Cost Load Balancing Research Articles

An Improved Particle Swarm Algorithm for Multi-Objectives Based Optimization in MPLS/GMPLS Networks

Particle swarm optimization (PSO) is a swarm-based optimization technique capable of solving different categories of optimization problems. Nevertheless, PSO has a serious exploration issue that makes it a difficult choice for multi-objectives constrained optimization problems (MCOP). At the same time, Multi-Protocol Label Switched (MPLS) and its extended version Generalized MPLS, has become an emerging network technology for modern and diverse applications. Therefore, as per MPLS and Generalized MPLS MCOP needs, it is important to find the Pareto based optimal solutions that guarantee the optimal resource utilization without compromising the quality of services (QoS) within the networks. The paper proposes a novel version of PSO, which includes a modified version of the Elitist learning Strategy (ELS) in PSO that not only solves the existing exploration problem in PSO, but also produces optimal solutions with efficient convergence rates for different MPLS/ GMPLS network scales. The proposed approach has also been applied with two objective functions; the resource provisioning and the traffic load balancing costs. Our simulations and comparative study showed improved results of the proposed algorithm over the well-known optimization algorithms such as standard PSO, Adaptive PSO, Bat and Dolphin algorithm.

Stochastic simulator for priority based task in grid environment

Among the different types of distributed computing, grid computing is that type of computing which includes huge collection of resources and applications, which may be at the same or different locations owned by same or different governments. Availability of right resource at right time with suitable task is managed by scheduling process. Task scheduling goal in grid computing is to attain better throughput and to match the requirement of application with the available set of resources. With the increase in the size of task and grid, complexity of scheduling problem increases and it becomes a NP-complete problem to be solved. This has given birth to a new research issue. In this paper authors have given a solution to the scheduling problem by proposing a probability based scheduling approach. To represent the scheduling scenario, authors have used the concept of directed acyclic graph. Proposed approach not only handles the tasks on the basis of their priority but also handle them in parallel. Gridsim simulator has been used to compare and test the proposed approach with other existing approaches found in the literature. Parameters like makespan, cost, through put and percentage of load balancing has been considered to check the performance of probabilistic scheduling approach.

Quantum genetic algorithm based scheduler for batch of precedence constrained jobs on heterogeneous computing systems

Abstract Distributed systems are efficient means of realizing High-Performance Computing (HPC). They are used in meeting the demand of executing large-scale high-performance computational jobs. Scheduling the tasks on such computational resources is one of the prime concerns in the heterogeneous distributed systems. Scheduling requires optimizing either single objective or multiple objectives. Load Imbalance (LIB) and Load balancing Cost Ratio (LCR) are two such objectives. LIB and LCR are used in deciding the distribution of load over available resources for utilization and the computational cost respectively. Scheduling jobs with precedence constraints are NP-complete in nature. Scheduling requires either heuristic or metaheuristic approach for sub-optimal but acceptable solutions. Quantum computing is one such metaheuristic approach. It has proven to be promising in addressing the multi-objective scheduling problems with an effective exploration of the search space. This work proposes a dual-objective Quantum-inspired Genetic Algorithm based Load Balancing Strategy (QGLBS) for workflow application with the objective of optimizing both LIB and LCR. The DAG representation of the batch of jobs helps in effective exploitation of the parallelism available at the job level as well as the sub-job level with the modules at the same level of the batch executing at the same time. The strategy ensures that the communication cost and the critical path are considered in making scheduling decisions. QGLBS exploits the features of Quantum Computing and Genetic Algorithm to establish the Pareto fronts using non-dominated sorting based on NSGA-II. Hypervolume measures are used to compare the results of the quality assessment of Pareto fronts. Simulation study on Gridsim ensures that the QGLBS approach works effectively for real life scenario with various workflow applications.

Frequent Itemsets Mining for Big Data: A Comparative Analysis

Abstract Itemset mining is a well-known exploratory data mining technique used to discover interesting correlations hidden in a data collection. Since it supports different targeted analyses, it is profitably exploited in a wide range of different domains, ranging from network traffic data to medical records. With the increasing amount of generated data, different scalable algorithms have been developed, exploiting the advantages of distributed computing frameworks, such as Apache Hadoop and Spark. This paper reviews Hadoop- and Spark-based scalable algorithms addressing the frequent itemset mining problem in the Big Data domain through both theoretical and experimental comparative analyses. Since the itemset mining task is computationally expensive, its distribution and parallelization strategies heavily affect memory usage, load balancing, and communication costs. A detailed discussion of the algorithmic choices of the distributed methods for frequent itemset mining is followed by an experimental analysis comparing the performance of state-of-the-art distributed implementations on both synthetic and real datasets. The strengths and weaknesses of the algorithms are thoroughly discussed with respect to the dataset features (e.g., data distribution, average transaction length, number of records), and specific parameter settings. Finally, based on theoretical and experimental analyses, open research directions for the parallelization of the itemset mining problem are presented.

Near-Optimal Routing Protection for In-Band Software-Defined Heterogeneous Networks

Facing the spectrum supply-demand gap, heterogeneous network (HetNet) is a promising approach to achieve drastic gains in network coverage and capacity compared with macro-only networks, thus making it especially attractive to network operators. On the other hand, software-defined networking brings a number of advantages along with many challenges. One particular concern is on the resilience for in-band fashioned control plane. Existing approaches mainly rely on a local rerouting policy when performing the routing protection for the target sessions in software-defined networks. However, such a policy would potentially bring congestions in the neighbouring links of the failed one. To this end, we study a weighted cost-minimization problem, where the traffic load balancing and control-channel setup cost are jointly considered. Because this problem is NP-hard, we first propose a near-optimal Markov approximation-based approach for in-band-fashioned software-defined HetNets. We then extend our solution to an online case that handles a single-link failure. We also conduct theoretical analysis on the performance fluctuation due to the single-link failure. We finally carry out experiments by experimental simulation. The extensive numerical results show that the proposed algorithm has fast convergence and high efficiency in resource utilization.

ANALISIS PENGARUH KONFIGURASI EIGRP EQUAL DAN UNEQUAL COST LOAD BALANCING TERHADAP KINERJA ROUTER

Routing protocol is tasked with finding the best route to send the packet. Assessed using the metric. If there is more than one route with the same metric value, Routing Information Path (RIP), Open Shortest Path First (OSPF), and Enhanched Interior Gateway Routing Protocol (EIGRP) support equal cost load balancing to send packets to the destination. If there is more than one route with a different metric values, EIGRP can do unequal cost load balancing. Research needs to be conducted to determine the effect of the configuration of EIGRP equal and unequal cost load balancing on the performance of the router which can be used as a proof-of-concept testing that is part of the project design document on a network. Research networks using EIGRP as the routing protocol. After the equal and unequal load balancing is enabled by configuring the variance, CEF, per-destination load balancing, per-packet load balancing, or traffic sharing and analyzing its effect on the neighbor table, topology table, routing table, the data transmission, survivability, convergence, throughput, and utilization. This study used an emulator GNS3 as Cisco 2691 Router with Cisco IOS version 12:24 (25) c and advanced enterprise-adventerprisek9 image c2691-mz.124-25c.bin, and OPNET Modeler 14.5 for simulation. The results of the study can be used as a proof-of-concept testing in the design document for later use as contemplated in the manufacture of plan implementation and verification plan.

Cost and Time Evaluation of Load Balancing and Service Broker Strategies in Multiple Data Centers

Cloud computing is a boon for almost every sector as it reduces the overall expenditure by making I.T. infrastructure, services and platform on “on-demand” basis. The “pay as you go” policy of cloud technology enables the users to use and pay for the services and computing infrastructure only when they need it without worrying about the installation and underlying hardware or software infrastructure. Data centers are located at different locations and provide services to the users in their closest proximity. As the user has to pay for the service utilization, so selection of data center plays an important role for maximum resource utilization at minimum expenditure and best response time. An important feature in cloud is that according to user location and demand, the closest and the cheapest data center is assigned in that region. In this paper, the authors have analyzed the existing service broker policies i.e., Closest Data center, optimize response time, and reconfigure dynamically load, in data center selection.

Optimal Load Balancing and Energy Cost Management for Internet Data Centers in Deregulated Electricity Markets

Along with the rapid increasing energy consumption, the energy cost of Internet data centers (IDCs) has been skyrocketing. A novel scheme of geographical load balancing was proposed to reduce electricity bills for service providers. However, one important challenge faced by service providers has not been considered properly. In service systems, the service delay faced by consumers includes the queuing delay and the transmission delay. While existing work only consider the queuing delay, the transmission delay introduced by geographical load balancing has been overlooked. It is one of the most important factors affecting the quality of service for real-time service systems. In this paper, we take the transmission delay into our design consideration and formulate a mixed-integer nonlinear programming (MINLP) problem with coupled constraint to achieve the optimal load balancing and energy cost management for IDCs while meeting the service-level agreements (SLA) of consumers. A novel heuristic based branch and bound with feedback (HBBF) algorithm is proposed to decouple the MINLP problem with coupled constraint efficiently. Extensive performance evaluations based on real electricity price data and site-to-site transmission delay data demonstrate the effectiveness of our proposed algorithm.

Quality Engineering of Distributed System with Hassle Free Load Curve Optimization

Distributed system presents the possible approaches for distribution and aggregation of diverse set of resources such as computers, storage space systems and other specific devices. Normally, these set of resources are processed by the different set of users in different organizations. So, the users might contain diverse objectives and plans and their activities is very complex to differentiate. And also when the demand over the system increases, the load balancing problem arises and the traffic over the systems increases. So, it is necessary to improve the load balancing in the distributed system by means of distributing the application load with software engineering traffic principles. Jun Zhang et.al., presented a novel nonparametric technique for traffic classification which can develop the performance of the classification efficiently by integrating associated information into the classification method. However the traffic classification using non-parametric and non engineering approach fails to balance the load effectively in distributed system. So, in this work, we plan to improve the quality of load balancing in the distributed system by adapting Hassle-free Curved Optimization (HCO) traffic engineering principle. The HCO traffic engineering principle is presented with this research for effective load balancing and communication cost in distributed system. Heuristic algorithm is followed in HCO principle with given communication pattern to produces an optimal solution for effective distributed data/service communication. The experimental performance of HCO traffic engineering principle is evaluated with Solar Flare Data Set from UCI repository against existing non-parametric approach for effective load balancing scheme to attain quality assurance, effective load balancing with minimal communication cost.

Building Structured Peer-to-Peer Resource Sharing Platform Using Object Encapsulation Approach

This paper designed and built a distributed hash table based computing resource sharing platform named OE-P2RSP. By employing Peer-to-Peer services, OE-P2RSP allowed users to submit jobs to be run in the system and to run jobs submitted by other users on any resources available over the Internet, essentially allowing a group of users to form an Ad hoc set of shared resources. OE-P2RSP is based on structured peer on network share, improves the communication mode of system by encapsulated object, object group, and physical proximity principle, and the use of physical proximity principle, and gathers computing resources better in the same research institution or enterprise within local area network. The experimental results obtained via simulations show that the system can reliably execute scientific applications on a widely distributed set of resources with good load balancing and low matchmaking cost and that OE-P2RSP has good efficiency, load balancing and scalability. DOI: http://dx.doi.org/10.11591/telkomnika.v11i2.2056

Replica-aided load balancing in overlay networks

In recent years, there have been rapid advances in network infrastructure and technologies for end-user communication. However, because of network dynamics and resource limitation, providing scalable end-user communication services is challenging when the applications are utilized on a large-scale. To address this challenge, a replica-aided load balancing scheme (RALB) is proposed for enabling the nodes in an overlay networks to support the communication applications for a large number of users. This paper makes three unique contributions. First, we study the existing load balancing schemes and identify their weakness in handling time-varying workloads with frequent load fluctuations. Second, we introduce a sophisticated cost model for load balancing cost estimation, which captures the dependencies between the factors (e.g., the load, message number, and link latency). Third, we propose a performance tuning technique to minimize the load balancing cost. The extensive experiments show that RALB effectively reduces the load imbalance and eliminates the load balancing cost when compared to the existing load balancing schemes.

A bi-criteria minimum spanning tree routing model for MPLS/overlay networks

The MPLS platform enables the implementation of advanced multipath and multicast routing schemes. This work develops and analyses the performance of a new bi-criteria minimum spanning tree model intended for routing broadcast messages in MPLS networks or constructing tree-based overlay networks. The aim of the model is to obtain spanning trees which are compromise solutions with respect to two important traffic engineering metrics: load balancing cost and average delay bound. An exact solution to the formulated bi-criteria optimization problem is presented, which is based on an algorithm that enables the computation of the set of supported non-dominated spanning trees. An application model and a set of experiments on randomly generated Internet type topologies will also be presented. Finally a network performance analysis of the model considering three network performance metrics will be shown.

On a moving direction pattern based MAP selection model for HMIPv6 networks

In a large-scale mobile IPv6 network, usually there are several coexisting mobility anchor points (MAPs) for networking robustness and traffic sharing. Therefore, it is a challenging issue for an arriving mobile node to choose the most appropriate MAP to bind. This task must be carried out by considering the issues of load balancing, binding update and packet delivery cost minimization. This paper proposes a novel MAP selection scheme for hierarchical mobile IPv6 networks to allow a mobile node to discover the most appropriate MAP when there are multiple coexisting MAPs. This scheme is an enhancement to the adaptive MAP selection scheme. The proposed scheme improve the overall performance due to the consideration of the movement (direction) pattern of mobile nodes. Simulation results show that this scheme outperforms the existing cost models in terms of total binding update and packet delivery costs, ensuring a level of load balance similar to adaptive MAP selection scheme.

Some Compartmentalized Secure Task Assignment Models for Distributed Systems

This paper formulates abstract problems of assigning subtasks to agents (processors) in a distributed system with a goal that they can perform its global task efficiently. The paper models the distributed system with a graph that describes the communication capabilities of the constituting agents. This graph is referred to as the "organizational graph.” In addition, the desired task-performing activity is modeled with another graph describing the required communications. Then, a few variants of the task assignment problem are formulated with potentially conflicting objectives (or constraints) of load balancing and communication costs. For some of these variants this paper provides efficient algorithms that solve the assignment problem. Some problems are proven NP-complete, and some others are left open.

Performance evaluation of a dynamic load-balancing library for cluster computing

The performance of scientific applications in heterogeneous environments has been improved with the research advances in dynamic scheduling at application and runtime system levels. This paper presents the performance evaluation of a library as a result of an integrated approach to dynamic load balancing. This approach combines the advantages of optimising data migration via novel dynamic loop-scheduling strategies with the advances in resource management and task migration capabilities offered by a recently developed parallel runtime system. The performance of the library has been investigated by its use in three scientific applications: the N-body simulations, the profiling of automatic quadrature routines and the solution to the 3D heat equation. The investigations focus on the performance degradation owing to the overhead induced by the runtime system software layer. The experimental results obtained indicate only a slight increase in the cost of load balancing owing to this overhead. The results validate the suitability of the runtime system as an implementation platform for dynamic load-balancing schemes and underscore the significance of using the integrated approach, as well as the benefits of using the library especially in cluster applications characterised by irregular and unpredictable behaviour.

Distributed Processing System by History for Load Balancing

This proposal system is distributed processing system by History for load balancing. This system is considered the load balancing and communication cost. History is a database of load states average to time of all hosts, and learned with Self-Organization Feature Map. Capsuled objects of class described in object oriented programming language are distributed. Remote Class Controller (RCC) and Remote Class (RC) are generated from the source class. RCs are sent to the servers suggested by History, and, by linking RCCs. a main program can commu­nicate with RCs. As a result, the load was balanced and the communication cost in executing became low.

Optimizer-Assisted Load Balancing Techniques for Multicomputer Database Management Systems

Although load balancing incurs processing costs, and therefore can have a profound influence on the optimized execution plan of a query, none of the existing parallelizing query optimizers consider this factor. In this paper, we address this issue by introducing the cost of load balancing as a new factor for query optimization. Specifically, we implemented three new optimizers for multiway join queries that take the load balancing issue into consideration. To evaluate the efficiency of these schemes, we also implemented a simulator for the parallel execution of multiway joins. To provide more faith, our simulation model was validated by comparing the simulation results to those produced by the actual implementation of the same algorithms running on a multicomputer system. This simulator was used in our study to compare the new techniques to a more conventional system in which load balancing is performed at runtime, but it is not a factor for query optimization. Our extensive simulation results confirm that the new methods, indeed, provide very significant savings. Most interestingly, the best scheme displays a performance which is essentially immune from the skew effect. Furthermore, we observed that these new optimizers can consistently achieve the same level of performance gain regardless of the CPU power, I/O, and communication capabilities of the computing system. This indicates that our approaches are generally useful for all hardware platforms.

A Uniform Framework for Dynamic Load Balancing Strategies in Distributed Processing Systems

Load balancing plays a central role in processor utilizations in distributed systems. Several strategies have been proposed in the literature to achieve load balancing. Usually, these strategies attempt to achieve a tradeoff between reducing the execution time of an application and minimizing the synchronization and the communication overhead. In this paper, we present a general model in which load balancing decisions are reached by enforcing performance metrics which may be adapted to reflect the specific requirements of different environments. Many of the load balancing schemes that have been suggested in the literature can be viewed as specific instances of the general framework presented in this paper. The basic scheme in this framework uses a load contention number that accounts for the load of the processors, the communication cost and the distance among processors. It is meant to be adaptable to the overall load on the system, the load on the communication devices, the run time characteristics of the tasks, and the configuration of the system. Furthermore, its implementation is not computationally complex. Thus, the gains made by load balancing are not overshadowed by the load balancing cost.