With the rapid development of Internet technology and the emergence of more flexible applications, dynamic allocation of network resources is becoming more and more significant. Since bandwidth resource is limited, us...
详细信息
ISBN:
(纸本)9781467372183
With the rapid development of Internet technology and the emergence of more flexible applications, dynamic allocation of network resources is becoming more and more significant. Since bandwidth resource is limited, users' dynamic allocation requests cannot always be granted. Therefore, scheduling the dynamic requests to achieve optimal bandwidth utilization or service rate is of great importance. However, this is difficult to achieve in current networks due to the lack of programming interface of network devices for effective and centralized control and management. In recent years, the emergence of software-defined networks (SDN) provides us an opportunity to realize the goal. Centralized control and programmability of SDN let network managers configure, manage and optimize network resources very easily and quickly. In this paper, a dynamic scheduling algorithm based on an optimization model for scheduling reservation-based bandwidth allocation requests is proposed to achieve optimal resource utilization. Simulation result shows that the scheduling algorithm proposed in this paper could achieve higher bandwidth utilization compared to another two scheduling algorithms.
One of the important issues in building differentiated services is how to support relative differentiated queueing delay services for packets from different classes. This paper presents a new packet scheduling algorit...
详细信息
ISBN:
(纸本)0780379454
One of the important issues in building differentiated services is how to support relative differentiated queueing delay services for packets from different classes. This paper presents a new packet scheduling algorithm, called Virtual-Length, to address this issue. It provides effective control over the relative average queueing delay ratio between two classes of traffic by monitoring their average arrival rates and cumulative queue lengths. The packets of different classes are forwarded according to their weighted queue lengths with respect to predefined proportional delay parameters. The simulation results show that VirtualLength is capable of providing controllable and predictable services for classes with different load distributions under various load conditions. In particular, it can provide more accurate control over classes of traffic in short timescales than other packet scheduling algorithms, and comparable service in long timescales. The results also show that the performance of VirtualLength is independent of the distributions of packet arrivals and sizes.
Most embedded real lime systems are based on uniprocessor architecture, and deadline miss ratio is always a mainstay for any type of real time system. This paper presents the design and analysis of a new scheduling al...
详细信息
ISBN:
(纸本)0780392477
Most embedded real lime systems are based on uniprocessor architecture, and deadline miss ratio is always a mainstay for any type of real time system. This paper presents the design and analysis of a new scheduling algorithm, Dynamic Queue Deadline First (DQDF) to handle scheduling of a dynamic multiple tasks in real time systems. Soft real-time processing is a realtime processing in which some or all applications are allowed to miss deadlines, particularly, in situations of system load. In this paper, a new scheduler is introduced to minimize the overhead and percentage of deadline misses to improve the system utilization and fairness. We have used Earliest Deadline First (EDF) with dynamic queuing technique, as it is an optimal scheduling algorithm for uniprocessor systems. Simulation results show that with DQDF algorithm, the deadline miss ratio is improved and fairness is restored by using static priority approach. The comparison of the new scheduler DQDF with simple EDF scheduler results the performance improvement upto 68%.
Designing a scheduling algorithm for data centers is a challenging task. In this paper, we propose a scheduling algorithm for cloud computing in data centers. A major feature of the proposed scheduling algorithm is th...
详细信息
ISBN:
(纸本)9781467379106
Designing a scheduling algorithm for data centers is a challenging task. In this paper, we propose a scheduling algorithm for cloud computing in data centers. A major feature of the proposed scheduling algorithm is that it determines the schedulability of incoming tasks before scheduling them. The performance of the proposed scheduling algorithm is evaluated through extensive simulations. We evaluate the impact of the number of tasks and the size of the workload on the energy consumed by a data center.
Modern computational experiments imply that the resources of the cloud computing environment are often used to solve a large number of tasks, which differ only in the values of a relatively small set of simulation par...
详细信息
ISBN:
(纸本)9783319556680;9783319556697
Modern computational experiments imply that the resources of the cloud computing environment are often used to solve a large number of tasks, which differ only in the values of a relatively small set of simulation parameters. Such sets of tasks may occur while implementing multivariate calculations aimed at finding the simulation parameter values, which optimize certain characteristics of the computational model. Applications of this type make a large percentage of modern HPC systems load, which implies a need for methods and algorithms for efficient allocation of resources in order to optimize systems for solving such problems. The aim of this work is to implement a PO-HEFT problem-oriented scientific workflow scheduling algorithm and to compare it with other workflow scheduling algorithms.
Due to the low cost, high bandwidth and compatibility of Ethernet, Ethernet-based fronthaul has been utilized to carry sampled radio frequency (RE) signals from radio equipment (RE) to the radio equipment controller (...
详细信息
ISBN:
(纸本)9781728129990
Due to the low cost, high bandwidth and compatibility of Ethernet, Ethernet-based fronthaul has been utilized to carry sampled radio frequency (RE) signals from radio equipment (RE) to the radio equipment controller (REC). Meeting the stringent performance requirements regarding jitter for the Common Public Radio Interface (CPRI) over Ethernet is challenging. In this paper, a time division multiplexing (TDM) slot-based scheduling (TSS) algorithm is proposed to minimize the jitter;this algorithm consists of two main modules: slot greedy allocation and low-delay collaboration. Specifically, low complexity slot greedy allocation aims to solve the NP-hard problem caused by the optimal slot allocation, and the low-delay collaboration among different switches is employed to avoid the large delays caused by a strict slot assignment. In the Ethernet based fronthaul network, the simulation results demonstrate that, compared with the conventional benchmark algorithm, the TSS algorithm ensures lower jitter, and this significant decrease is achieved without incurring a large delay.
During data transmission, uneven distribution of network communication resources occurs frequently due to competition for limited network communication resources, which results in low efficiency of high priority data ...
详细信息
ISBN:
(纸本)9781665440899
During data transmission, uneven distribution of network communication resources occurs frequently due to competition for limited network communication resources, which results in low efficiency of high priority data transmission. To solve the above problems, this paper proposes the priority load-aware scheduling round-robin (PLAWRR) scheduling algorithm. It firstly introduces the parameters of average data packet size and remaining queue length to weight calculation in order to allocate network communication resources in a balanced manner. Secondly, it dynamically adjusts the weight by quantifying the load priority to ensure that the high priority queue can transmit more efficiently. The sImulation results show that the proposed algorithm reduces the scheduling delay and the packet loss rate of data transmission as well as increasing the throughput.
The rapid evolution of the wireless communication technologies along with the growing number of connected vehicles and the limitation of available radio resources make scheduling tasks very important for LTE-V2X vehic...
详细信息
ISBN:
(纸本)9781538677476
The rapid evolution of the wireless communication technologies along with the growing number of connected vehicles and the limitation of available radio resources make scheduling tasks very important for LTE-V2X vehicular networks. Accordingly, this paper presents a new optimal scheduling algorithm for V2X connections called DSA-PA, which improves performances in terms of Quality of Service (QoS), throughput and fairness. The proposed algorithm consider the traffic classification imposed by V2X exigency and aims to maximize the system throughput by taking into account the channel conditions of Vehicular User Equipment (VUE) expressed by the Signal-to-Interference-plus-Noise Ratio (SINR) value. The optimal distribution of the available Resources Blocks (RBs) is based on a dynamic reallocation process depending in the average blocking rate value (ABR) especially for vehicular Safety Traffic (ST).
5G promises very high throughput and low latency. It will be able to support the exchange of large amounts of data with a wide variety of communications, including Vehicle-to-Everything (V2X) services. Conventional re...
详细信息
5G promises very high throughput and low latency. It will be able to support the exchange of large amounts of data with a wide variety of communications, including Vehicle-to-Everything (V2X) services. Conventional resource allocation techniques based on orthogonal multiple access (OMA) appear to be unsuitable for a dense network due to limited resources available. This work presents a new planning algorithm called SAVCN (scheduling algorithm for V2X Communication based on NOMA), for the 5G network. The main feature of NOMA is the same resource can be shared by multiple users. The purpose of our algorithm is to improve network performance in terms of throughput, equity, the number of V2X users served and error rates. In fact, SAVCN efficiently allocates available resource blocks (Rbs) to maximize system throughput by taking into account a well-defined criterion on the minimum distance between transmitters and receivers. The simulation results indicate promising performance for SAVCN.
A new scheduling, algorithm, which aims to provide proportional and controllable QoS in terms of burst loss probability for OBS (optical burst switching) networks, is proposed on the basis of a summary of current QoS ...
详细信息
ISBN:
(纸本)0819455792
A new scheduling, algorithm, which aims to provide proportional and controllable QoS in terms of burst loss probability for OBS (optical burst switching) networks, is proposed on the basis of a summary of current QoS schemes in OBS. With simulations, performance analyses and comparisons are studied in detail. The results show that, in the proposed scheme, burst loss probabilities are proportional to the given factors and the control of QoS performance can be achieved with better performance. This scheme will be beneficial to the OBS network management and the tariff policy making.
暂无评论