Proposed Load Balancing Research Articles

A particle swarm optimization inspired global and local stability driven predictive load balancing strategy

In distributed systems and parallel computing, optimal load balancing is difficult. These abstract addresses load balancing in distributed situations, highlighting current solutions' flaws and emphasizing the need for new ones. Load balancing research includes centralized and distributed algorithms, heuristics, and predictive models. Despite various successful methods, workload adaptability, overhead reduction, and scaling to large systems remain unresolved. This study proposes a particle swarm optimization (PSO) load balancing method that considers global and local stability considerations. The proposed method uses PSO principles to balance exploration and exploitation and allocate resources among distributed nodes. Predictive components improve preventative load management by predicting workload changes. Global and local load balancing stability criteria distinguish this study. The recommended method considers global system-wide performance indicators, local node-level characteristics, and micro-level stability to maximize system efficiency. A dual-focus technique distinguishes the proposed load balancing strategy from others, solving dynamic distributed system challenges. The study examines load balancing system advances and suggests improvements and further research. More accurate prediction modeling, stability measures, and application-specific enhancements may be studied in the future. Experimental validation and real-world implementation of the recommended approach are necessary to determine its practicality and ability to handle modern distributed computing systems.

Load-aware continuous-time optimization for multi-agent systems: toward dynamic resource allocation and real-time adaptability

In the realm of next-generation mobile communication networks, characterized by dynamic and evolving workloads, the efficient resource allocation becomes paramount for achieving optimal performance. This paper addresses the intricate challenges of multiagent resource management through the integration of stochastic learning automata and continuous-time optimization techniques, presenting an innovative solution for real-time adaptability and dynamic load balancing. The central aim is to achieve optimal values for local time-varying cost functions while taking into account resource constraints and the feasibility of local allocation. In the context of multi-agent resource allocation systems, where conditions vary over time, the algorithm exhibits a dynamic adaptability, continuously adjusting to the evolving optimal solutions tailored for each agent. Utilizing stochastic learning automata, the algorithm effectively addresses the time-varying load-balancing function in response to dynamic user demands. Furthermore, we introduced an scalable solution for mobility robustness optimization based on deep reinforcement learning (DRL-MRO). This method dynamically identifies network-wide optimal parameter values across the entire network to accommodate diverse mobility patterns, ensuring the effectiveness of load balancing parameters that seamlessly adapt to the dynamic configuration of the network. The overarching goal is to maintain a consistent quality of service level for each agent, thereby enhancing the overall system performance. Simulation outcomes unequivocally affirm the superior performance of the proposed load balancing approach, rooted in continuous-time optimization and stochastic learning automata, surpassing existing schemes across diverse system configurations.

Load balancing for heterogeneous serverless edge computing: A performance-driven and empirical approach

Serverless edge systems simplify the deployment of real-time AI-based Internet of Things (IoT) applications at the edge. However, the heterogeneity of edge computing nodes – in terms of both hardware and software – makes load balancing challenging in these systems. In this paper, we propose a performance-driven, empirical weight-tuning approach to achieve effective load balancing based on the characteristics and capabilities of the nodes. By extensively profiling the nodes, we gather knowledge on performance metrics such as throughput, energy efficiency, response time, AI accuracy, and cost. Using this acquired knowledge, we introduce a weighted round-robin strategy to optimize the performance metrics according to their observed significance. To address multiple objectives, we introduce a multi-objective method that aims to strike a balance between any arbitrary set of performance objectives simultaneously. Additionally, we explore a coordinated distributed approach to overcome the limitations of centralized load balancing. Next, we introduce Hedgi, a heterogeneous serverless edge architecture designed to efficiently configure and utilize the derived load balancing policies, validated empirically. To demonstrate the practicality of Hedgi, we containerize and serverlessize a real-time object detection application. Extensive empirical studies are conducted using Hedgi to evaluate the performance of the proposed load balancing approach. The results provide valuable insights into the design trade-offs of various load balancing policies and system designs in the heterogeneous serverless edge.

Optimizing Cloud Computing Applications with a Data Center Load Balancing Algorithm

Delivering scalable and on-demand computing resources to users through the usage of the cloud has become a common paradigm. The issues of effective resource utilisation and application performance optimisation, however, become more pressing as the demand for cloud services rises. In order to ensure efficient resource allocation and improve application performance, load balancing techniques are essential in dispersing incoming network traffic over several servers. The workload balancing in the context of cloud computing, particularly in the Infrastructure as a Service (IaaS) model, continues to be difficult. Due to available virtual machines and the limited resources, efficient job allocation is essential. To prevent prolonged execution delays or machine breakdowns, cloud service providers must maintain excellent performance and avoid overloading or underloading hosts. The importance of task scheduling in load balancing necessitates compliance with Service Level Agreement (SLA) standards established by cloud developers for consumers. The suggested technique takes into account Quality of Service (QoS) job parameters, VM priorities, and resource allocation in order to maximise resource utilisation and improve load balancing. The proposed load balancing method is in line with the results in the body of existing literature by resolving these problems and the current research gap. According to experimental findings, the Dynamic LBA algorithm currently in use is outperformed by an average resource utilisation of 78%. The suggested algorithm also exhibits excellent performance in terms of accelerated Makespan and decreased execution time.

Balancing of four wire loads using linearized H-bridge static synchronous compensators

<p>In this paper, a load balancing system is designed to balance the secondary phase currents of 11 kV/380 V, 50 Hz, 100 kVA power transformer in a three phase 4-wire, distribution network. The load balancing system is built of six identical modified static synchronous compensators (M-STATCOMs). Each M-STATCOM is constructed of a voltage source converter-based H-bridge controlled in capacitive and inductive modes as a linear compensating susceptance. The M-STATCOM current is controlled by varying its angle such that it exchanges pure reactive current with the utility grid. Three identical M-STATCOMs are connected in delta-form to balance the active phase currents of the power transformer, whereas the other three identical M-STATCOMs are connected in star-form to compensate for reactive currents. The M-STATCOMs in the delta-connected compensator are driven by 380 V line-to-line voltages, whilst, those connected in star-form are driven by 220 V phase voltages. The results of the 220 V and 380 V M-STATCOMs have exhibited linear and continuous control in capacitive and inductive regions of operation without steady-state harmonics. The proposed load balancing system has offered high flexibility during treating moderate and severe load unbalance conditions. It can involve any load unbalance within the power transformer current rating and even unbalance cases beyond the power transformer current rating.<strong></strong></p>

An improved energy-efficient cloud-optimized load-balancing for IoT frameworks

As wireless communication grows, so does the need for smart, simple, affordable solutions. The need prompted academics to develop appropriate network solutions ranging from wireless sensor networks (WSNs) to the Internet of Things (IoT). With the innovations of researchers, the necessity for enhancements in existing researchers has increased. Initially, network protocols were the focus of study and development. Regardless, IoT devices are already being employed in different industries and collecting massive amounts of data through complicated applications. This necessitates IoT load-balancing research. Several studies tried to address the communication overheads produced by significant IoT network traffic. These studies intended to control network loads by evenly spreading them across IoT nodes. Eventually, the practitioners decided to migrate the IoT node data and the apps processing it to the cloud. So, the difficulty is to design a cloud-based load balancer algorithm that meets the criteria of IoT network protocols. Defined as a unique method for controlling loads on cloud-integrated IoT networks. The suggested method analyses actual and virtual host machine needs in cloud computing environments. The purpose of the proposed model is to design a load balancer that improves network response time while reducing energy consumption. The proposed load balancer algorithm may be easily integrated with peer-existing IoT frameworks. Handling the load for cloud-based IoT architectures with the above-described methods. Significantly boosts response time for the IoT network by 60 %. The proposed scheme has less energy consumption (31 %), less execution time (24\\%), decreased node shutdown time (45 %), and less infrastructure cost (48\\%) in comparison to existing frameworks. Based on the simulation results, it is concluded that the proposed framework offers an improved solution for IoT-based cloud load-balancing issues.

Improving a procedure of load balancing in distributed IoT systems

The object of this research is the process of load balancing in distributed Internet of Things (IoT) systems. Within this work, a complex of problems related to efficient load distribution has been addressed. The authors conducted an analysis of existing load-balancing approaches and their drawbacks and proposed an enhanced architecture for the MQTT broker. Additionally, methods and algorithms for load balancing were developed based on multi-criteria server monitoring. Furthermore, the authors created a mathematical model to assess the uniformity of load distribution in the system and introduced a corresponding metric – the load distribution coefficient. In order to evaluate the proposed load balancing methods, a series of experiments were conducted, including the simulation of a distributed IoT system with non-deterministic load. The main goal of these experiments was to assess the uniformity of MQTT load distribution by the broker. The results of the experiments confirmed the hypothesis of improved load distribution efficiency through multi-criteria monitoring-based balancing. The utilization of the proposed load-balancing methods allowed for a more efficient utilization of computational resources. It was found that when using the proposed methods, in the case of non-deterministic load in the IoT system, the load distribution coefficient on average exceeded the corresponding indicator of existing methods by 70 %. In addition, the value of this coefficient for the proposed methods remains virtually unchanged throughout the experiment, which is evidence of the stable operation of the system as a whole. The results obtained can be useful in the development of modern IoT systems.

Spatial–temporal graph neural network traffic prediction based load balancing with reinforcement learning in cellular networks

Balancing network traffic among base stations poses a primary challenge for mobile operators because of the escalating demand for enhanced data speeds in large-scale 5G radio applications. Within cellular networks, traffic flow prediction constitutes a pivotal issue in numerous applications, such as resource allocation, load balancing, and network slicing. In this paper, traffic prediction based load balancing framework with reinforcement learning is proposed to optimize neighbor cell relational parameters that can better balance traffic within a defined geographical cluster. Spatial–temporal-event cross attention graph convolution neural network (STECA-GCN) is put forward to predict the precise traffic flow. The model takes event dimension features into account, while also incorporating direct cross-fusion among diverse features. Concurrently, we have developed a strategy based on deep reinforcement learning to facilitate dynamic load balancing decisions. Simulation results show that our proposed load balancing framework can improve overall system performance. In particular, the combination of loading rate and energy efficiency can achieve a 12% improvement. The load balancing of the base station can better deal with social emergencies.

Optimal load balancing in cloud: Introduction to hybrid optimization algorithm

One of the leading-edge technologies that have made it possible to pay as you go and distribute appropriate resources on demand is cloud computing. The process of optimizing the resource on cloud virtual machines (VMs) is called load balancing. The main goal is to evenly distribute tasks and resources among all machines to ensure that no node is overloaded or inactive. However, the load balancing methods currently used in cloud environments are ineffective in lowering the amount of time needed to respond to requests. This work intends to propose a novel optimal load balancing (LB) method in the cloud. For this a new Intercrossed Chimp and Bald Eagle Algorithm (IC&BA) algorithm is introduced that combines the algorithms like bald eagle search (BES) optimization and chimp optimization (CA) algorithm. Thereby, the proposed load balancing model works under the consideration of (i) Energy Consumption (ii) execution cost (iii) migration cost (iv) Make span and (v) load balancing parameters like response time, Turnaround time.

SAMBA: Sparsity Aware In-Memory Computing Based Machine Learning Accelerator

Machine Learning (ML) inference is typically dominated by highly data-intensive Matrix Vector Multiplication (MVM) computations that may be constrained by the memory bottleneck due to massive data movement between processor and memory units. Although analog in-memory computing (IMC) ML accelerators have been proposed to execute MVM with high efficiency, the latency and energy of such analog computing systems can be dominated by the large latency and energy costs from analog-to-digital converters (ADC). Leveraging sparsity in ML workloads, reconfigurable ADCs can improve the MVM energy and latency by reducing the required ADC bit precision. However, such improvement in latency can be hindered by non-uniform sparsity of the weight matrices mapped into hardware. Moreover, data movement between MVM processing cores may become another factor that delays the overall system-level performance. To address these issues, we propose SAMBA, Sparsity Aware IMC Based Machine Learning Accelerator. First, we propose load balancing at the mapping of weight matrices into physical crossbars. The goal of load balancing is to accommodate reconfigurable ADCs and address the possible delay of MVM processing caused by non-uniform sparsity in the weight matrices. Second, we propose optimizations in arranging and scheduling the tiled MVM hardware to minimize the overhead of data movement across multiple processing cores. Our evaluations show that the proposed load balancing technique can achieve performance improvement by eliminating the non-uniformity in the sparsity of mapped matrices. Moreover, we demonstrate that data movement optimizations can further improve both performance and energy efficiency regardless of sparsity condition. With the combination of load balancing and data movement optimization in conjunction with reconfigurable ADCs, our proposed approach provides up to 2.38x speed-up and 1.54x energy efficiency over state-of-art analog IMC based ML accelerators for ImageNet datasets on Resnet-50 architecture.

Artificial neural network based load balancing scheme for top of rack switches in optical data centers

Abstract Data centers serve as dedicated facilities for housing computer systems and their related components, including telecommunications and storage systems. They typically have high levels of security and environmental controls to ensure that the equipment housed within them functions optimally. Data center networks (DCNs) often employ load balancing algorithms to handle large volumes of traffic and ensure that all servers and switches are utilized equally, keeping the network running smoothly. However, as load on the server varies, therefore dynamic traffic management systems that can adjust traffic flow in real-time based on the current traffic state is required. This study presents an artificial neural network-based load balancing method. By training a feed-forward artificial neural network (ANN) using a back propagation (BP) learning algorithm, it evenly distributes workload over all of the nodes. Simulation results are also presented to prove the usefulness of the proposed load balancing mechanism. It is found that the load balancing scheme can reduce the packet blocking probability (PBP) by 10 folds and delay by about nearly 11 percent.

Improved QoS with Fog computing based on Adaptive Load Balancing Algorithm

As the number of sensing devices rises, traffic on the cloud servers is boosting day by day. When a device connected to the IoTwants access to data, cloud computing encourages the pairing of fog & cloud nodes to provide that information. One of the key needs in a fog-based cloud system, is efficient job scheduling to decrease the data delay and improve the QoS (Quality of Service). The researchers have used a variety of strategies to maintain the QoS criteria. However, because of the increased service delay caused by the busty traffic, job scheduling is impacted which leads to the unbalanced load on the fog environment. The proposed work uses a novel model which curates the features and working style of Genetic algorithm and the optimization algorithm with the load balancing scheduling on the fog nodes. The performance of the proposed hybrid model is contrasted with the other well-known algorithms in contrast to the fundamental benchmark optimization test functions. The proposed work displays better results in sustaining the task scheduling process when compared to the existing algorithms, which include Round Robin (RR) method, Hybrid RR, Hybrid Threshold based and Hybrid Predictive Based models, which ensures the efficacy of the proposed load balancing model to improve the quality of service in fog environment.

Load balanced locality-aware parallel SGD on multicore architectures for latent factor based collaborative filtering

We investigate the parallelization of Stochastic Gradient Descent (SGD) for matrix completion on multicore architectures. We provide an experimental analysis of current SGD algorithms to find out their bottlenecks and limitations. Grid-based methods suffer from load imbalance among 2D blocks of the rating matrix, especially when datasets are skewed and sparse. Asynchronous methods, on the other hand, can face cache issues due to their memory access pattern. We propose bin-packing-based block balancing methods that are alternative to the recently proposed BaPa method. We then introduce Locality Aware SGD (LASGD), a grid-based asynchronous parallel SGD algorithm that efficiently utilizes cache by changing nonzero update sequence without affecting factor update order and carefully arranging latent factor matrices in the memory. Combined with our proposed load balancing methods, our experiments show that LASGD performs significantly better than alternative approaches in parallel shared-memory systems.

Load balancing in 5G heterogeneous networks based on automatic weight function

Load balancing is a major challenge in heterogeneous networks (HetNets) consisting of 5G and 6G ultra-dense small cells with long-term evaluation advanced (LTE-A) networks. A key factor in achieving efficient load balancing during user mobility is creating appropriate optimisation for handover control parameters (HCP). This paper proposes a coordinated load balancing algorithm for LTE-A/fifth generation (5G) HetNets. The algorithm automatically optimises HCP settings for a given user based on three bounded functions (the signal-to-interference-plus-noise ratio (SINR) of the user equipment (UE), the number of physical resource blocks (PRBs) per UE and the UE’s speed) as well as their automatic weight levels. A two-step target cell determination strategy is implemented according to the cell load level and RSRP criteria, ensuring that users are handed over to low-loaded target cells. A new HO procedure that considers the pilot signal power is also proposed, which includes the number of PRBs per UE and the RSRP. Cells with freer PRBs are prioritised in user association to provide load balance and enhanced throughput. The proposed load balancing algorithm is compared with five other load balancing algorithms selected from the literature. The simulation results reveal that under various mobile speed scenarios, the proposed load balancing scheme enhances network performance in terms of load level, throughput, spectral efficiency and call dropping ratio (CDR).

Efficient Load Balancing for Blockchain-Based Healthcare System in Smart Cities

Smart cities are emerging rapidly due to the provisioning of comfort in the human lifestyle. The healthcare system is an important segment of the smart city. The timely delivery of critical human vital signs data to emergency health centers without delay can save human lives. Blockchain is a secure technology that provides the immutable record-keeping of data. Secure data transmission by avoiding erroneous data delivery also demands blockchain technology in healthcare systems of smart cities where patients’ health history is required for their necessary treatments. The health parameter data of each patient are embedded in a separate block in blockchain technology with SHA-256-based cryptography hash values. Mining computing nodes are responsible to find a 32-bit nonce (number only used once) value for each data block to compute a valid SHA-256-based hash value in blockchain technology. Computing nonce for valid hash values is a time-taking process that may cause life losses in the healthcare system. Increasing the mining nodes reduces this delay; however, the uniform distribution of mining data blocks to these nodes by considering the priority data is a challenging task. In this work, an efficient scheme is proposed for scheduling nonce computing tasks at the mining nodes to ensure the timely execution of these tasks. The proposed scheme consists of two parts, the first one provides a load balancing scheme to distribute the nonce execution tasks among the mining nodes such that makespan is minimized and the second part prioritizes more sensitive patient data for quick execution. The results show that the proposed load balancing scheme effectively allocates data blocks in different mining nodes as compared to round-robin and greedy algorithms and computes hash values of most of the higher-risk patients’ data blocks in a reduced amount of time.

PGWO‐AVS‐RDA: An intelligent optimization and clustering based load balancing model in cloud

SummaryLoad balancing and task scheduling in cloud have gained a significant attention by many researchers, due to the increased demand of computing resources and services. For this purpose, there are various load balancing methodologies are developed in the existing works, which are mainly focusing on allocating the tasks to Virtual Machines (VMs) based on their priority, order of tasks, and execution time. Still, it facing the major difficulties in finding the best tasks for allocation, because the sequence of patterns are normally used to categorize the relevant tasks with respect to the load. Thus, this research work intends to develop an intelligent group of mechanisms for efficiently allocating the tasks to the VMs by finding the best tasks with respect to the scheduling parameters. Initially, the user tasks are given to the load balancer unit, where the Probabilistic Gray Wolf Optimization (PGWO) technique is used to find the best fitness value for selecting the tasks. Then, the Adaptive Vector Searching (AVS) methodology is utilized to cluster the group of tasks for efficiently allocating the tasks with improved Quality of Service (QoS). Finally, the Recursive Data Acquisition (RDA) based scheduler unit can allocate the clustered tasks to the appropriate VMs in the cloud system by analyzing the properties of storage capacity, balancing load of VM, CPU usage, memory consumption, and execution time of tasks. During evaluation, the performance of the proposed load balancing model is validated by using various measures. Then, the obtained results are compared with some state‐of‐the‐art models for proving the betterment of the proposed scheme.

Effective load balancing strategy using fuzzy golden eagle optimization in fog computing environment

Fog computing (FC) is becoming popular for connecting real-time Internet of Things (IoT) applications such as driverless cars and health care. Although fog has low latency compared to the cloud, the fog becomes overloaded when serving a large number of IoT devices. FC struggles while distributing the load to the resources that may result in certain fog nodes being overloaded or underloaded. To overcome these issues, an effective load balancing algorithm named Fuzzy Golden Eagle Load Balancing (FGELB) is proposed. The proposed load balancing strategy consists of three stages, namely: assigning priority to the tasks, resource ranking and scheduling, and power management. Initially, the priority is assigned to the incoming task by considering the deadline time, task size, and predefined priority using a fuzzy algorithm. This prioritization of tasks helps to execute the important task without delay. Then the Golden Eagle Optimization algorithm (GEOA) is used to rank and schedule the resources. This ranking and scheduling help the tasks to be executed with suitable resources. Ranked resources execute the tasks in the resource engine, and the power engine manages the power by enabling/disabling the resources based on their necessity. The performance of the approach is tested and it is compared with the results of existing methods in terms of energy consumption, failure rate, computational cost, communication overhead, average turnaround time, and waiting time. The results show that the proposed FGELB approach achieves higher performance than the existing load balancing approaches.

Enhanced performance of long-term evolution small-cell networks using improved mobility algorithms

With an ever-increasing number of user equipment (UE) and bandwidth demands of new applications, the deployment of dense heterogeneous cellular networks has been adopted in many network scenarios. The small cells are unable to unload the traffic due to the random UEs mobility and cells deployment. This degrades the network performance such as handover success, throughput, and load distribution. To address such a problem, we propose an enhanced mobility load-balancing algorithm for small cells. The conventional mobility load balancing (MLB) algorithms study only the fixed or adaptive thresholds of the network separately to consider the load balance process, while other MLB algorithms consider the neighboring cells of the network those experience unnecessary MLB actions. The proposed load balancing algorithms study overloaded cells and neighbors using the proposed efficiency parameter, B. To identify the overloaded cells, we propose B which compares between a pre-defined threshold and the network threshold to categorize medium and overloaded cells, based on B, the algorithm is triggered. Then, to control the shifted load to a target neighbor cell, we propose the Rescue factor, f. the f ensures the load of the target cell after handover is equal or less than Ɓ. The simulation results showed a lower standard deviation and higher throughput and physical resource block utilization than surveyed algorithms.

Towards Investigation of Various Load Balancing Techniques in Cloud Computing

Cloud computing is an on-demand service where customers may access any time common IT resources, information, software and other equipment. It is a web-based development that offers virtual resources over the Internet as a service. The higher the cloud use, the higher the charge. The allocation of loads to components processing is a challenging phenomenon.In a multi-node system, there is a very good possibility that some nodes will be idle while others will be overloaded. The load balancing algorithms' objective is to keep the load on each processing element constant. A technique for Random Load Balancing in Cloud Computing is devised, in which the amount of user requests is instantly assigned to the resources. This method employs a random strategy to decrease the request's waiting time. This proposed load balancing technique is simulated in Cloud Analyst tool. The performance comparison study between proposed random method and other available load balancing methods is conducted. It is found that the performance of proposed random method is better in terms of response time.Keywords: Cloud Computing, Load Balancing, Response Time, Virtual Machine, Sorting

OpenDaylight vs. Floodlight: Comparative Analysis of a Load Balancing Algorithm for Software Defined Networking

This paper presents the proposal of a load balancing algorithm implemented in two of the most popular controllers for Software Defined Networks (SDN): OpenDaylight and Floodlight. A comparative study in terms of the available bandwidth and delay time of the packet forwarding was performed by means of simulation modeling in a base network in which a shortest path algorithm was implemented as well. The results show that the proposed load balancing algorithm improves significantly the performance of a SDN in terms of the offered QoS of a OpenDaylight based controller. The effect of the proposed load balancing algorithm in the Floodlight controller shows a smaller impact mainly on the bandwidth allocation due to its in-build modules that by default perform specific routing and forwarding operations efficiently according to the traffic demand.