Load Balancing Capabilities Research Articles

GSADDQN: Combining GraphSAGE and reinforcement learning for routing optimization in software-defined optical transport network

As a result of rapid development of network communication technology, optical transport network (OTN) traffic has also experienced rapid growth in terms of information volume scale, traffic complexity, and spatiotemporal distribution dynamics. Because the OTN traffic demand has complex spatiotemporal fluctuations, the traditional deep reinforcement learning (DRL) algorithm has been applied to the routing optimization of software-defined OTNs. However, the traditional DRL algorithm has problems such as slow convergence, weak generalization ability, and load imbalance when performing routine tasks. To address these issues, we propose a graph sampling and aggregation (GraphSAGE)-based dueling deep Q-network (GSADDQN) algorithm for software-defined OTN routing optimization. First, we design a DRL-based routing decision model to find the best routing strategy for each optical network’s source–destination traffic demand. Second, considering the sparse connection characteristics of optical network nodes, we use sampling neighbors and a deep aggregation mechanism as the neural network model of the Dueling Deep Q-Network (Dueling DQN) algorithm so that the reinforcement learning agent can consciously aggregate important network information and improve the model’s convergence performance and generalization ability. Finally, we design simulation routing experiments based on Gym and evaluate the algorithm’s load balancing and generalization capabilities for different network topologies. The experimental results show that the GSADDQN algorithm has good convergence performance and load balancing ability in routing optimization of optical transmission networks and can generalize new network structures, maintaining good decision-making ability even during network node failures.

On the Reliability of Wireless Sensor Networks with Multiple Sinks.

The convergence of heterogeneous wireless sensor networks provides many benefits, including increased coverage, flexible load balancing capabilities, more efficient use of network resources, and the provision of additional data by different types of sensors, thus leading to improved customer service based on more complete information. However, despite these advances, the challenge of ensuring reliability and survivability remains due to low-cost sensor requirements and the inherent unreliability of the wireless environment. Integrating different sensor networks and unifying protocols naturally leads to the creation of a network with multiple sinks, necessitating the exploration of new approaches to rational reliability assurance. The failure of some sensors does not necessarily lead to a shutdown of the network, since other sensors can duplicate information and deliver data to sinks via an increased number of alternative routes. In this paper, the reliability indicator is defined as the probability that sinks can collect data from a given number of sensors. In this context, a dedicated reliability metric is introduced and examined for its effectiveness. This metric is computed using an algorithm rooted in the modified factoring method. Furthermore, we introduce a heuristic algorithm designed for optimal sink placement in wireless sensor networks to achieve the highest level of network reliability.

Deep reinforcement learning-based resource scheduling for energy optimization and load balancing in SDN-driven edge computing

Traditional techniques for edge computing resource scheduling may result in large amounts of wasted server resources and energy consumption; thus, exploring new approaches to achieve higher resource and energy efficiency is a new challenge. Deep reinforcement learning (DRL) offers a promising solution by balancing resource utilization, latency, and energy optimization. However, current methods often focus solely on energy optimization for offloading and computing tasks, neglecting the impact of server numbers and resource operation status on energy efficiency and load balancing. On the other hand, prioritizing latency optimization may result in resource imbalance and increased energy waste. To address these challenges, we propose a novel energy optimization method coupled with a load balancing strategy. Our approach aims to minimize overall energy consumption and achieve server load balancing under latency constraints. This is achieved by controlling the number of active servers and individual server load states through a two stage DRL-based energy and resource optimization algorithm. Experimental results demonstrate that our scheme can save an average of 19.84% energy compared to mainstream reinforcement learning methods and 49.60% and 45.33% compared to Round Robin (RR) and random scheduling, respectively. Additionally, our method is optimized for reward value, load balancing, runtime, and anti-interference capability.

A Local Pre-Rerouting Algorithm to Combat Sun Outage for Inter-Satellite Links in Low Earth Orbit Satellite Networks

Satellite networks show the development trend in global coverage, flexible access, and reliable transmission. They are the key to building a wide coverage, massive connection, three-dimensional, all-round, all-weather, space-, air- and ground-integrated information network. Due to its high frequency and large bandwidth, the laser inter-satellite link (ISL) is easily interfered with by sunlight. When sun outage occurs, the laser inter-satellite link will be interrupted for a long time. Most of the existing satellite link failure-routing algorithms are distributed and have poor load balancing capabilities. The centralized algorithm calculates paths independently for different services, and the effect is not ideal. In this paper, a local pre-rerouting algorithm (LPRA) is proposed. The center node collects the link information from the limited area before sun outage and the service information in the link that is about to be interrupted, and reroutes these services before sun outage occurs. After the occurrence of sun outage, services that originally needed to go through the failed link are forwarded according to the path calculated in advance, thereby avoiding link congestion and reducing the packet loss rate. Simulation results show that the proposed algorithm performs better than existing algorithms in terms of packet loss ratio, total delay and load balancing capabilities.

Optimizing QoS Metrics for Software-Defined Networking in Federated Learning

In the modern and complex realm of networking, the pursuit of ideal QoS metrics is a fundamental objective aimed at maximizing network efficiency and user experiences. Nonetheless, the accomplishment of this task is hindered by the diversity of networks, the unpredictability of network conditions, and the rapid growth of multimedia traffic. This manuscript presents an innovative method for enhancing the QoS in SDN by combining the load-balancing capabilities of FL and genetic algorithms. The proposed solution aims to improve the dispersed aggregation of multimedia traffic by prioritizing data privacy and ensuring secure network load distribution. By using federated learning, multiple clients can collectively participate in the training process of a global model without compromising the privacy of their sensitive information. This method safeguards user privacy while facilitating the aggregation of distributed multimedia traffic. In addition, genetic algorithms are used to optimize network load balancing, thereby ensuring the efficient use of network resources and mitigating the risk of individual node overload. As a result of extensive testing, this research has demonstrated significant improvements in QoS measurements compared to traditional methods. Our proposed technique outperforms existing techniques such as RR, weighted RR, server load, LBBSRT, and dynamic server approaches in terms of CPU and memory utilization, as well as server requests across three testing servers. This novel methodology has applications in multiple industries, including telecommunications, multimedia streaming, and cloud computing. The proposed method presents an innovative strategy for addressing the optimization of QoS metrics in SDN environments, while preserving data privacy and optimizing network resource usage.

Creating a call center test bench for load balancing Asterisk servers in a cluster

The article deals with the issues of increasing throughput in call centers. The current solution is to cluster call servers and evaluate their characteristics to ensure efficient operation and the necessary fault tolerance. It is shown that one of the main aspects of the quality functioning of the call center is load balancing of servers in the cluster.
 The features of the call-center modeling process are considered. The organization scheme of the call center of the company and the network model of the call center have been created. Virtualization technology was used to create a network model of a call center. The VMWare ESXI 6.7 hypervisor and the vCenter client were used as a network configuration platform.
 An analysis of load balancing was carried out using different algorithms and strategies.
 Asterisk PBX was configured and a server cluster was created. A test bench was developed and configured using the Zabbix open source product to explore a cluster of call servers. A bandwidth characteristic for the Asterisk _1 server and a network map were obtained, which actually represents a simulated structure of the call center network.
 The process of load testing on three Asterisk servers and the implementation of Zabbix load balancing capabilities are shown. A custom SIPp session script has been created for accepting calls and load testing. The number of calls that the call center is able to handle is calculated. It has been determined that a single Asterisk server with its current settings can handle a maximum of 915 concurrent calls. The process of load balancing on a cluster of SIP servers has been launched. A cluster of Asterisk servers has been found to be capable of handling 2550 simultaneous calls.

DDCM: a decentralized density clustering and its results gathering approach

The use of distributed clustering is an important method of solving large-scale data mining problems. There are still some problems associated with distributed clustering, such as a performance bottleneck on the master node and network congestion caused by global broadcasting. This paper proposes a decentralized clustering method based on density clustering and the content-addressable network technique. It can form a cluster with excellent scalability and load balancing capabilities based on several surrounding nodes. In addition, a method is presented for optimizing the way clustering results are gathered in different application scenarios. Based on our extensive experiments, the proposed approach performs three times better than benchmark algorithms in terms of efficiency and has a stable expanding ratio of about 0.6 for large-scale data sets.

Implementing ML Models in Load Balancing to Improve Application Performance

In modern distributed systems, load balancing plays a critical role in ensuring optimal performance and user experience. However, traditional static load balancing mechanisms often fail to adapt to dynamic traffic patterns, leading to performance degradation, increased latency, and inefficient resource utilization. This paper presents a novel approach that leverages machine learning (ML) models to enhance load balancing by predicting traffic fluctuations and intelligently distributing workloads in real time. By training ML models on historical traffic data and application performance metrics, we enable the system to make proactive decisions about resource allocation. This approach improves the ability to handle traffic surges during peak periods, minimizes latency, and optimizes infrastructure usage. The research outlines the implementation of various ML techniques, such as reinforcement learning and neural networks, into a microservices-based architecture, demonstrating how these models enhance both load balancing and auto-scaling capabilities. Empirical results from the study reveal that ML-driven load balancing reduces latency by up to 40%, improves resource efficiency, and lowers infrastructure costs by 30%, compared to traditional methods. The paper concludes by discussing the technical challenges, future possibilities of using more advanced ML algorithms, and the broader implications for cloud-native application performance.

Supercharging the APGAS Programming Model with Relocatable Distributed Collections

In this article, we present our relocatable distributed collection library. Building on top of the AGPAS for Java library, we provide a number of useful intranode parallel patterns as well as the features necessary to support the distributed nature of the computation through clearly identified methods. In particular, the transfer of distributed collections’ entries between processes is supported via an integrated relocation system. This enables dynamic load-balancing capabilities, making it possible for programs to adapt to uneven or evolving cluster performance. The system we developed makes it possible to dynamically control the distribution and the data flow of distributed programs through high-level abstractions. Programmers using our library can, therefore, write complex distributed programs combining computation and communication phases through a consistent API. We evaluate the performance of our library against two programs taken from well-known Java benchmark suites, demonstrating superior programmability and obtaining better performance on one benchmark and reasonable overhead on the second. Finally, we demonstrate the ease and benefits of load balancing and a more complex application, which uses the various features of our library extensively.

Research on Generalized Intelligent Routing Technology Based on Graph Neural Network

Aiming at the problems of poor load balancing ability and weak generalization of the existing routing algorithms, this paper proposes an intelligent routing algorithm, GNN-DRL, in the Software Defined Networking (SDN) environment. The GNN-DRL algorithm uses a graph neural network (GNN) to perceive the dynamically changing network topology, generalizes the state of nodes and edges, and combines the self-learning ability of Deep Reinforcement Learning (DRL) to find the optimal routing strategy, which makes GNN-DRL minimize the maximum link utilization and reduces average end-to-end delay under high network load. In this paper, the GNN-DRL intelligent routing algorithm is compared with the Open Shortest Path First (OSPF), Equal-Cost Multi-Path (ECMP), and intelligence-driven experiential network architecture for automatic routing (EARS). The experimental results show that GNN-DRL reduces the maximum link utilization by 13.92% and end-to-end delay by 9.48% compared with the superior intelligent routing algorithm EARS under high traffic load, and can be effectively extended to different network topologies, making possible better load balancing capability and generalizability.

Cloud Resource Scheduling Method based on Markov Process and the Cuckoo Search

With the rapid development of cloud computing, the computing reliability of cloud servers has become an important research area. To improve the load balancing capability of cloud servers, an autonomous adaptive architecture cloud computing framework based on software-defined networking and network function virtualization were proposed. And a cloud resource scheduling method based on the Markov process model and cuckoo search algorithm to realize the load balancing of cloud nodes was designed. The cloud platform verified by the experiment shows that the method can effectively guarantee the service reliability of cloud computing.

Efficient and Automated Deployment Architecture for OpenStack in TianHe SuperComputing Environment

Recently, with the large-scale outbreak of the global financial crisis and public safety incidents (such as COVID-19), high-performance computing has been widely applied to risk prediction, vaccine development, and other fields. In scenarios where high-performance computing infrastructure responds to the instantaneous explosion of computing demands, a crucial issue is to provide large-scale flexible allocation and adjustment of computing capability by rapidly constructing computing clusters. Existing large-scale computing cluster deployment solutions usually utilize source code deployment or other deployment tools. The great challenge of existing deployment methods is to reduce excessive image distribution time and refrain from configuration defects. In this paper, we design an intelligent distributed registry deployment (IDRD) architecture based on the OpenStack cloud platform, which adaptively places distributed image repositories using the containerized deployment of multiple registries. We propose a server load priority algorithm to solve multiple registries placement problems in IDRD. Furthermore, we devise a clustering algorithm based on demand density that can optimize the global performance of IDRD and improve large-scale cluster load balancing capabilities, which has been implemented in the TianHe Supercomputing environment. Extensive experimental results demonstrate that IDRD can effectively reduce 30%-50% of the distribution time of component images and significantly improve the efficiency of large-scale cluster deployment.

Dynamic Distributed Multi-Path Aided Load Balancing for Optical Data Center Networks

Benefiting from dense connections in data center networks (DCNs), load balancing algorithms are capable of steering traffic into multiple paths for the sake of preventing traffic congestion. However, given each path’s time-varying and asymmetrical traffic state, this may also lead to worse congestion when some paths are overutilised. Especially in the two-tier hybrid optical/electrical DCNs (Hoe-DCNs), the port contentions and large-grained optical packets of the fast optical switch (FOS) require the top-of-rack (TOR) switch to have microsecond-level load balancing capability for microburst traffic. This paper establishes a leaf-spine Hoe-DCN model to illustrate the principal characteristic of dynamic load balancing in TOR switches for the first time. Moreover, we propose the dynamic distributed multi-path (DDMP) load balancing algorithm that relies on dynamic hashing computing for network flow distribution in DCNs, which dynamically adjusts traffic flow distribution at microsecond level according to the inverse ratio of the buffer occupancy. The simulation results show that our proposed algorithm reduces the TOR-to-TOR latency by 15.88% and decreases the packet loss by 22.06% compared to conventional algorithms under regular load conditions, which effectively improves the overall performance of the Hoe-DCNs. Moreover, our proposed algorithm prevents more than 90% packet loss under low load conditions.

KubFBS: A fine‐grained and balance‐aware scheduling system for deep learning tasks based on kubernetes

AbstractThe past decade witnessed a remarkable increase in deep learning (DL) workloads which require GPU resources to accelerate the training process. However, the existing coarse‐grained scheduling mechanisms are agnostic to information other than the number of GPUs or GPU memory, which results in performance degradation of DL tasks. Moreover, the common assumption held by the existing balance‐aware DL task scheduling strategies, a DL task consumes resources once it starts, fails to reduce resource contention, and further limits execution efficiency. To address these problems, this article proposes a fine‐grained and balance‐aware scheduling model (FBSM) which considers the resource consumption characteristic of the DL task. Based on FBSM, we propose customized GPU sniffer (GPU‐S) and balance‐aware scheduler (BAS) modules to construct a scheduling system called KubFBS. The experimental results demonstrate KubFBS accelerates the execution of DL tasks while improving the load balancing capability of the cluster.

SCADA-Based Message Generator for Multi-Vendor Smart Grids: Distributed Integration and Verification of TASE.2.

Recent developments in massive machine-type communication (mMTC) scenarios have given rise to never-seen requirements, which triggered the Industry 4.0 revolution. The new scenarios bring even more pressure to comply with the reliability and communication security and enable flawless functionality of the critical infrastructure, e.g., smart grid infrastructure. We discuss typical network grid architecture, communication strategies, and methods for building scalable and high-speed data processing and storage platform. This paper focuses on the data transmissions using the sets of standards IEC 60870-6 (ICCP/TASE.2). The main goal is to introduce the TASE.2 traffic generator and the data collection back-end with the implemented load balancing functionality to understand the limits of current protocols used in the smart grids. To this end, the assessment framework enabling generating and collecting TASE.2 communication with long-term data storage providing high availability and load balancing capabilities was developed. The designed proof-of-concept supports complete cryptographic security and allows users to perform the complex testing and verification of the TASE.2 network nodes configuration. Implemented components were tested in a cloud-based Microsoft Azure environment in four geographically separated locations. The findings from the testing indicate the high performance and scalability of the proposed platform, allowing the proposed generator to be also used for high-speed load testing purposes. The load-balancing performance shows the CPU usage of the load-balancer below 15% while processing 5000 messages per second. This makes it possible to achieve up to a 7-fold improvement of performance resulting in processing up to 35,000 messages per second.

RETRACTED: A border surveillance system to sense terrorist outbreaks

Wireless Sensor Networks (WSN) are widely used in military applications such as border surveillance. This research focuses on developing a WSN network for implementing in the India-Pakistan border. The traditional approaches used for border surveillance has challenges such as energy efficiency, low cost, latency, information integrity, computation, and communication limitations and needs manual intervention. The conventional method of employing military personnel, fences, and building bulletproof walls and excavation caves at borders are replaced by WSN based sensors, radars, surveillance cameras and Unmanned Ariel Vehicles (UAV) by creating a virtual fence. A hybrid WSN based Border Surveillance System (BSS) architecture using activation scheduling strategy for high energy efficiency, load balancing capabilities, and network lifetime increment is proposed. The proposed integrated BSS architecture reduces manual intervention to a high extent by deploying sensors radars, cameras, and UAVs in place of humans. The results show that the proposed structure and activation methodology are accurate and precise in surveillance and require less maintenance, involve low-cost deployment and provide improved reliability. The experimental analysis shows that the system outperforms the other alternatives in detecting intrusion in boundary fields and has an increased lifetime via effectively handling the network resources.

A Performance Evaluation Method of Load Balancing Capability in SaaS Layer of Cloud Platform

Cloud platform load balancing has become one of the essential elements in modern cloud platform operations. However, there are still some problems in the evaluation methods of cloud platform load balance, such as unreasonable evaluation indexes and lack of comprehensive evaluation methods. This paper first proposes a performance evaluation method of cloud platform load balancing capabilities. This method analyzes the characteristics of the cloud platform and the actual needs of users, designs and proposes a cloud platform load balancing capability evaluation system, quantitatively analyzes and quantifies specific indexes, and uses the Analytic Hierarchy Process to determine the weight of each index. Finally, we deploy and test our method on the OpenStack cloud platform. Experiments show that the evaluation results of the method proposed in this paper are consistent with the actual results, which verifies the feasibility and rationality of the method proposed in this paper.

Intelligent Government Control System Construction Based on 5G IoT

This paper presents an in‐depth analysis of the construction of an intelligent government control system in the context of 5G IoT. The research has conducted performance tests and network tests on the samples of the components which debugged the problems that occurred during the tests and analyzed and improved the system. The system designed in this study is not only easy to maintain but also very reliable and safe, with a high level of accuracy, and can be used for remote “five remote” and management of switchgear and other related power equipment. Based on this model, a dual middleware PaaS service selection algorithm is proposed, and a utility evaluation strategy is used to load balance the PaaS service. The test results show that the DMPS algorithm has a high selection accuracy and load balancing capability. Combined with the experience of local government investment and financing platform financing risk management and control practices, the financing risk control system established through the study is made more feasible. From the development history of local government investment and financing platforms, we study local government investment and financing platforms from the perspective of historical development and explore the root causes of their high financing risks and possible defects in the management process of daily work. Concerning the existing laws and regulations, practical operation norms, the framework of the local government investment and financing platform system, and other objective conditions, we start to build a financing risk control system for local government investment and financing platforms from both practical and operational management aspects, to find a financing risk control system that is in line with the regulations and effective. Finally, the system will be put into practice, and the effectiveness of the system will be tested by using the simulation operation method, and the results will be adjusted in time for the shortcomings of the system. The system is expected to help local government investment and financing platforms to reduce their risks so that they can provide better services for urban infrastructure construction and become a solid link in the economic development chain.

Load balancing and neural dynamic model to optimize replicator dynamics controllers for vibration reduction of highway bridge structures

Earthquakes cause irreparable damages to the built environment, which has led bridge engineers to develop structural control systems to mitigate damage and improve vibration reduction in real-time. Among control systems, base isolation is one of the most commonly used passive control strategies for seismic protection of civil structures. Yet, it lacks real-time adaptability, has lower energy dissipation, and poor performance during near-fault earthquakes. To overcome these limitations, a hybrid control system comprised of semi-active magneto-rheological (MR) dampers and passive base isolation bearings is installed at the deck and piers for vibration reduction of highway bridge structures. This paper, inspired by evolutionary game theory and artificial intelligence, proposes data-driven replicator dynamic control algorithms to distribute the command voltage to the current driver of the semi-active MR dampers. It incorporates a load balancing strategy to reallocate additional resources. To achieve a high-performance design of the game-theory-inspired controllers, a patented neural dynamic model is used to optimize the control parameters. The evaluation of the proposed methodology uses a benchmark control problem based on the 91/5 highway bridge in Southern California subjected to near-field earthquake accelerograms. The performance of five different proposed controllers is compared with conventional Lyapunov and fuzzy logic control algorithms using 21 performance criteria. Results show the load balancing capability of the proposed control algorithms to mitigate the vibrations experienced by the bridge structure and further increase the durability of semi-active devices. The novelty of the methodology impacts how game-theory controllers make control decisions among multiple devices in engineering problems.

Perspectives on energy efficiency and smart energy systems from the 5th SESAAU2019 conference

This short paper gives some perspectives on the recent development in the concept of Smart Energy Systems while introducing the contents of selected papers from the 5th International Conference on Smart Energy Systems, 4th Generation District Heating, Electrification, Electrofuels and Energy Efficiency – SESAAU2019. All of the papers elaborate on or contribute to the theoretical scientific understanding on how to design and implement an energy efficient green transition with an emphasis on system integration and sector-coupling. Low-temperature 4th generation district heating and its possibilities of providing flexible demand is particularly emphasized in the papers presented. However, in order to reach 100% renewability in the energy system, further research is required particularly on the transformation of the transport and industrial sectors. Methods for accommodating increased levels of renewable energy sources, while maintaining proper load-balancing capability and security of supply, should also be further investigated – an area where smart energy systems play an important role.