In MPLS networks, CR-LDP (Constraint-Based Routed Label Distribution Protocol) provides traffic engineering and QoS (Quality of Service) by distributing labels along the path. One of the innovative ideas in CR-LDP is right in the capability of bandwidth preemption. Bandwidth preemption allows a high-priority traffic flow to pre-empt the low-priority traffic flow when there is not enough bandwidth for conveying the high-priority flow. However, it is not effective in assuring QoS by simply pre-empting the low-priority flow. Therefore, in the thesis, we present a bandwidth preemption with negotiation (BPN) architecture for MPLS networks. In BPN, a high-priority flow will have to negotiate the bandwidth with Ingress Switch Router (ISR) before it can actually pre-empt the bandwidth being used by the low-priority flow. A network link-state database in ISR is designed to record the remaining bandwidth for each priority class. ISR determines whether a high-priority flow can pre-empt a low-priority flow based on the condition that the low-priority flow is possible to switch to other paths with an equal bandwidth along the path. In order to evaluate the performance of our proposed BPN mechanisms, we modify the MNS (MPLS Network Simulator) by adding a bandwidth negotiation module. For the comparisons, we design two topologies for simulating the proposed BPN and the traditional bandwidth preemption with force mode. It is observed that the BPN have exhibited better performance in average throughput and packet loss rate than the traditional bandwidth preemption, not matter either a complex or a regular topology is used. Finally, we analyze the BPN algorithm complexity by some network parameters, and compare the complexity with that of traditional bandwidth preemption

In an ad hoc networks, it is an important problem about how to increase the bandwidth utilization and guarantee Quality-of-Service. In order to increase the bandwidth utilization under the situation of limited bandwidth, we want to achieve the target by using the bitty residual bandwidth in the ad hoc networks. Hence we integrate several single paths into a multi-path that satisfies QoS requirement. In this paper, we present a policy to look up, maintain, and reroute the virtual circuit in the ad hoc networks. From the result of simulation, we obtain that our proposal indeed can increase the bandwidth utilization of ad hoc networks. But the condition of splitting also increases the overhead of connection, and affects the forced terminated ratio and packets lost ratio under mobility.

Future processors will integrate an increasing number of cores because the scaling of single-thread performance is limited and because smaller cores are more power efficient. Off-chip memory bandwidth that is shared between those many cores, however, scales slower than the transistor (and core) count does. As a result, in many future systems, off-chip bandwidth will become the bottleneck of heavy demand from multiple cores. Therefore, optimally managing the limited off-chip bandwidth is critical to achieving high performance and efficiency in future systems. In this dissertation, I will develop techniques to optimize the shared use of limited off-chip memory bandwidth in chip-multiprocessors. I focus on issues that arise from the sharing and exploit the differences in memory access characteristics, such as locality, bandwidth requirement, and latency sensitivity, between the applications running in parallel and competing for the bandwidth. First, I investigate how the shared use of memory by many cores can result in reduced spatial locality in memory accesses. I propose a technique that partitions the internal memory banks between cores in order to isolate their access streams and eliminate locality interference. The technique compensates for the reduced bank-level parallelism of each thread by employing memory sub-ranking to effectively increase the number of independent banks. For three different workload groups that consist of benchmarks with high spatial locality, low spatial locality, and mixes of the two, the average system efficiency improves by 10%, 7%, 9% for 2-rank systems, and 18%, 25%, 20% for 1-rank systems, respectively, over the baseline shared-bank system. Next, I improve the performance of a heterogeneous system-on-chip (SoC) in which cores have distinct memory access characteristics. I develop a deadline-aware shared memory bandwidth management scheme for SoCs that have both CPU and GPU cores. I show that statically prioritizing the CPU can severely constrict GPU performance, and propose to dynamically adapt the priority of CPU and GPU memory requests based on the progress of GPU workload. The proposed dynamic bandwidth management scheme provides the target GPU performance while prioritizing CPU performance as much as possible, for any CPU-GPU workload combination with different complexities. / text

ITC/USA 2013 Conference Proceedings / The Forty-Ninth Annual International Telemetering Conference and Technical Exhibition / October 21-24, 2013 / Bally's Hotel & Convention Center, Las Vegas, NV / Telemetry Network System (TmNS) contains a centralized link manager which allows efficient use of the frequency spectrum by dynamically allocating capacity to transmit based on need and priority. To verify the accurate operation of the telemetry system link management algorithm prior to system demonstration, a combination of novel techniques were leveraged in the areas of modeling and simulation, and test bed verification. This paper will cover the process of verifying the link management algorithm from the use of the OPNET iNET simulation to test bed radio simulators along with the developed test bed tools used to capture the results.

ITC/USA 2012 Conference Proceedings / The Forty-Eighth Annual International Telemetering Conference and Technical Exhibition / October 22-25, 2012 / Town and Country Resort & Convention Center, San Diego, California / In the Telemetry Network System (TmNS) the prioritization of mission traffic is performed by internal radio queuing disciplines and the Link Manager performing adjustments to each radio transmit window in the TDMA network. These processes combine to provide the QoS traffic handling across the range. The radios provide the transmit packet prioritization using queuing disciplines which can be assigned to mission traffic flows. The Link Manager on the ground receives periodic reports of queue levels from each radio and performs transmit capacity adjustments based on internal radio and network wide conditions. Presented in this paper is the design of this TmNS RF Link Management QoS process with OPNET modeling and simulation results.

High Speed Downlink Packet Access (HSDPA) is a cellular system that was standardized by the 3rd Generation Partnership Project (3GPP). HSDPA can support data rates of up to 14.4 Mbps through the use of a shared channel. Due to its high transmission rates, the highly popular multimedia applications are converging over this network. Moreover, as the shared channel is assigned to a single user in a given time interval, the scheduling decision is considered as a crucial one. Conventional HSDPA scheduling schemes utilize the fluctuations in channel condition to maximize system throughput by selecting users with relatively good radio conditions. However, this raises the issue of fairness as users with relatively poor channel conditions might not be served and consequently may suffer from starvation. Furthermore, Real-Time (RT) applications have strict delay constraints and require that packets are transmitted within a certain delay threshold. In this thesis, a Delay Based Scheduler (DBS) is proposed for HSDPA which aims at minimizing the average queuing delay at the packet scheduler without compromising system throughput and fairness. In addition, the scheme can balance the tradeoff between throughput maximization and the minimization of queuing delay through the attunement of a parameter, thus allowing the service provider to choose between these two metrics. The DBS maintains the delay constraints of RT applications by defining delay thresholds for each traffic class and dropping packets that exceed their delay limit. The DBS accommodates Quality of Service (QoS) prioritization by defining and utilizing desired QoS parameters in the scheduling assignment. Finally, it was mathematically shown that the DBS can converge to a Non-Real-Time (NRT) scheme known as the Max CIR algorithm, allowing the scheduler to support RT and NRT applications simultaneously. The performance of the DBS was evaluated and compared to other well known schemes. It was found that the DBS can minimize the aggregate queuing delay of the system and maintain similar throughput and fairness. / Thesis (Master, Computing) -- Queen's University, 2007-08-15 16:12:36.22

Voice over Internet Protocol (VoIP) provides a new telephony approach where the voice traffic passes over Internet Protocol shared traffic networks. VoIP is a significant application of the converged network principle. The research aim is to model VoIP over a hybrid Universal Mobile Telecommunications System (UMTS) network and to identify an improved approach to applying the ITU-T Recommendation G.107 (E-Model) to understand possible Quality of Service (QoS) outcomes for the hybrid UMTS network. This research included Modeling the hybrid UMTS network and carrying out simulations of different traffic types transmitted over the network. The traffic characteristics were analysed and compared with results from the literature. VoIP traffic was modelled over the hybrid UMTS network and the VoIP traffic was generated to represent different loads on the network from light to medium and heavy VoIP traffic. The VoIP over hybrid UMTS network traffic results were characterized and used in conjunction with the E-Model to identify VoIP QoS outcomes. The E-Model technique was implemented and results achieved were compared with results for other network types highlighted in the literature. The research identified an approach that permits accurate Modeling of VoIP QoS over a hybrid UMTS network. Accurate results should allow network design to facilitate new approaches to achieving an optimal network implementation for VoIP.

Master of Engineering by Research / Service-level-agreement (SLA) monitoring measures network Quality-of-Service (QoS) parameters to evaluate whether the service performance complies with the SLAs. It is becoming increasingly important for both Internet service providers (ISPs) and their customers. However, the rapid expansion of the Internet makes SLA monitoring a challenging task. As an efficient method to reduce both complexity and overheads for QoS measurements, sampling techniques have been used in SLA monitoring systems. In this thesis, I conduct a comprehensive study of sampling methods for network QoS measurements. I develop an efficient sampling strategy, which makes the measurements less intrusive and more efficient, and I design a network performance monitoring software, which monitors such QoS parameters as packet delay, packet loss and jitter for SLA monitoring and verification. The thesis starts with a discussion on the characteristics of QoS metrics related to the design of the monitoring system and the challenges in monitoring these metrics. Major measurement methodologies for monitoring these metrics are introduced. Existing monitoring systems can be broadly classified into two categories: active and passive measurements. The advantages and disadvantages of both methodologies are discussed and an active measurement methodology is chosen to realise the monitoring system. Secondly, the thesis describes the most common sampling techniques, such as systematic sampling, Poisson sampling and stratified random sampling. Theoretical analysis is performed on the fundamental limits of sampling accuracy. Theoretical analysis is also conducted on the performance of the sampling techniques, which is validated using simulation with real traffic. Both theoretical analysis and simulation results show that the stratified random sampling with optimum allocation achieves the best performance, compared with the other sampling methods. However, stratified sampling with optimum allocation requires extra statistics from the parent traffic traces, which cannot be obtained in real applications. In order to overcome this shortcoming, a novel adaptive stratified sampling strategy is proposed, based on stratified sampling with optimum allocation. A least-mean-square (LMS) linear prediction algorithm is employed to predict the required statistics from the past observations. Simulation results show that the proposed adaptive stratified sampling method closely approaches the performance of the stratified sampling with optimum allocation. Finally, a detailed introduction to the SLA monitoring software design is presented. Measurement results are displayed which calibrate systematic error in the measurements. Measurements between various remote sites have demonstrated impressively good QoS provided by Australian ISPs for premium services.

Doctor of Philosophy(PhD) / Recent advances in miniaturization and low power design have led to a flurry of activity in wireless sensor networks. Sensor networks have different constraints than traditional wired networks. A wireless sensor network is a special network with large numbers of nodes equipped with embedded processors, sensors, and radios. These nodes collaborate to accomplish a common task such as environment monitoring or asset tracking. In many applications, sensor nodes will be deployed in an ad-hoc fashion without careful planning. They must organize themselves to form a multihop, wireless communication network. In sensor network environments, much research has been conducted in areas such as power consumption, self-organisation techniques, routing between the sensors, and the communication between the sensor and the sink. On the other hand, real-time communication with the Quality of Service (QoS) concept in wireless sensor networks is still an open research field. Most protocols either ignore real time or simply attempt to process as fast as possible and hope that this speed is sufficient to meet the deadline. However, the introduction of real-time communication has created additional challenges in this area. The sensor node spends most of its life routing packets from one node to another until the packet reaches the sink; therefore, the node functions as a small router most of the time. Since sensor networks deal with time-critical applications, it is often necessary for communication to meet real time constraints. However, research that deals with providing QoS guarantees for real-time traffic in sensor networks is still in its infancy.This thesis presents a real-time communication framework to provide quality of service in sensor networks environments. The proposed framework consists of four components: First, present an analytical model for implementing Priority Queuing (PQ) in a sensor node to calculate the queuing delay. The exact packet delay for corresponding classes is calculated. Further, the analytical results are validated through an extensive simulation study. Second, report on a novel analytical model based on a limited service polling discipline. The model is based on an M/D/1 queuing system (a special class of M/G/1 queuing systems), which takes into account two different classes of traffic in a sensor node. The proposed model implements two queues in a sensor node that are served in a round robin fashion. The exact queuing delay in a sensor node for corresponding classes is calculated. Then, the analytical results are validated through an extensive simulation study. Third, exhibit a novel packet delivery mechanism, namely the Multiple Level Stateless Protocol (MLSP), as a real-time protocol for sensor networks to guarantee the traffic in wireless sensor networks. MLSP improves the packet loss rate and the handling of holes in sensor network much better than its counterpart, MMSPEED. It also introduces the k-limited polling model for the first time. In addition, the whole sending packets dropped significantly compared to MMSPEED, which it leads to decrease the consumption power. Fourth, explain a new framework for moving data from the sink to the user, at a low cost and low power, using the Universal Mobile Telecommunication System (UMTS), which is standard for the Third Generation Mobile System (3G). The integration of sensor networks with the 3G mobile network infrastructure will reduce the cost of building new infrastructures and enable the large-scale deployment of sensor networks

Made available in DSpace on 2014-06-12T15:50:41Z (GMT). No. of bitstreams: 1 license.txt: 1748 bytes, checksum: 8a4605be74aa9ea9d79846c1fba20a33 (MD5) Previous issue date: 2008 / A popularização da Internet e a demanda por acesso de alta velocidade levou ao desenvolvimento da Broadband Wireless Access. Apesar do seu grande potencial, a comunicação via rádio impõe alguns desafios. Uma grande limitação é o próprio meio de transmissão devido a efeitos inerentes à propagação de radio como o path loss, frequency selective fading, espalhamento Doppler e multipath delay-spread. Nesse contexto, o OFDM é uma tecnologia promissora por causa de sua tolerância a problemas de perdas e multi-caminho. Devido à combinação de canais independentes, é possível usar diferentes modulações em cada sub-carrier, de acordo com as condições do canal. Esta técnica é conhecida como adaptive modulation and coding. Além disso, em uma arquitetura ponto a multi-ponto, múltiplos usuários podem compartilhar o espectro ao se atribuir diferentes conjuntos de sub-carriers, tirando vantagem do um efeito conhecido como diversidade multi-usuário. Em comparação com outras técnicas de múltiplo acesso, o OFDMA permite um melhor aproveitamento da diversidade multi-usuário com a possibilidade de uma alocação com alta granularidade. Muitas pesquisas têm investigado técnicas adaptativas capazes de melhorar a eficiência espectral em sistemas multi-usuário. Essas técnicas são normalmente formuladas como constraint optimization problems, conhecidos por serem NP-hard. Neste trabalho, adotamos uma abordagem heurística para lidar com esse tipo de problema. O objetivo principal é desenvolver uma estratégia de alocação fazendo uso eficiente dos recursos disponíveis e maximizando a eficiência espectral total. Entretanto, um estratégia que apenas procura maximizar a eficiência espectral pode gerar um problema relacionado à justiça no compartilhamento de recursos. Outrossim, com a popularização das redes sem fio, é esperado que elas sejam capazes de prover uma maior variedade de serviços com diferentes requisites de QoS e largura de banda. Portanto, procuramos desenvolver um algoritmo que permita ao operador da rede definir esses requisitos. De acordo com eles, o algoritmo deve fornecer o maior throughput possível dentro dos limites estabelecidos por essas restrições