Future generation wireless networks are envisioned to provide ubiquitous networking to a wide number of mobile users, promising them the ability to access the various data networks anywhere and anytime. Such networks have motivated the research into efficient management and allocation of the wireless network's limited resources. Heterogeneity also exists amongst the subscribers, i.e. there are those who are willing to spend a little extra on their subscriptions in the prospect of obtaining a better level of service. This work proposes a framework for efficient resource management, while satisfying the heterogeneous QoS demands of the different subscribers. Part of the proposed framework was used to generate mathematical models for the purpose of analyzing the behavior of the system under two different resource management schemes. / February 2006

Future generation wireless networks are envisioned to provide ubiquitous networking to a wide number of mobile users, promising them the ability to access the various data networks anywhere and anytime. Such networks have motivated the research into efficient management and allocation of the wireless network's limited resources. Heterogeneity also exists amongst the subscribers, i.e. there are those who are willing to spend a little extra on their subscriptions in the prospect of obtaining a better level of service. This work proposes a framework for efficient resource management, while satisfying the heterogeneous QoS demands of the different subscribers. Part of the proposed framework was used to generate mathematical models for the purpose of analyzing the behavior of the system under two different resource management schemes.

Future generation wireless networks are envisioned to provide ubiquitous networking to a wide number of mobile users, promising them the ability to access the various data networks anywhere and anytime. Such networks have motivated the research into efficient management and allocation of the wireless network's limited resources. Heterogeneity also exists amongst the subscribers, i.e. there are those who are willing to spend a little extra on their subscriptions in the prospect of obtaining a better level of service. This work proposes a framework for efficient resource management, while satisfying the heterogeneous QoS demands of the different subscribers. Part of the proposed framework was used to generate mathematical models for the purpose of analyzing the behavior of the system under two different resource management schemes.

Video streaming over the best-effort networks is a challenging problem due to the time-varying and uncertain characteristics of the links. When multiple video streams are present in a network, they share and compete for the common bandwidth. In such a setting, a bandwidth allocation algorithm is required to distribute the available resources among the streams in a fair and efficient way. Specifically, it is desired to establish fairness across end-users' Quality of Experience (QoE). In this research, we propose three novel methods to provide QoE-fair network bandwidth allocation among multiple video streaming sessions. First, we formulate the problem of bandwidth allocation for video flows in the context of Network Utility Maximization (NUM) framework, using sigmoidal utility functions, rather than conventional but unrealistic concave functions. An approximation algorithm for Sigmoidal Programming (SP) is utilized to solve the resulting nonconvex optimization problem, called NUM-SP. Simulation results indicate improvements of at least 60% in average utility/QoE and 45% in fairness, while using slightly less network resources, compared to two representative methods. Subsequently, we take a collaborative decision-theoretic approach to the problem of rate adaptation among multiple video streaming sessions, and design a multi-objective foresighted optimization model for network resource allocation. A social welfare function is constructed to capture both fairness and efficiency objectives at the same time. Then, assuming a common altruistic goal for all network users, we use multi-agent decision processes to find the optimal policies for all players. We propose a Decentralized Partially Observable Markov Decision Process (Dec-POMDP) model for the conventional IP networks and a Multi-agent Markov Decision Process (MMDP) model for the SDN-enabled wireless networks. By planning these cooperative decision process models, we find the optimal network bandwidth allocation that leads to social welfare maximization. Distributed multi-agent reinforcement learning algorithms are also designed and proposed as a low-complexity model-free solution to these optimization problems. Simulations of the proposed methods show that the resulting optimal policies of the novel Social Utility Maximization (SUM) framework outperform existing approaches in terms of both efficiency and fairness. The Dec-POMDP model applied to a server-side rate adaptation results in 25% improvement in efficiency and 13% improvement in fairness, compared to one popular protocol of congestion control for multimedia streaming. Our performance evaluations also show that the MMDP model applied to a client-side rate adaptation like DASH improves efficiency, fairness, and social welfare by as much as 18%, 24%, and 25%, respectively compared to current state-of-the-art.

Multimedia applications in wireless communication have shown notable increases over recent years. Specifically, Quality of Service (QoS) has become an important support mechanism in the context of a variety of applications which utilize network resources. The IEEE 802.16 standard for Wireless Metropolitan Area Networks (WirelessMAN) provides a complete QoS control structure designed to enable flow isolation and service differentiation over the common wireless interface. Although the media access control (MAC) mechanisms defined in this standard can offer predefined QoS provisioning on a pre-connection basis (additionally, many algorithms have been addressed to support QoS guarantees for various kinds of applications), the design for a stable, efficient and flexible MAC scheduling algorithm for such QoS provisioning still remains to be discovered. In this paper, we propose a QoS control scheme, Efficient Bandwidth Allocation (EBA), for the WirelessMAN system. Notably, this schedule enables predefined service parameters to control the service provided for each uplink connection and it provides each connection with different service opportunities such that the BS could allocate the most suitable bandwidth constantly for each connection with the various QoS parameters. Moreover, through MAC layer resource allocation, the proposed algorithm is capable of providing QoS guarantee for the SSs under different distance.

In this thesis, we consider a network environment which consists of wired Internet and a wireless broadband network (WiMAX); data from wired or wireless network are all conveyed through WiMAX links to its destination. In order to promise the quality of real-time traffic and allow more transmission opportunity for other traffic types, we propose an Adaptive Bandwidth Allocation (ABA) algorithm for BS to adequately allocate bandwidth. Our ABA algorithm would first reserve required minimum bandwidth for high-priority traffic, such as video streaming. By allocating minimum bandwidth to real-time traffic, the delay time constraint can be satisfied. Other traffic types, such as non-real-time, which have no real-time requirement, may gain extra bandwidth to improve their throughput. For best-effort traffic, the remaining bandwidth can be used to avoid any possible starvation. We build four-dimension Markov chains to evaluate the performance of the proposed ABA algorithm. In the analytical model, we first divide transmission on WiMAX into upload and download phases, and analyze the ABA performance by using Poisson process to generate traffic. At last, by comparing to a previous work, we observe the impacts of different traffic parameters on WiMAX network performance in terms of average delay time, average throughput, and average packet-drop ratio.

In a WiMAX network, four traffic types with different priorities are defined. They are Unsolicited Grant Service (UGS), real-time Polling Service (rtPS), non-real-time Polling Service (nrtPS), and Best Effort (BE). In this thesis, we propose a Dynamic Bandwidth Allocation (DBA) scheme for BS to schedule the four above-mentioned traffic types. By adopting Deficit Round Robin (DRR) scheduling, DBA first assigns minimum quantum to each traffic type for transmission. When rtPS packets exceed their delay constraints, without sacrificing the minimum requirements of nrtPS and BE traffic, DBA borrows some quantum from nrtPS and BE to satisfy the delay requirements of rtPS traffic. When nrtPS packets can not reach the minimum transmission rate, without starving the BE traffic, DBA borrows some quantum from BE to support the required throughput of nrtPS traffic. According to the history record of borrowed quantum, DBA dynamically adjusts the assigned quantum for the three traffic types. For the purpose of evaluation, we use NS-2 to simulate the proposed DBA. We adjust the traffic load to analyze the performance in terms of average packet delay, average throughput, and average packet loss ratio. The simulation results show that the DBA, in comparison to a previous work, can promise the delay constraints of rtPS, maintain the average throughput of nrtPS, and avoid the starvation of BE, when the traffic load is high.

A networked control system (NCS) is a control system where sensors, actuators, and controllers are interconnected over a communication network. This dissertation presents a framework for modeling, stability analysis, optimal control, and bandwidth allocation of the NCS. A ball magnetic-levitation (maglev) system, four DC motor speed-control systems, and a wireless autonomous robotic wheelchair are employed as test beds to illustrate and verify the theoretical results of this dissertation. This dissertation first proposes an output feedback method to stabilize and control the NCSs. The random time delays in the controller-to-actuator and sensor-to-controller links are modeled with two time-homogeneous Markov chains while the packet losses are treated with Dirac delta functions. An asymptotic mean-square stability criterion is established to compensate for the network-induced random time delays and packet losses in the NCS. Then, an algorithm to implement the asymptotic mean-square stability criterion is presented. Experimental results illustrate effectiveness of the proposed output feedback method compared to conventional controllers. The proposed output feedback controller could reduce the errors of the NCS by 13% and 30–40% for the cases without and with data packet losses, respectively. The optimal bandwidth allocation and scheduling of the NCS with nonlinear-programming techniques is also presented in the dissertation. The bandwidth utilization (BU) of each client is defined in terms of its sampling frequency. Two nonlinear approximations, exponential and quadratic approximations, are formulated to describe the system performance governed by discrete-time integral absolute error (DIAE) versus sampling frequency. The optimal sampling frequencies are obtained by solving the approximations with Karush-Kuhn-Tucker (KKT) conditions. Simulation and experimental results are given to verify the effectiveness of the proposed approximations and the bandwidth allocation and scheduling algorithms. In simulations and experiments, the two approximations could maximize the total BU of the NCS up to about 98% of the total available network bandwidth.

The increasing bandwidth demands in access networks motivates network operators, networking devices manufacturers, and standardization institutions to search for new approaches for access networks. These approaches should support higher bandwidth, longer distance between end user and network operator, and less energy consumption. Ethernet Passive Optical Network (EPON) is a favorable choice for broadband access networks. EPONs support transmission rates up to 10 Gbps. EPONs also support distance between end users and central office up to 20 Km. Moreover, optical networks have the least energy consumption among all types of networks. In this dissertation, we focus on reducing delay and saving energy in EPONs. Reducing delay is essential for delay-sensitive traffic, while minimizing energy consumption is an environmental necessity and also reduces the network operating costs. We identify five challenges, namely excess bandwidth allocation, frame delineation, congestion resolution, large round trip time delay in long-reach EPONs (LR-EPONs), and energy saving. We provide a Dynamic Bandwidth Allocation (DBA) approach for each challenge. We also propose a novel scheme that combines the features of the proposed approaches in one highly performing scheme. Our approach is to design novel DBA protocols that can further reduce the delay and be simultaneously simple and fair. We also present a dynamic bandwidth allocation scheme for Green EPONs taking into consideration maximizing energy saving under target delay constraints. Regarding excess bandwidth allocation, we develop an effective DBA scheme called Delayed Excess Scheduling (DES). DES achieves significant delay and jitter reduction and is more suitable for industrial deployment due to its simplicity. Utilizing DES in hybrid TDM/WDM EPONs (TWDM-EPONs) is also investigated. We also study eliminating the wasted bandwidth due to frame delineation. We develop an interactive DBA scheme, Efficient Grant Sizing Interleaved Polling (EGSIP), to compensate the unutilized bandwidth due to frame delineation. Our solution achieves delay reduction ratio up to 90% at high load. We also develop a Congestion Aware Limited Time (CALT) DBA scheme to detect and resolve temporary congestion in EPONs. CALT smartly adapts the optical networking unit (ONU) maximum transmission window according to the detected congestion level. Numerical results show that CALT is more robust at high load compared to other related published schemes. Regarding LR-EPONs, the main concern is large round trip delay mitigation. We address two problems, namely bandwidth over-granting in Multi-Thread Polling (MTP) and on-the-fly void filling. We combine, with some modifications, EGSIP and DES to resolve bandwidth over-granting in MTP. We also manage to adaptively tune MTP active running threads along with the offered load. Regarding on-the-fly void filling, Our approach, Parallel Void Thread (PVT), achieves large delay reduction for delay-sensitive traffic. PVT is designed as a plus function to DBA and can be combined with almost all DBA schemes proposed before. The powerful feature of our proposed solutions is integrability. We integrate our solutions together and form a multi-feature, robust, fairly simple, and well performing DBA scheme over LR-TWDM-EPONs. Our final contribution is about energy saving under target delay constraints. We tackle the problem of downstream based sleep time sizing and scheduling under required delay constraints. Simulation results show that our approach adheres to delay constraints and achieves almost ideal energy saving ratio at the same time.

Variable bit rate (VBR) compressed video traffic, such as live video news, is expected to account for a large portion of traffic in future integrated networks. This real-time video traffic has strict delay and loss requirements, and exhibits burstiness over multiple time scales, thus imposing a challenge on network resource allocation and management. The renegotiated VBR (R-VBR) scheme, dynamically allocating resources to capture the burstiness of VBR traffic, substantially increases network utilization while satisfying any desired quality of service (QoS) requirements. This thesis focuses on the performance evaluation of R-VBR in the context of different R-VBR approaches. The renegotiated deterministic VBR (RED-VBR) scheme, proposed by Dr. H. Zhang et al., is thoroughly investigated in this research using a variety of real-world videos, with both high quality and low quality. A new Virtual-Queue-Based RED-VBR is then developed to reduce the implementation complexity of RED-VBR. Simulation results show that this approach obtains a comparable network performance as RED-VBR: relatively high network utilization and a very low drop rate. A Prediction-Based R-VBR based on a multiresolution learning neural network traffic predictor, developed by Dr. Y. Liang, is studied and the use of binary exponential backoff (BEB) algorithm is introduced to efficiently decrease the renegotiation frequency. Compared with RED-VBR, Prediction-Based R-VBR obtains significantly improved network utilization at a little expense of the drop rate. This work provides evaluations of the advantages and disadvantages of several R-VBR approaches, and thus provides a clearer big picture on the performance of the studied R-VBR approaches, which can be used as the basis to choose an appropriate R-VBR scheme to optimize network utilization while enabling QoS for the application tasks. / Master of Science