Load balancing in processor sharing systems

We investigate optimal load balancing strategies for a multi-class multi-server processor-sharing system with a Poisson input stream, heterogeneous service rates, and a server-dependent holding cost per unit time. Specifically, we study (i) the centralized setting in which a dispatcher routes incoming jobs based on their service time requirements so as to minimize the weighted mean sojourn time in the system; and (ii) the decentralized, distributed non-cooperative setting in which each job, aware of its service time, selects a server with the objective of minimizing its weighted mean sojourn time in the system. For the decentralized setting we show the existence of a potential function, which allows us to transform the non-cooperative game into a standard convex optimization problem. For the two aforementioned settings, we characterize the set of optimal routing policies and obtain a closed form expression for the load on each server under any such policy. Furthermore, we show the existence of an optimal policy that routes a job independently of its service time requirement. We also show that the set of servers used in the decentralized setting is a subset of set of servers used in the centralized setting. Finally, we compare the performance perceived by jobs in the two settings by studying the so-called Price of Anarchy (PoA), that is, the ratio between the decentralized and the optimal centralized solutions. When the holding cost per unit time is the same for all servers, it is known that the PoA is upper bounded by the number of servers in the system. Interestingly, we show that the PoA for our system can be unbounded. In particular this indicates that in our system, the performance of selfish routing can be extremely inefficient.     

Optimal control of a removable and non-reliable server in a markovian queueing system with finite capacity

This paper studies the economic behavior of a removable and non-reliable server in an Markovian queueing system with finite capacity under steady-state conditions. The removable server applies the N policy which turns the server on when the queue length reaches the value N, and turns the server off when the system is empty. The server may break down only if operating and require repair at a repair facility. Interarrival and service times of the customers, and breakdown and repair times of the server, are assumed to follow a negative exponential distribution. A cost model is developed to determine the optimal operating N policy numerically in order to minimize the total expected cost per unit time.     

Cost minimization in placing service chains for virtualized network functions

Network function virtualization (NFV) places network functions onto the virtual machines (VMs) of physical machines (PMs) located in data centers. In practice, a data flow may pass through multiple network functions, which collectively form a service chain across multiple VMs residing on the same or different PMs. Given a set of service chains, network operators have two options for placing them: (a) minimizing the number of VMs and PMs so as to reduce the server rental cost or (b) placing VMs running network functions belonging to the same service chain on the same or nearby PMs so as to reduce the network delay. In determining the optimal service chain placement, operators face the problem of minimizing the server cost while still satisfying the end‐to‐end delay constraint. The present study proposes an optimization model to solve this problem using a nonlinear programming (NLP) approach. The proposed model is used to explore various operational problems in the service chain placement field. The results suggest that the optimal cost ratio for PMs with high, hybrid, and low capacity, respectively, is equal to 4:2:1. Meanwhile, the maximum operating utilization rate should be limited to 55% in order to minimize the rental cost. Regarding quality of service (QoS) relaxation, the server cost reduces by 20%, 30%, and 32% as the end‐to‐end delay constraint is relaxed from 40 to 60, 80, and 100 ms, respectively. For the server location, the cost decreases by 25% when the high‐capacity PMs are decentralized rather than centralized. Finally, the cost reduces by 40% as the repetition rate in the service chain increases from 0 to 2. A heuristic algorithm, designated as common sub chain placement first (CPF), is proposed to solve the service chain placement problem for large‐scale problems (eg, 256 PMs). It is shown that the proposed algorithm reduces the solution time by up to 86% compared with the NLP optimization model, with an accuracy reduction of just 8%.     

Towards a Performance-Based Resource Server Access Control in Future Grid Communication Environments

With the rapid advances in communication and networking, especially in Grid technique, an increasing number of applications will involve computing systems. These applications in turn create an increasing demand for efficient resource management, request handling policies and access control. In this paper, we propose an efficient access control algorithm to protect the critical resource of server and improve the performance of the future grid communication computing system. Stability of CPU utilization is aimed to protect the server from overload and under-load. It is then beneficial to keep a satisfactory response time of requests, high throughput and less potential loss of service. We analyze the stability in detail and present a method for tuning control gains in order to guarantee the system stability. Finally, we perform simulations to evaluate the performance of the proposed algorithm. Simulation results demonstrate that the proposed algorithm stabilizes the utilization of CPU in the computing system if the control gains are appropriately chosen on the basis of system stability. It then achieves satisfactory performance.     

Integration of firefly optimization and Pearson service correlation for efficient cloud resource utilization

Cloud Computing (CC) environment presents a simplified, centralized platform or resources to usage while necessitated at minimum cost. In CC, the main processes in is the allocation of resources of web applications. However, with the increasing demands of Cloud User (CU), an efficient resource allocation technique for web applications is required. According to the request made by the user and response obtained, the cost of resources has also to be optimized. To overcome such limitations, Pearson service correlation‐based firefly resource cost optimization (PSC‐FRCO) technique is designed. Pearson service correlation‐based firefly resource cost optimization technique not only improves the performance of cost aware resource allocation but also achieves higher efficiency while rendering services in cloud computing environment for web applications. Pearson service correlation‐based firefly resource cost optimization technique initially uses Pearson service correlation in which the user‐required service is identified by correlating the available services provided by cloud owner. This helps in improving the Response Time (RT) of cloud service provisioning. Next, firefly resource cost optimization algorithm is applied to identify and allocate the cost‐optimized cloud resources to users to afford required service from the cloud server. Thus, PSC‐FRCO technique improves the Resource Utilization Efficiency (RUE) of web applications with minimal computational cost. This technique conducts experimental works on parameters such as RT, Bandwidth Utilization Rate (BUR) computational cost, Energy Consumption (EC), and RUE. Experimental analysis reveals that PSC‐FRCO technique enhances enhances RUE and lessens RT as compared to state‐of‐the‐art works.     

Resource allocation and dynamic power control for D2D communication underlaying uplink multi-cell networks

Underlaying device-to-device (D2D) communication is suggested as a promising technology for the next generation cellular networks (5G), where users in close proximity can transmit directly to one another bypassing the base station. However, when D2D communications underlay cellular networks, the potential gain from resource sharing is highly determined by how the interference is managed. In order to mitigate the resource reuse interference between D2D user equipment and cellular user equipment in a multi-cell environment, we propose a resource allocation scheme and dynamic power control for D2D communication underlaying uplink cellular network. Specifically, by introducing the fractional frequency reuse (FFR) principle into the multi-cell architecture, we divide the cellular network into inner region and outer region. Combined with resource partition method, we then formulate the optimization problem so as to maximize the total throughput. However, due to the coupled relationship between resource allocation and power control scheme, the optimization problem is NP-hard and combinational. In order to minimize the interference caused by D2D spectrum reuse, we solve the overall throughput optimization problem by dividing the original problem into two sub-problems. We first propose a heuristic resource pairing algorithm based on overall interference minimization. Then with reference to uplink fractional power control (FPC), a dynamic power control method is proposed. By introducing the interference constraint, we use a lower bound of throughput as a cost function and solve the optimal power allocation problem based on dual Lagrangian decomposition method. Simulation results demonstrate that the proposed algorithm achieves efficient performance compared with existing methods.     

An on-line, business-oriented optimization of performance and availability for utility computing

Utility computing provides a pay-as-you-go approach to information systems in which application providers (e.g., web sites) can better manage their costs by adding capacity in response to increased demands and shedding capacity when it is no longer needed. This paper addresses application providers who use clusters of servers. Our work develops a framework to determine the number of servers that minimizes the sum of quality-of-service (QoS) costs resulting from service level penalties and server holding costs for the server cluster. The server characteristics considered are service rate, failure rates, repair rates, and costs. The contributions of this paper are: 1) a model for the performance and availability of an e-Commerce system that is consistent with data from a multisystem testbed with an e-Commerce workload; 2) a business-oriented cost model for resource allocation for application providers; 3) a closed form approximation for the optimal allocation of servers for an application provider based on the performance model in 1) and the cost model in 2); and 4) a simple criteria for utility owners and server manufacturers to make tradeoffs between server characteristics.     

Estimation‐Based Load‐Balancing with Admission Control for Cluster Web Servers

The growth of the World Wide Web and web‐based applications is creating demand for high performance web servers to offer better throughput and shorter user‐perceived latency. This demand leads to widely used cluster‐based web servers in the Internet infrastructure. Load balancing algorithms play an important role in boosting the performance of cluster web servers. Previous load balancing algorithms suffer a significant performance drop under dynamic and database‐driven workloads. We propose an estimation‐based load balancing algorithm with admission control for cluster‐based web servers. Because it is difficult to accurately determine the load of web servers, we propose an approximate policy. The algorithm classifies requests based on their service times and tracks the number of outstanding requests from each class in each web server node to dynamically estimate each web server load state. The available capacity of each web server is then computed and used for the load balancing and admission control decisions. The implementation results confirm that the proposed scheme improves both the mean response time and the throughput of clusters compared to rival load balancing algorithms and prevents clusters being overloaded even when request rates are beyond the cluster capacity.     

Low Complexity Modified Viterbi Decoder with Convolution Codes for Power Efficient Wireless Communication

To attain high quality of service (QoS) with efficient power consumption with minimum delay through Wireless Local Area Network (WLAN) through mesh network is an important research area. But the existing real-time routing system involves multiple hops with time varying mobility channels for fastest data propagation is greatly degraded with power utilization factor through congestion traffic queue. Required allocation and resource management through desired access points plays vital roles in which multiple hops demands delay rates by interconnected data nodes. In order to achieve high throughput with minimum delay the QoS in real-time data communication have to be concentrated by using Viterbi decoder with convolution codes. By undertaking IEEE 802.11 WLAN physical layers afford multiple transmission rates by engaging various modulations and channel coding schemes, major point arises to pinpoint the desired transmission rate to enhance the performance. Because each node exhibits different dynamic characteristics based on the token rings passed from the server to the end links. In order to validate the real-time traffic with power consumption and average delay communication, an improved Viterbi decoder is designed with convolution codes to determine accurate channel estimation based on learning the utilization ration of the needed to execute the current wireless channel optimization. The proposed methodology can attain accurate channel estimation without additional implementation effort and modifications to the current 802.11 standard. And each node is capable to choose the optimized transmission rate, so that the system performance can be improved with very minimum power with high packet transmission ratio with minimum traffic rate to improve QoS. The proposed scheme also offers an appealing combination of the allocation of transmission rate and the current link condition. Based on the basic relationship between them, the proposed decoding scheme maximizes the throughput with periodic learning of channel variation and system status.

     

Analysis of a finite-buffer polling system with exhaustive servicebased on virtual buffering

This paper presents an exact analysis of a general asymmetric polling system which has a finite-capacity buffer and exhaustive service discipline at each station. To analyze the system we consider a model system in which each station is assumed to have a temporary virtual buffer with infinite capacity attached to the real buffer, while the server is unavailable. It is introduced to store all the messages which are to be lost due to the limitation of the real buffer capacity during the vacation period. At the instant when the server visits a station, using the marginal queue length distribution of the station including the virtual buffer, the intervisit time distribution of the station is derived, The system behavior using the relationship between performance measures and the intervisit time distribution is then investigated     

保证响应时间百分比下的云中心资源优化配置

为了保证响应时间百分比SLA的同时,最小化商用云中心资源配置成本,基于排队理论提出了云中心模型,从经济角度分析云中心资源配置成本,并推导出云中心响应时间百分比的数学表达式。给出了计算最小配置成本算法（MPCA）,通过大量实验分析了负载流量、响应时间百分比和基准响应时间对配置成本的影响。实验结果表明,通过MPCA对云中心服务器数量和运行速率进行优化,可以使配置成本随着负载流量的增加而呈线性增长。     

Queueing model based resource optimization for multimedia cloud

Multimedia cloud is a specific cloud computing paradigm, focusing on how cloud can effectively support multimedia services. For multimedia service providers (MSP), there are two fundamental concerns: the quality of service (QoS) and the resource cost. In this paper, we investigate these two fundamental concerns with queueing theory and optimization methods. We introduce a queueing model to characterize the service process in multimedia cloud. Based on the proposed queueing model, we study resource allocation problems in three different scenarios: single-service scenario, multi-service scenario, and priority-service scenario. In each scenario, we formulate and solve the response time minimization problem and the resource cost minimization problem, respectively. We conduct extensive simulations with practical parameters of Windows Azure. Simulation results demonstrate that the proposed resource allocation schemes can optimally allocate cloud resources for each service to achieve the minimal response time under a certain budget or guarantee the QoS provisioning at the minimal resource cost.     

Dynamic inter-domain MDS approach with secure seamless handover based on IEEE 802.21 MIH

The accessibility of available wireless access technologies with increasing demand for real time multimedia application becomes an essential part for mobile communication. Mobile users resourcefully utilize the heterogeneous environment for best quality of service (Qos) anywhere and anytime. Efficient handover optimization and intelligent mobility management is a key requirement for designing next generation wireless networks. Therefore, a novel IEEE 802.21 media independent handover (MIH) standard is adopted to provide an associated service for intelligent handover procedures. In addition, dynamic mobility management decision server (MDS) and IEEE 802.21a security extension for MIH services are also integrated in the proposed architectures to support fast, seamless and secure handover optimization in inter-domain mobility. Simulation results prove that the presented work resourcefully minimizes the packet loss, unnecessary handover probability and vertical handover delay by avoiding time consuming scanning process for target network discovery. The system thus achieves Qos guarantee by balancing the network load and throughput improvement for different applications with Proxy MIPv6 mobility management protocol.     

Application-layer anycasting: a server selection architecture anduse in a replicated Web service

Server replication improves the ability of a service to handle a large number of clients. One of the important factors in the efficient utilization of replicated servers is the ability to direct client requests to the “best” server, according to some optimality criteria. In the anycasting communication paradigm, a sender communicates with a receiver chosen from an anycast group of equivalent receivers. As such, anycasting is well suited to the problem of directing clients to replicated servers. This paper examines the definition and support of the anycasting paradigm at the application-layer, providing a service that uses an anycast resolver to map an anycast domain name and a selection criteria into an IP address. By realizing anycasting in the application-layer, we achieve flexibility in the optimization criteria and ease the deployment of the service. As a case study, we examine the performance of our system for a key service: replicated Web servers. To this end, we develop an approach for estimating the response time that a client will experience when accessing given servers. Such information is maintained in the anycast resolver that clients query to obtain the identity of the server with the best estimated response time. Our performance collection technique combines server push with resolver probes to estimate the expected response time without undue overhead. Our experiments show that selecting a server using our architecture and estimation technique can improve the client response time by a factor of two over nearest server selection and by a factor of four over random server selection     

Performance enhancement of adaptive orthogonal modulation in wireless CDMA systems

Recent research in wireless code-division multiple-access systems has shown that adaptive rate/power control can considerably increase network throughput relative to systems that use only power or rate control. In this paper, we consider joint power/rate optimization in the context of orthogonal modulation (OM) and investigate the additional performance gains achieved through adaptation of the OM order. We show that such adaptation can significantly increase network throughput, while simultaneously reducing the per-bit energy consumption relative to fixed-order modulation systems. The optimization is carried out under two different objective functions: minimizing the maximum service time and maximizing the sum of user rates. For the first objective function, we prove that the optimization problem can be formulated as a generalized geometric program (GGP). We then show how this GGP can be transformed into a nonlinear convex program, which can be solved optimally and efficiently. For the second objective function, we obtain a lower bound on the performance gain of adaptive OM (AOM) over fixed-modulation systems. Numerical results indicate that relative to an optimal joint rate/power control fixed-order modulation scheme, the proposed AOM scheme achieves significant throughput and energy gains.     

Analysis of Erlangian network services

This paper presents two equivalent analytical models to study the performance of finite queuing systems with Erlangian network services. Erlangian services are commonly seen in the processing of received network packets by many network servers. Our models allow us to derive equations for key features and performance measures of engineering and design significance. These features and measures include throughput, packet loss, packet delay, and server CPU availability. Numerical examples are given to study the performance while varying the number of service stages and the size of finite buffer. Discrete-event simulation has been used to verify the proposed analytical models.     

An optimization framework for balancing throughput and fairness in wireless networks with QoS support

Quality-of-service (QoS) provisioning, high system throughput, and fairness assurance are indispensable for heterogeneous traffic in future wireless broadband networks. With limited radio resources, increasing system throughput and maintaining fairness are conflicting performance metrics, leading to a natural tradeoff between these two measures. Balancing system throughput and fairness is desired. In this paper, we consider an interference-limited wireless network, and derive a generic optimization framework to obtain an optimal relationship of system throughput and fairness with QoS support and efficient resource utilization, by introducing the bargaining floor. From the relationship curve, different degrees of performance tradeoff between throughput and fairness can be obtained by choosing different bargaining floors. In addition, our framework facilitates call admission control to effectively guarantee QoS of. multimedia traffic. The solutions of resource allocation obtained from the optimization framework achieve the pareto optimality, demonstrating efficient use of network resources.     

Response Time in Data Broadcast Systems: Mean,Variance and Tradeoff

Data broadcast has been suggested as a promising method of information dissemination [2,33]. In such an environment, the information server cannot afford to serve the requests from a large population of users individually. Instead, the server uses a broadcast channel to deliver information to all users. A single transmission of a data item satisfies all pending requests for that item. The response time of a request depends on the broadcast time of the desired data item, which is scheduled by the server according to the overall demands for various data items. Therefore, the response time may vary in a large range. We argue that, in addition to mean response time, the variance of response time should also be taken into account by the broadcast scheduler. In this paper, we address the issue of variance optimization in regard to response time. Building on our previous research on mean response time optimization, we propose an algorithm which can minimize the variance of response time. Furthermore, we evaluate an algorithm that facilitates a tradeoff between the mean and variance of response time. Numerical examples that illustrate the performance of our algorithms are also presented.     

Optimization‐based workload distribution in geographically distributed data centers: A survey

Energy efficiency is a contemporary and challenging issue in geographically distributed data centers. These data centers consume significantly high energy and cast a negative impact on the energy resources and environment. To minimize the energy cost and the environmental impacts, Internet service providers use different approaches such as geographical load balancing (GLB). GLB refers to the placement of data centers in diverse geolocations to exploit variations in electricity prices with the objective to minimize the total energy cost. GLB helps to minimize the overall energy cost, achieve quality of service, and maximize resource utilization in geo‐distributed data centers by employing optimal workload distribution and resource utilization in the real time. In this paper, we summarize various optimization‐based workload distribution strategies and optimization techniques proposed in recent research works based on commonly used optimization factors such as workload type, load balancer, availability of renewable energy, energy storage, and data center server specification in geographically distributed data centers. The survey presents a systemized and a novel taxonomy of workload distribution in data centers. Moreover, we also debate various challenges and open research issues along with their possible solutions.     

Link Adaptation Using Dynamically Allocated Thresholds and Power Control

Link adaptation technique, in which the modulation and coding used in a communication system is changed as per the channel conditions is a very well investigated topic for link throughput maximization with widespread application in wireless access networks. Most of the known algorithms dynamically adjust transmitter data rate by comparing instantaneous SNR with pre-defined SNR switching thresholds, in order to maximize throughput while maintaining the desired quality of service. However, the use of incorrect or stale values of these pre-defined switching thresholds often leads to selection of erroneous modulation and coding schemes resulting in unsatisfactory throughput or quality of service. This work introduces a novel scheme which achieves the maximum possible throughput while maintaining the target quality of service by dynamically acquiring the threshold values of different modulation and coding schemes used in the system for a given value of block error rate based on measurement at the receiver. This helps in keeping the threshold look up table up to date, so that proper threshold values for mode switching is present for all channel conditions. Also, a relationship between the throughput and the accuracy of the threshold value calculation is provided so that these can be optimized depending on the user requirements. The performance evaluation shows that the proposed system outperforms the conventional link adaptation in various operating scenarios where pre-determined look up tables are not available.