Distributed Load Balancing Research Articles

Dynamic Load Balancing for Congestion Avoidance using Adaptive Neuro-Fuzzy Inference System in Mobile Communication Network

Network congestion is one of the key challenges to mobile communication services. This is because of rapid and constant growth in the mobile subscriber base that has led to an increase in network traffic. The more the network traffic the more economic opportunity from the business perspective for the operators and the challenges it poses on the network. If not followed by network capacity expansion it will defiantly pose a serious network issue like network congestion that leads to call drop and poor Quality of Service (QoS). Many research works adopted handover mechanism as a means of reducing congestion in a network. However, the decision on when to initiate a handover, which cell to receive the Mobile Station (MS) and how to ensure that the QoS requirements are maintained are the paramount research questions that must be resolved. Therefore, a handover process using soft computing in the Adaptive Neuro-Fuzzy Inference System (ANFIS) to ensure balanced traffic load distribution in-network and reduce the probability of congestion is proposed. The study allows simultaneous evaluation of three major network parameters Received Signal Strength (RSS), Received Signal Quality (RxQual) and network traffic using ANFIS to improve system performance. The results show that when the Hysteresis value approaches 6 the handover processes are triggered. The hysteresis value of the concerned MS with neighbouring cells is considered to determine the most suitable cell to handover. This work will be able to achieve dynamic load balancing, congestion avoidance and avoided the ‘ping pong’ effect that is often an issue with handover with less computation. At the end customer satisfaction will be achieve Quality of experience (QoE).

Impairment characteristics of static balance and plantar load distribution of patients undergoing tibial cortex transverse distraction for diabetic foot ulcers

ObjectiveTibial cortex transverse distraction (TCTD) has been recently reported for the treatment of diabetic foot ulcers. Herein, we explored the characteristics of the impairments in static balance and plantar load distribution in patients.MethodsWe performed a retrospective study of 21 patients with diabetic foot ulcers who underwent TCTD, who were regularly followed up for > 1 year after surgery, and 20 healthy individuals (control group). A pressure platform was used to assess the standing balance functions of the lower extremities and the plantar load distribution.ResultsOne patient underwent amputation because of severe infection. In patient group, center of pressure (COP) ellipse sway area, COP path length and angle θ were all larger, compared with those of control group (250.15 ± 98.36 mm2 vs. 135.67 ± 53.21 mm2, 145.15 ± 67.43 mm vs. 78.47 ± 34.15 mm, 39.75 ± 17.61° vs. 22.17 ± 14.15°), with statistically significant differences (P < 0.01). The average plantar load and backfoot load of the unaffected side was significantly larger than that of the affected side (58.4 ± 5.5% vs. 41.6 ± 5.5%, 45.3 ± 6.4% vs. 36.5 ± 5.6%), but they were similar for the two feet of members of the control group.ConclusionsAlthough TCTD may represent an appropriate method for the treatment of diabetic foot ulcers, postoperative impairments in static balance and plantar load distribution remain in the long term. These potential long-term problems should be taken into account in further rehabilitation planning.Type of study/level of evidence: Therapeutic III.

A Fair, Dynamic Load Balanced Task Distribution Strategy for Heterogeneous Cloud Platforms Based on Markov Process Modeling

Load balancing techniques in cloud computing can be applied at three different levels: Virtual machine load balancing, task load balancing, and resource load balancing. At all levels, load balancing should also be implemented in an efficient manner, to increase system performance. In this paper, we propose a fair, in terms of added workload per VM, task load balancing strategy, that aims to improve the average response time and the makespan of the system in the cloud environment. The problem is formulated as an irreducible finite state Markov process, which is known to have a balance equation for each state. From the balance state probabilities we derive the expected utilizations for the virtual machines (VM), which play a vital role in our task allocation approach. In our model, the Load Balancer (LBer) acts as a central server, which uses our proposed fair task allocation scheme to distribute the incoming tasks in a fair, balanced manner among the virtual machines, taking into account their current state as well as their processing capabilities. Our scheme has been compared to recent algorithms that use the particle swarm optimization and the Honey bee foraging scheme to achieve load balancing. Our experimental results show that our proposed scheme outperforms other state of the art schemes in terms of makespan, average response time, and resource utilization and provides lower degree of imbalance.

Comparative Analysis of Selected Open-Source Solutions for Traffic Balancing in Server Infrastructures Providing WWW Service

As the number of users increased over the years, pioneering technologies and solutions in given areas ceased to be sufficient even in terms of performance. Therefore, there was a need for their development or even redesign and redefinition. One of the issues that undoubtedly has a huge impact on the current shape of the global network and the way information is processed in it is the issue of traffic balancing, especially the one in the server infrastructure related to the WWW service, providing users with the possibility of efficient and reliable web browsing. The paper presents a comparative analysis of selected open-source solutions used for traffic balancing in server infrastructures providing WWW service based on selected criteria. The designed architecture of the test environment and the test results of selected tools implementing the traffic-balancing functionality are presented. Methodologies, test plans, and comparison criteria are proposed. A comparative analysis of results based on specific criteria was performed. The balance between network traffic optimization and load balancing distribution among servers is crucial for the development of energy-efficient data processing centers.

Impairments and Compensations of Static Balance and Plantar Load Distribution in Patients With Chronic Stroke: An Observational Study

ObjectivesThe purpose of this study was to explore impairment and compensation characteristics of static balance and plantar load distribution in patients with chronic stroke. MethodsWe recruited 68 patients who had chronic stroke who could stand independently (stroke group) and 30 healthy individuals (control group) with the Zebris FDM platform. Static balance parameters, including center of pressure (COP) ellipse sway area, COP path length, and angle-θ between y and major axis, were compared between 2 groups under standard standing posture. In the stroke group, balance parameters were re-tested under their preferring standing posture. Plantar load distribution was also assessed. Another 8 patients with chronic stroke who could not stand independently and had to rely on a crutch were enrolled to analyze the characteristics of balance compensation. ResultsIn the stroke group, the ellipse sway area, COP path length, and angle-θ were significantly larger than those of the control group. Sixty-one (89.7%) patients preferred standing with the affected foot outward-forward supporting, and their preferring standing balance was better than that of standard standing. All patients who could not stand independently tended to compensate for balance with a crutch supporting laterally and also preferred standing with a typical posture- the affected foot was outward-forward. ConclusionsIn patients post stroke, static balance is impaired mainly at the lateral direction, and patients commonly locate the crutch laterally for compensation of lateral balance. Patients preferred standing with the affected foot outward-forward supporting, and their preferring standing balance was better than that of standard standing, which challenged the necessity of training standing symmetrically.

Proprioceptive Neuromuscular Facilitation Improves Symptoms Among Older Adults With Knee Osteoarthritis During Stair Ascending: A Randomized Controlled Trial.

This study aimed to investigate the effects of a 6-wk proprioceptive neuromuscular facilitation stretching on pain, proprioception, joint range of motion, and joint moments during stair ascending among older adults with knee osteoarthritis. This study is a randomized, controlled, and assessor-blinded trial. Thirty-six older adults with knee osteoarthritis were randomly assigned to the proprioceptive neuromuscular facilitation and the control groups. They received proprioceptive neuromuscular facilitation stretching and health lecture series, respectively, for 6 wks. Final data analysis included 14 participants of the proprioceptive neuromuscular facilitation group and 13 of the control group. Pain score, joint proprioception, range of motion, and joint moments during stair ascending were measured before and after the stretching. Two-way (group by time) analysis of variance with repeated measures was used to evaluate stretching effects. Significant interactions were detected in pain score, joint proprioception, external knee adduction moment, and external knee extension moment. Compared with week 0, the pain score, joint proprioception threshold, and external knee adduction moment decreased, whereas the external knee extension moment increased among older adults in the proprioceptive neuromuscular facilitation group at week 7. Proprioceptive neuromuscular facilitation could be recommended as one of the clinical treatments for knee osteoarthritis to relieve pain, improve proprioception, and balance load distribution between medial and lateral compartments at the knee.

Research on Remote Upgrade Method of Distributed Load Balancing for Power Monitoring Terminal

With the increasing application of power monitoring terminal and the increasing demand for its function, it is necessary to upgrade the field embedded software remotely.The existing online upgrade schemes generally deal with the terminals one by one, and there are also distributed upgrade schemes. However, due to the large differences in the communication performance of terminals in different places, there is a problem of load imbalance, resulting in the low efficiency of the whole upgrade. Aiming at the efficiency of remote online upgrade of power monitoring terminal, a distributed load balancing remote upgrade method is proposed. The power monitoring terminal is upgraded at the same time through the cloud platform, and the genetic algorithm is used for load balancing processing, so as to improve the efficiency. The practical results show that this remote upgrade method improves the upgrade efficiency of the terminal by comparing a variety of upgrade schemes, compared with no load balancing method, the average time of upgrading a terminal is reduced by 36%, which achieves good application effect in practical application.

Fault‐tolerant strategies for open‐circuit switch failures in open‐end winding PMSM drives

Many safety-critical drive systems require post-fault operation after an open-circuit switch failure occurs in the inverter. The open-end winding inverter topology facilitates the injection of zero-sequence current such that a faulty switch leg can still be used to generate sufficient voltage states to continue operation. Therefore, only a minimal increase in the post-fault torque ripple is experienced. However, the injection of this zero-sequence current results in an increase in the copper loss which de-rates the maximum post-fault torque of the machine. Existing post-fault techniques that inject zero-sequence current, during an open-circuit switch fault, do not minimize the copper loss resulting in sub-optimal post-fault torque per ampere. A method for reducing copper loss by optimizing the zero-sequence current waveform to achieve the maximum post-fault torque per ampere is proposed here. The proposed strategy has been experimentally assessed for both three-phase and five-phase drive systems. The experimental results demonstrate the continuous post-fault operation of the drive system with a more balanced thermal load distribution between phases. A 2.8% and 5.6% improvement in post-fault torque per ampere for three-phase and five-phase systems are demonstrated.

Novel distributed load balancing algorithms in cloud storage

From last decade, there is a rapid expanded upon in the data in cyberspace. In order to manage them efficiently, distributed storage came into the world. Cloud storage is a type of distributed storage based on cloud computing technology. Cloud storage acts as a repository in which data stored, managed and made accessible to users. Largest generated application datasets can flexibly be stored or deleted in the cloud and from here end users access this data by using cloud storage services interface, without accessing any storage server in real. Cloud storage system is considered of hundreds of independent storage servers which are distributed geographically, and handle millions client requests concurrently. Some of the storage servers get huge clients requests and some servers remain under loaded. Due to this unequal distribution of load on storage servers degrades the performance of overall system and increases the response time. This work addresses issues regarding to efficient utilization of storage servers in cloud storage. Handling various challenges regarding to the load balancing in the cloud storage is the one of the main objectives of this research. Though analyzing the contribution of other authors in this area, in this work two distributed load balancing algorithm CDLB and DDLB are proposed by exploiting the different parameter of storage server. The first proposed algorithm considers the service rate and queue length as a main parameter of the server. The second proposed algorithm considers extra server parameter such as service time and deadline time of the client request. This work monitors various aspects which leverage the overall performance of cloud storage. Both the algorithms try to balance the load of storage servers as well as effectively utilize the server capabilities. From the simulation results, it can be concluded that proposed algorithms balance the load, efficiently utilize the server capabilities, reduce the response time and leverage the overall system performance.

Research on Load Balancing Dispatching Method of Power Network Based on Cloud Computing

With the continuous development of big data era and cloud computing(CC) technology, there are still a series of problems in the planning performance and related algorithms of CC grid, such as unbalanced grid load and uneven grid spatial distribution. However, with the continuous development of social economy and the continuous improvement of people’s living standards and quality, the scale of power grid construction is also expanding, the load distribution of power grid is unbalanced, and the power distribution capacity is insufficient. In order to effectively analyze the load balance and scheduling of power grid, CC technology can accurately improve the accuracy of data analysis and processing. In the process of network load balancing and distribution, the use of CC can not only improve the scheduling efficiency, but also improve the quality and speed of work. Aiming at the problem of poor scheduling performance, this paper proposes a fuzzy iterative algorithm and applies it to CC environment. Plan load balancing for the current network. The experimental process of the algorithm is introduced in detail, the balanced distribution parameters of the load data are extracted, and the load model is established. Realize and improve the network load balance distribution. Experimental results show that the proposed method can improve the load balancing configuration of power grid, improve the management efficiency of power grid, and has good experimental performance.

Functional foot trimming to balance load distribution between the paired forelimb claws in dairy cows: An experimental study

In standing dairy cows, the medial claws carry most of the load of the forelimb. The goal of our experimental study was to investigate how even load distribution can be achieved between the paired forelimb claws through foot trimming. We measured vertical ground reaction forces, mean and maximum pressures, and contact areas under the sound forelimb claws of 28 nonlame dairy cows using a force plate and a pressure plate. The experimental setup consisted of 3 measurement situations: (1) baseline measurements made after routine foot trimming, (2) measurements made with plywood plates, 3 mm in thickness, attached to both lateral claws, and (3) measurements made with the same plywood plates attached to both medial claws. After routine foot trimming, the medial claws carried 70.1% of the entire forelimb load with its palmar zone bearing 54.0% of the entire forelimb load. Even load distribution between the paired claws was achieved when the plywood plates were attached to the lateral claws, whereas plates on the medial claws increased the load discrepancy. The pressures on the medial claws were reduced by attaching the plywood plates to the lateral claws but remained uneven between paired claws and claw zones. The palmar zone of the medial claws was the region subjected to the greatest force and highest pressures in all situations. A drawback of the study was that the plywood plates increased the contact area and therefore reduced the mean and maximum pressures in the claws they were attached to. We concluded that equal load distribution between the forelimb claws was achieved in our setting when the lateral claws were left 3 mm higher than the medial claws.

A new task offloading algorithm in edge computing

In the last few years, the Internet of Things (IOT), as a new disruptive technology, has gradually changed the world. With the prosperous development of the mobile Internet and the rapid growth of the Internet of Things, various new applications continue to emerge, such as mobile payment, face recognition, wearable devices, driverless, VR/AR, etc. Although the computing power of mobile terminals is getting higher and the traditional cloud computing model has higher computing power, it is often accompanied by higher latency and cannot meet the needs of users. In order to reduce user delay to improve user experience, and at the same time reduce network load to a certain extent, edge computing, as an application of IOT, came into being. In view of the new architecture after dating edge computing, this paper focuses on the task offloading in edge computing, from task migration in multi-user scenarios and edge server resource management expansion, and proposes a multi-agent load balancing distribution based on deep reinforcement learning DTOMALB, a distributed task allocation algorithm, can perform a reasonable offload method for this scenario to improve user experience and balance resource utilization. Simulations show that the algorithm has a certain adaptability compared to the traditional algorithm in the scenario of multi-user single cell, and reduces the complexity of the algorithm compared to the centralized algorithm, and reduces the average response delay of the overall user. And balance the load of each edge computing server, improve the robustness and scalability of the system.

Achieving Balanced Load Distribution with Reinforcement Learning-Based Switch Migration in Distributed SDN Controllers

Distributed controllers in software-defined networking (SDN) become a promising approach because of their scalable and reliable deployments in current SDN environments. Since the network traffic varies with time and space, a static mapping between switches and controllers causes uneven load distribution among controllers. Dynamic migration of switches methods can provide a balanced load distribution between SDN controllers. Recently, existing reinforcement learning (RL) methods for dynamic switch migration such as MARVEL are modeling the load balancing of each controller as linear optimization. Even if it is widely used for network flow modeling, this type of linear optimization is not well fitted to the real-world workload of SDN controllers because correlations between resource types are unexpectedly and continuously changed. Consequently, using the linear model for resource utilization makes it difficult to distinguish which resource types are currently overloaded. In addition, this yields a high time cost. In this paper, we propose a reinforcement learning-based switch and controller selection scheme for switch migration, switch-aware reinforcement learning load balancing (SAR-LB). SAR-LB uses the utilization ratio of various resource types in both controllers and switches as the inputs of the neural network. It also considers switches as RL agents to reduce the action space of learning, while it considers all cases of migrations. Our experimental results show that SAR-LB achieved better (close to the even) load distribution among SDN controllers because of the accurate decision-making of switch migration. The proposed scheme achieves better normalized standard deviation among distributed SDN controllers than existing schemes by up to 34%.

A Wireless Network Communication Capacity Control Technology Based on Fuzzy Wavelet Neural Network

The communication capacity control of the computer wireless network is the basis for realizing the efficient communication of massive data. In order to study the communication capacity control technology of the computer wireless network, improve the control effect of a large amount of data communication, and calculate the capacity of the wireless network in real time, this paper uses the fuzzy wavelet neural network to predict the wireless network channel. After the interference‐free channel is obtained, the load balancing strategy of the ant colony optimization algorithm is used to filter the channel, and the channel allocation sequence with the most balanced load distribution is obtained, and a priority selection list is generated. After discretizing the channels in the largest discretization selection list, the channel sequence is allocated to the pair of nodes with communication requests according to the greedy coloring algorithm, so as to realize the communication capacity control of the computer wireless network. The test results show that the technology can guarantee good communication performance in both static and dynamic networks and can effectively complete network communication of massive data, and the communication capacity control effect is good.

Belief Propagation for Energy Efficiency Maximization in Wireless Heterogeneous Networks

In this article, we study an energy efficient management of two-tier heterogeneous cellular networks (HetNets) which consist of one macro base station (BS) and multiple micro base stations. This article presents a distributed user association algorithm that maximizes the network-wide energy efficiency (EE) in HetNets. A subset of BSs that support only a small number of users can be turned off to save the energy consumption. By turning off BSs in the HetNet and offloading serving users to adjacent active BSs, the network-wide energy consumption is minimized, while the sum throughput is maximized. To solve the problem efficiently, we introduce a new approach based on a message-passing framework and derive a distributed load balancing algorithm. The proposed method provides a very efficient solution with reduced computational complexity compared to existing schemes. Simulation results verify that the proposed algorithm outperforms other conventional load balancing strategies.

PackStealLB: A scalable distributed load balancer based on work stealing and workload discretization

The scalability of high-performance, parallel iterative applications is directly affected by how well they use the available computing resources. These applications are subject to load imbalance due to the nature and dynamics of their computations. It is common that high performance systems employ periodic load balancing to tackle this issue. Dynamic load balancing algorithms redistribute the application’s workload using heuristics to circumvent the NP-hard complexity of the problem However, scheduling heuristics must be fast to avoid hindering application performance when distributing the workload on large and distributed environments. In this work, we present a technique for low overhead, high quality scheduling decisions for parallel iterative applications. The technique relies on combined application workload information paired with distributed scheduling algorithms. An initial distributed step among scheduling agents group application tasks in packs of similar load to minimize messages among them. This information is used by our scheduling algorithm, PackStealLB, for its distributed-memory work stealing heuristic. Experimental results showed that PackStealLB is able to improve the performance of a molecular dynamics benchmark by up to 41%, outperforming other scheduling algorithms in most scenarios over almost one thousand cores.

Fairness-Aware Mechanism for Load Balancing in Distributed Systems

When a set of self-interested users shares multiple resources in a distributed system, we face the problem of allocating resources, called the load balancing problem. In particular, load balancing is defined as allocating the load to the servers of the distributed system such that jobs’ response time is minimized, and the utilization of servers is improved. In this article, the load balancing problem in a distributed system consists of a finite set of servers, and a finite set of users is studied. The load balancing problem considered here is a bi-objective problem with two highly probable conflicting objectives: (i) minimizing jobs’ response time (ii) providing the fair utilization of servers. In order to satisfy these two objectives simultaneously, both the objectives are considered in an integrated manner. Next, the load balancing problem is formulated as a noncooperative game; and to solve the game (i.e., to find the Nash equilibrium), a distributed load balancing algorithm (DLBA) is proposed. An experimental study is carried out to ascertain the efficacy of the proposed DLBA. Further, we compare DBLA with three existing load balancing approaches to evaluate its comparative effectiveness. The experimental results validate the effectiveness of the DLBA over the existing approaches.

Near Optimum Random Routing of Uniformly Load Balanced Nodes in Wireless Sensor Networks Using Connectivity Matrix

Wireless sensor networks (WSNs) are ad hoc networks usually made up of miniaturized autonomous units called sensor nodes that communicate in between by radio links. The WSN has attracted a lot of interests for scientific research, particularly because of new problems of routing under strong constraints of network lifetime and low processing capacity of the nodes. Many algorithms for routing such networks are proposed that are generally either network structured or protocol operated. This paper implies that regardless of routing algorithms studied and developed already in literature, all of the WSNs can be routed efficiently in long term if they are employed with uniformly distributed load balancing mechanisms reducing nodes failures in the neighborhood of sink node. To show this, first a WSN is modeled and distribution of nodes is illustrated using a defined overall point factor. Next, an algorithm is developed and simulated the underlying WSN model. Finally, near optimum solutions with their efficiency factors are extracted to deduce that it is possible to quickly route uniformly load balanced networks by random searching of a tiny margin of solutions space.

Spotlight: Scalable Transport Layer Load Balancing for Data Center Networks

Load Balancing plays a vital role in cloud data centers to distribute traffic among instances of network functions or services. State-of-the-art load balancers dispatch traffic obliviously without considering the real-time utilization of service instances and therefore can lead to uneven load distribution and sub-optimal performance. In this article, we design and implement Spotlight, a scalable and distributed load balancing architecture that maintains connection-to-instance mapping consistency at the edge of data center networks. Spotlight uses a new stateful flow dispatcher which periodically polls instances’ load and dispatches incoming connections to instances in proportion to their available capacity. Our design utilizes a distributed control plane and in-band flow dispatching; thus, it scales horizontally in data center networks. Through extensive flow-level simulation and packet-level experiments on a testbed with HTTP traffic on unmodified Linux kernel, we demonstrate that compared to existing methods Spotlight distributes traffic more efficiently and has near-optimum performance in terms of overall service utilization. Compared to existing solutions, Spotlight improves aggregated throughput and average flow completion time by at least 20 percent with infrequent control plane updates. Moreover, we show that Spotlight scales horizontally as it updates the switches at O(100ms) and is resilient to lack of control plane convergence.

FlowMan: QoS-Aware Dynamic Data Flow Management in Software-Defined Networks

In this paper, we study the problem of data flow management in the presence of heterogeneous flows — elephant and mice flows — in software-defined networks (SDNs). Most of the researchers considered the homogeneous flows in SDN in the existing literature. The optimal data flow management in the presence of heterogeneous flows is NP-hard. Hence, we propose a game theory-based heterogeneous data flow management scheme, named FlowMan. In FlowMan, initially, we use a generalized Nash bargaining game to obtain a sub-optimal problem, which is NP-complete in nature. By solving it, we get the Pareto optimal solution for data-rate associated with each switch. Thereafter, we use a heuristic method to decide the flow-association with the switches, distributedly, which, in turn, helps to get a Pareto optimal solution. Extensive simulation results depict that FlowMan is capable of ensuring quality-of-service (QoS) for data flow management in the presence of heterogeneous flows. In particular, FlowMan is capable of reducing network delay by 77.8–98.7%, while ensuring 24.6–47.8% increase in network throughput, compared to the existing schemes such as FlowStat and CURE. Additionally, FlowMan ensures that per-flow delay is reduced by 27.7% with balanced load distribution among the SDN switches.