operator (MVNO) which is a middleman who buys physical resource from the InP, bundling them into virtual resources called slides before selling off the service providers. To solve this problem, we propose a distributed solution framework based on Lagrangian relaxation to find a suboptimal decision about slice and transmit power allocations. Furthermore, by exploiting the concept of a matching game, we propose a low complexity solution that makes our proposal much more practical and robust in the virtualized wireless network environment. In all of these scenarios, the proposed frameworks are evaluated based on the simulation results. We show that the proposed frameworks can be implemented in a distributed manner and require a small number of iterations to converge. The experimental results show that the proposed frameworks are better than those of the other frameworks.

Development of compressed block and green mortar using loess (Hwangto)

Le, Anh Tuan Yeungnam Univ. 2009 국내박사

Synthetic Study on Alkaloids via New Cyclization Strategy

Le Anh Tuan 충남대학교 대학원 2010 국내박사

The family of alkaloids is one of the most important families in natural products due to their biological activities. Therefore, many efforts have been devoted for the total synthesis of important naturally-occurring alkaloids such as taxol, vinblastine or even unnatural alkaloids e.g. oseltavimir phosphate. For the synthesis of quinolizidine alkaloids, homopumilotoxin 223G (1) and epiquinamide (2), the common intermediate 1-hydroxyquinolizidinone has been an attractive target for many research groups recently. To prepare this important intermediate, we have developed an asymmetric and divergent route. Sharpless asymmetric dihydroxylation was manipulated to prepare the necessary precursor, azido-mesylate lactone, and the subsequent sequential cyclization produced stereoselectively (-)-1-hydroxyquinolizidinone or its epimer. Tetrahydroisoquinoline family which exhibit antitumor activities has been studied intensively over 30 years. The two core structures of this family are the quinone and the aromatic core. The tetrahydroisoquinolines include potent cytotoxic agents displaying a range of antitumor activities, antimicrobial activity, and other biological properties. The saframycin, naphthyridinomycin/bioxalomycin, and quinovarcin/tetrazomine families belong to this natural product group. To develop a new stereoselective way to the core structure, we used L-dopa as starting material and constructed the core ring skeleton by Bischler-Napieralski reaction followed by stereoselective reduction to the syn-piperidine moiety. And the -iifollowing arrangement of the functional groups would provide the desired intermediate. Over the years, the Sceletium alkaloids have still gained much attention from synthetic chemists due to the interesting cis-3a-arylhydroindole skeleton. The problematic challenge needed to overcome is the stereoselective construction of the sterical quarternary carbon center (C3a). During the syntheses of the members of this alkaloid family, a number of methods have been published, however, a more efficient, stereoselective strategy still needs to develop. Our strategy involved the stereoselective cyclization of the key intermediate 1,3-diketone amide. This compound prepared by palladium-catalyzed coupling reaction followed by Claisen rearrangement and amide synthesis, provided a (1:3) diastereoisomeric mixture of tetrahydroindole-2,4-diones under thermodynamically controlled-condition. Deoxygenation of the ketone group and reduction with Et3SiH in CF3COOH furnished the cis-3a-aryloctahydroindol-2- ones, the precursors for the synthesis of mesembrane isomers.

Coordinated Congestion Control Algorithms for Multipath Transport Protocol

Le Tuan Anh 경희대학교 2012 국내박사

improved throughput, and load-balancing. Multimedia streaming is becoming more popular on the Internet and various versions of TCP dominates Internet trac. MPTCP halves the data rate upon detecting packet drop, thereby its transmission rate is highly variant, so MPTCP is not well suited for many multimedia applications. We extend the uid model of binomial algorithms for single-path transmission to support the concurrent transmission of packets across mul- tiple paths. We focus on two particular algorithms multipath square-root (MPSQRT) and inverse-increase/additive decrease (MPIIAD). MPSQRT increases its window size inversely proportionally and decreases it proportionally to square-root of the current con- gestion window. MPIIAD increases its window size inversely proportionally to the current congestion window and decreases additively. Hence, their data rate are smoother than MPTCP while preserving fairness to both TCP and MPTCP. Another challenge for MPTCP is ecient in high BDP networks, which consist of high bandwidth and/or long delay links. However, MPTCP was designed to be backward- compatible with regular TCP (TCP Reno), which has proven to have low utilization of bandwidth available in high BDP networks. We introduce an extended version of Cubic TCP for multiple paths, called MPCubic. The extension process is approached from an analysis model of Cubic by using coordinated congestion control between paths. MPCubic can spread its trac across paths in load-balancing manner, while preserving fair sharing with regular TCP, Cubic, and MPTCP at joint bottlenecks. Moreover, to improve resilience to link failure, we propose a multipath fast recovery algorithm, which can signicantly reduce the recovery time of data rate after restoration of failed links. Finally, MPTCP also faces signicant degradation of performance in wireless networks as packet losses are occurred by random error in wireless links rather than by network congestion as in wired networks. We again apply the multipath extension technique into conventional single-path TCP Westwood for multipath transmission, called MPTCPW. To start with the analysis model of TCPW, MPTCPW is designed as a coordinated congestion controller between paths which allows load-balancing, fair sharing to regular TCPW at bottleneck. Our simulation results show that MPTCPW can achieve higher throughput than MPTCP in wireless environments, fairness to regular TCPW.

Nonlinear Controls of the Overhead Cranes

Le Anh Tuan 경희대학교 2012 국내박사

This study proposes six improved nonlinear controllers for the overhead cranes by using two symbolic techniques of nonlinear controls composed of Sliding Mode Control (SMC) and Partial Feedback Linearization (PFL). Besides, the adaptive control approach called Model-Reference Adaptive Control (MRAC) is applied to design two other kinds of nonlinear controllers. The key contributions of this thesis are proposing the nonlinear controllers based on the combination of PFL and SMC, and the integration of MRAC to SMC scheme. The control schemes are designed based on two versions of the mathematical models of crane systems: one describes the Two-dimensional (2D) motions of overhead crane and the other is extended for Three-dimensional (3D) crane system. The controllers are not only theoretically analyzed but also verified by experimental research. The real-time overhead crane is installed for experiment with xPC Target solution of MathWorks. Overhead crane is an under-actuated system since the number of control inputs is higher than that of actuators. For 2D motion, the control variables composed of trolley movement, payload lifting displacement, and payload swing are driven by two actuators consisted of trolley moving motor and cargo hoisting one. For 3D crane system, five control outputs include bridge movement, trolley displacement, length of payload handling cable, and two cargo swing angles that are controlled by three inputs composed of bridge moving, trolley travelling, and cargo hosting forces. The controllers are first designed for 2D overhead cranes then are improved for 3D motion of crane systems. For 2D motion, the proposed control scheme concurrently carries out four duties: precisely tracking the trolley from the initial point to destination, lifting/lowering the payload to desired length of cable, remaining the cargo vibrating angle small during the transport process, and absolutely eliminating the payload swing at steady destination. In the case of 3D crane application, besides executing four abovementioned duties, the controller must additionally track the bridge motion. In this case, five state variables should be controlled by three input forces. Based on the application of PFL and SMC techniques, together with MRAC, this study proposes eight kinds of nonlinear controllers respectively represented as follows (i) The PFL based controller for 2D crane and its extended version for 3D motion. The proposed nonlinear controller is designed based on the partial feedback linearization of the overall crane dynamics. First, the mathematical model of an 2D overhead crane is separated into actuated and un-actuated dynamics. Then, two partial nonlinear control laws are proposed: one determined from the feedback linearization of the actuated model and the other derived from the feedback linearization of the unactuated model. To guarantee the stability of the system, including those of the actuated and un-actuated states, an upgraded nonlinear control scheme is designed based on the linear combination of the above-mentioned partial nonlinear control laws. Improving the above-proposed controller for 3D motion, another nonlinear controller is considered for 3D overhead crane in case of the most complicated task in which bridge moving, trolley travelling, and cargo hoisting are simultaneously operated. The control structure is designed based on the nonlinear feedback of both un-actuated and actuated states. First, a system dynamics of overhead cranes composed of five highly nonlinear second order differential equations is generated. The system involves two unactuated states and three actuated states driven by three control inputs. Since 3D overhead crane is an under-actuated system, its system dynamics should be separated into actuated model and un-actuated one. Next, the actuated dynamics is “linearized” by using nonlinear feedback technique, and thus, un-actuated model is considered as internal dynamics. Afterward, considering actuated states as system outputs, a nonlinear controller is proposed to track the output trajectories to desired values. However, this nonlinear control scheme does not guarantee the convergence of un-actuated states. Therefore, the structure of this nonlinear controller should be modified to satisfy the stability condition of both actuated and un-actuated states based on the feedback linearization of both system states. The controller structure is now the linear combination of two components: one is received from the feedback linearization of actuated dynamics and the other is acquired from that of un-actuated model. Hopefully, designed controller asymptotically pulls all system responses to desired values. (ii) The SMC based controller for 2D crane and its extended version for 3D motion. SMC approach is classified as Variable Structure Control (VSC) technique that has many advantages. The objective of SMC approach is to determine an asymptotical stable surface so that all state trajectories are attracted to this surface and slide along it until they reach their desired values. SMC is well known to be robust and easy to implement and insensitive with uncertainties and disturbance. It is especially suitable for under-actuated systems having uncertainties. As mentioned, overhead cranes are under-actuated systems containing uncertain components such as cargo mass, friction factors, and so on. Therefore, SMC technique is a proper selection in this study. Two SMC schemes are discussed. One is for 2D motion of crane and the other is improved for 3D crane system. The design process of controllers includes two steps. First, the sliding surface is chosen as a linear combination of position and velocity errors of both actuated and un-actuated states. Then, the control law is designed to force all system states asymptotically approach to sliding surface and roll to desired positions on this surface. To achieve this, the control law involves two components in which the first one, called equivalent input, is to attract all state trajectories to sliding surface, and the second one, called switched action, is to remain the outputs on this surface. The stability of sliding surface is fully proved using Lyapunov’s Direct Method. Furthermore, the constraint formulas of controller gains are established to guarantee the stabilization of all states of crane system based on Lyapunov’s Linearization Technique. As we will see in the simulation and experiment results, the controllers guarantee the robust behaviors of crane system responses: without changing the controller parameters, the shapes of outputs are varied very small in spite of changing the system parameters in a wide range. (iii) PFL & SMC combination based control scheme for 2D crane and its extended version for 3D motion. In this study case, an improved nonlinear control scheme for an 2D overhead crane is proposed applying the combination of two control design techniques. Using partial feedback linearization, the cargo swing vanishing mechanism is constructed. Trolley and cargo tracking control is designed based on the sliding mode technique. Afterwards, a combined control scheme is proposed by integrating the anti-swing structure into an SMC-based algorithm. Owing to the kinematic and geometric constraints between trolley traveling, cargo lifting and cargo swing, the actuators not only drive trolley moving and cargo hoisting directly but also control the cargo swing angle indirectly. Therefore, the control from the sliding mode tracking structure and the effect of the anti-swing mechanism can stabilize the cargo swing and simultaneously track the trolley and cargo to the desired positions accurately. Improving the above-mentioned controller for 3D crane motion, the other control structure is proposed. The control scheme is divided into two mechanisms in which each of them carry out particular duties, and both SMC and Feedback Linearization Control (FLC) techniques are used in control design. Firstly, an anti-swings mechanism is designed based on PFL technique. This mechanism is applied to reduce cargo vibrations in two motion directions and to completely suppress the payload swings at cargo destination. The second mechanism is calculated based on SMC method to accuracy control the positions of trolley and bridge, concurrently, drive the cargo moving up and down along the cable length. After that, a combined nonlinear controller is proposed by integrating PFL based scheme into SMC based structure. The kinematic and geometric relationships between trolley movement, bridge displacement, cable length, and cargo swing angles lead to the mutual effects between two components of combined controller. This kinematic conversation contributes in successful design of overall controller. (iv) The adaptive SMC controller for 2D crane and its extended version for 3D motion. The SMC controllers mentioned in section (ii) are designed in connection with robust control direction without adaptive behavior in which the controllers require to know all crane system parameters. Normally, the overhead crane transports the cargo of the various weights and volumes depending on each operation case. The friction factors characterized by damped coefficients are able to change in terms of operating environment. In this study case, an adaptive version of SMC schemes of section (ii) is developed in the case of no priori knowledge of payload mass and damped elements. The SMC controllers are improved by integrating the adaptation mechanisms into control diagram to estimate the system parameters. In this case, the controllers work well even if some system parameters, such as cargo mass and friction factors, are not clearly known. The controller design process involves two steps. First, a conventional SMC controller is constituted in the case that all system parameters are assumed to be known. Second, the SMC controller is converted into adaptive control structure in which the payload mass and friction factors are considered as adjustable system parameters. Based on Lyapunov stability, an adaptation mechanism is then constructed to adjust the estimated parameters of controller. We first design a controller for 2D motion of crane, and then upgrade it for 3D crane system. Our adaptive robust controllers contain two important properties composed of remaining the shapes of system responses consistently in spite of large variation of system parameters and requiring little information of system parameters. Simulation and real-time experiment. The responses of eight above-proposed controllers are investigated by both numerically simulating the crane dynamics and experimentally researching with the laboratory crane system. Both simulation and experiment results show that the designed controllers guarantee the asymptotical stability of all system trajectories with high qualities of system responses.

Exploration of novel Benzoxathiol-4-ones and Isoflavone type compound as inhibitors of NF-kB and Interleukin-5

Hoang Le Tuan Anh 충남대학교 대학원 2009 국내박사

Trapping and releasing hydrodynamic microarrays device for trapping microparticles : 미세입자의 포획 및 방출을 위한 정수압 마이크로 어레이 장치

Le, Tuan Anh 경북대학교 대학원 2014 국내석사

본 논문은 MEM공정을 이용하여 마이크로어레이 장치의 제작하였으며, 제작된 장치를 이용하여 미세입자의 포획 및 방출에 대하여 연구하였다. 정수압 마이크로어레이 장치는 포토리소그래피공정으로 제작된 마스터 몰드를 이용하여PDMS 캐스팅으로 제작된다. 제작된 정수압 마이크로 어레이 장치는 메인 채널, 보조채널, 마이크로 포스트 등으로 구성되었다. 미세입자로는 76 um 지름의 폴리비드를 계면활성제인 TWEEN 20이 첨가된 탈이온수에 혼합하여 사용하였다. COMSOL 멀티피직스 시뮬레이션을 이용하여 미세채널 내부 형상, 사이즈에 따른 유체거동에 대하여 연구하였다. 시뮬레이션 해석결과, 메인채널이 길이가 증가할수록 입자의 포획이 용이해짐을 확인할 수 있었다. 시뮬레이션 결과를 바탕으로 설계한 마이크로어레이 장치를 이용하여 실험한 결과, 미세입자의 포획률이 100%에 도달하였다. 포획된 입자들의 방출을 위하여 유속을 반대로 흐르게 하였을 때, 포획시 사용했던 유속보다 상대적으로 높은 유속에서 미세입자들이 다시 방출되었으며 유속이 높을 수록 방출률은 점차적으로 증가하는 것을 확인할 수 있었다. Microarrays have an extensive potential for bio-based scientific studies, drug-discovery, and diagnostics. There have been significant developments in two primary classes of microarray: static microarray which consists of bio-molecules immobilized on static substrates and dynamic microarrays which consist of bio-molecules immobilized on mobile substrates, such as microbeads. This thesis describes the fabrication of hydrodynamic microarrays device and trapping technique that can immobile microparticles and retrieve them for resettable purpose. The hydrodynamic microarrays device was fabricated by polydimethylsiloxane (PDMS) casting from SU-8 mold prepared by photolithography. The device has a structure with three parts: main channels, auxiliary channels and microposts. The microparticles were served in this experiment as microbeads with 76.94μm in diameter, were diluted in DI water and TWEEN 20 as a surfactant. This study showed that the efficiency of trapping microparticles does not depend on flow rate of solution, it depends on the dimension of microchannels. The fluid behaviors inside microchannels were also analyzed by simulation via COMSOL Multiphysics. In experiments, we succeeded in arraying 100% microparticles in microposts by using this structure. After that, the trapped microparticles also were retrieved by using high backward flow rates. Microbeads were not completely released from the devices but the results show that more microparticles can be retrieved when increasing the flow rate. Finally, the presented studies were successfully introduced a simple technique with high efficiency to immobile microparticles, the structure of microdevices can be improved to trap more microbeads and completely release microparticles for further purposes.

Development of Candida antarctica lipase B and Serum Paraoxonase 1 through in-silico design and molecular engineering

Le, Quang Anh Tuan Kwangwoon University Graduate School 2015 국내박사

Transparent conductive oxide front-electrodes for high-efficiency silicon-based solar cells

Le, Huy Tuan Anh Sungkyunkwan university 2018 국내석사

In silicon-based photovoltaic technologies, transparent conductive oxide (TCO) front electrodes play a critical role in the achievement of a high-efficiency of solar cells. The goal of this thesis is to develop high-quality TCO front electrodes for the improvement of cell performance of thin-film silicon (TFS) solar cells and silicon heterojunction with intrinsic thin-layer (HIT) solar cells. For the first part of the work, magnetron-sputtered aluminum-doped zinc oxide (AZO) films fabricated by a seed layer approach are reported as front electrodes in superstrate-type TFS solar cells. By controlling the quality of the seed layer through depositing conditions, excellent opto-electrical properties of the AZO films were achieved with a high average transmittance of 87.1% in the wavelength range from 400 to 1100 nm, a high Hall mobility (μ) of 54.04 cm2.V-1.s-1, and a low resistivity (ρ) of 276 μΩ.cm. Further, a crater-like morphology with lateral feature size around 1.2 μm, and a high average haze ratio of 43.3% in the wavelength range from 400 to 1100 nm were obtained, which indicated a good light-trapping capability of these films. Consequently, the thin-film amorphous silicon solar cells using the AZO films fabricated by the seed layer approach showed a high performance with an open-circuit voltage (Voc) of 0.94 V, a short-circuit current density (Jsc) of 15.44 mA/cm2, a fill factor (FF) of 71.89%, and an efficiency (η) of 10.43%. In the final part, an analysis of damage to passivation contact in the HIT solar cells by indium tin oxide (ITO) sputtering under various plasma excitation modes such as radio frequency (RF), direct current (DC), RF-superimposed DC, pulsed DC is presented. Regardless plasma excitation modes, a remarkable degradation of effective carrier lifetime (τeff) of precursor cells after ITO deposition at room temperature was observed. This was attributed to sputter-induced ion bombardment. In details, the τeff of the precursor cells using the ITO films deposited by the RF and RF+DC plasma excitation modes showed a smaller degradation in comparison with that of the precursor cells using the ITO films deposited by the DC and pulsed DC plasma excitation modes. The main reason of this benefit was attributed to a lower ion flux onto the substrate surface in the case of using the RF and RF-superimposed DC rather than a lower ion kinetic energy. The degradation of the τeff in this case could only be partially recovered by the curing process which is a mandatory step in metallization process of screen-printed HIT solar cells. Meanwhile, a small degradation of the τeff due to luminescence of the sputtering plasma was observed regardless plasma excitation modes, and this degradation was fully recovered after the curing process. For ITO deposition at 180 oC, no degradation of the τeff in the case of using RF and RF-superimposed DC and small degradation of the τeff in the case of using DC and pulsed DC were found. This was attributed to the trade-off between degradation and curing step that occurred in ITO deposition at high temperature. In this case, the curing steps in metallization process could not improve the τeff due to their lower temperature in comparison with ITO deposition temperature. The cell performance of HIT solar cells using ITO films deposited by various plasma excitation modes at 180 oC was subsequently discussed. The ITO films deposited by RF-superimposed DC exhibited a high potential as excellent opto-electrical properties and free of sputter-induced damage electrodes for HIT solar cell applications.

적외선 센서창의 냉각 시스템 설계 및 해석

Tuan, Tran le Anh 건국대학교 대학원 2011 국내석사

Aerodynamic heating is one of the most important factors which affect the high flying-speed vehicles. Interceptor missiles fly at a very high speed that leads to a strong effect of aerodynamic heating. It has significant impact on structure of the vehicles, especially on contacting surfaces between the structure and the air. Moreover, an infrared (IR) detection system is settled on the body of the missile. To ensure the properly working of this detection system, a special window is designed to protect the IR detection system from high heating load and allow IR signal to pass through. This special window is cooled down by using a cold nitrogen gas system flowing in the channels inside. To analyze the efficiency of the cooling system before manufacturing the window, simulations are carried out by using SINDA/FLUINT software. During simulations, many parameters of the window model, i.e. materials or geometric values, are changed to find out the best model which satisfies all requirements of working condition. Simulation results are then compared with the experiment results to make sure the model can work well. The temperature distribution on the window structure can be achieved and then applied to ANSYS model to do mechanical behavior analysis of the cooling system model.