Fig 1 - available via license: Creative Commons Attribution 4.0 International
Content may be subject to copyright.
Topology of P2P overlay network
Source publication
In P2P systems, large volumes of data are declustered naturally across a large number of peers. But itis very difficult to control the initial data distribution because every user has the freedom to share any datawith other users. The system scalability can be improved by distributing the load across multiple serverswhich is proposed by replication...
Similar publications
Massive multiuser virtual environments (MMVEs) are rapidly expanding both in the number of users and complexity of interactions. Their needs of computational resources offer new challenges for the computer scientists. In this paper we present some ideas on the implementation of a massive simulation environments, a particular MMVE, distributed over...
Grid and P2P environments often require to manage a wide number of distributed and mobile entities. In this paper we present a scalable partitioning technique of the entity space, based on Voronoi diagrams, that is well-suited for distributed applications with dynamically changing topology. In HyVonNe (Hybrid Voronoi Network) architecture, the enti...
Previous analytical studies of unstructured P2P resilience have assumed exponential user lifetimes and only considered age independent neighbor replacement. In this system, the limitations are overcome by introducing a general node-isolation model for heavy-tailed user lifetimes and arbitrary neighbor-selection algorithms. Using this model, analysi...
Salah satu permasalahan yang sering terjadi pada lingkungan komputasi klaster adalah terjadinya beban yang tidak seimbang yang dapat menurunkan Quality of Service (QoS). Oleh sebab itu, dibutuhkan metode load balancing yang handal dalam pendistribusian beban. Pada beberapa kasus, metode load balancing dinamis gagal berperan sebagai metode load bala...
In today’s business environment, reducing costs is crucial due to the variety of Internet of Things (IoT) devices and security infrastructure. However, applying security measures to complex business scenarios can lead to performance degradation, making it a challenging task. To overcome this problem, we propose a novel algorithm based on deep reinf...
Citations
... However, since each user is free to divide any data with additional users, it is exceedingly difficult to organize the original data distribution. In peer-to-peer contented distribution systems, load-balancing is offered using a cluster-based imitation architecture [12] [2]. It includes an effective load balancing strategy in addition to an intellectual duplicate position technique [13]. ...
... The proposed method offers higher performance value and lower delay time compared to the load balancing method of cluster-based replication architecture [44]. The existing method has a throughput is 1, yet the novel CSO have 2.6 × 10 10 Mbps. ...
Peer-to-Peer (P2P) networks are less expensive, simple to use, and do not require the traditional client–server model. It has particular advantages in data sharing and resource utilization, so it is recommended to use it for various applications. P2P networks have been used in many applications, especially in data sharing and resource utilization. Load balancing and security is an essential task to improve the performance of P2P networks. Hence, in this paper, probability-based load balancing control and security enhancement is developed in P2P networks. The probability of peer can be computed with chicken swarm optimization (CSO), which selects the best peer in P2P networks to achieve load balancing and resource utilization. The proposed method is developed to attain two main objective functions: load balancing control and security enhancement. A probability-based CSO algorithm is used to control load balancing. The security is achieved with Enhanced Rumour Riding protocol (ERR) and SXOR (Split XOR) operation. The proposed method is implemented in the NS2 platform, and the performance of the proposed method is analysed with performance metrics such as delay, delivery ratio, packet loss, encryption time, decryption time, and throughput. The proposed method is compared with existing methods such as Biased Contribution Index based Rumour Riding protocol (BCIRR), Ant Colony Optimization (ACO), and Catching Algorithms (CA). The proposed technique achieves a 98.75% packet delivery ratio, with a minimum 3.8 s delay. Ultimately the performance suggests that the proposed system can perform better for load balancing and security in the P2P network.
... Load balancing is a key component of highly-available infrastructures [2] commonly used to improve the performance and reliability of web sites, applications, databases and other services by distributing the workload across multiple servers. ...
Load balancing is a technique used to distribute the work load on single server, distributed severs and
cloud etc. It is used to provide reliability, high availability and scalability services to the end users. In this we have
studied different areas of load balancing and made a comparative analysis. We have also presented performance
measures for load balancing with a case study approach
... Especially with the coverage area expansion of distribution service and the scale of user collective, RPAs play a key role in efficient content distribution and network running. However, replica placement has to solve two problems: (1) which content replicas should be placed in CDNs (2) where these replicas should be placedThe replica placement algorithm reduces communication overhead and response delay by reasonable replica distribution in CDNs, hence they can improve the performance of network distribution and satisfy the benefits of operators. Most existing replica placement algorithms consider only ISPs' benefits, and do not consider user preferences for content to meet their personalized demand. ...
... Existing RPAs mainly use greedy, random or other improved algorithms, with optimal goals such as response delay [1], load balancing [2], [3], communication cost and storage overhead [4]- [6], and constraints of storage capacity, fewest replicas or others,which determine the replica placement scheme. The greedy algorithm places all content replicas on the source server initially, when the system overhead is the largest. ...
... With the increase of replica numbers, the average response time of every placement algorithm decreases gradually. As shown in Fig. 1, When the number of replicas is few (1)(2)(3) in the network, the average response time of UIARP is not significantly reduced compared with that of l-Greedy-Insert algorithm, but which is more significant than that of Greedy algorithm; with the increase of replica numbers, the average response time of UIARP is reduced by about 44% compared with that of l-Greedy-Insert, the average reduction about 60% compared with Greedy algorithm, thus, we can conclude that this algorithm has some certain advantages in reducing the average response time. Because of the constraints of replica server capacity and request response tolerance limit, the average response time tends to be stable after reaching the limit of the server capacity. ...
A user interest-aware content replica optimized placement algorithm (UIARP) is proposed. Firstly, the interest subjects of the user-collective are extracted from their content access logs by clustering algorithms, and according to the weighting of the individual interest degree, their collective interest degree would be got and updated in real time; then under the nonlinear optimization model, replicas of larger collective interest degree have priority to be placed, with the goal of minimizing the average response time, which achieves the maximum match between placing replicas and users' content demand. This algorithm not only ensures that users get interested replicas quickly, but also improves the system efficiency. From four aspects including average response time, the matching degree of request response, load balancing and the utilization rate of adjacent replicas, using 1-Greedy-Insert or others as compared algorithms, the simulation results show that each metric improves by 30% on average, which verifies the effectiveness of the proposed algorithm. ©, 2014, Editorial Board of Journal on Communications. All right reserved.
... Further, it is a major component in network systems to accomplish scalability. Load balancing describes the efficient process that describes the requirement of an application's processing across a different number of servers and is the main concern for all distributed systems; in fact, the scalability problem in client/server MMOGs is in essence related to load balancing [16]. However, when applying load balancing across MMOGs infrastructure, it must be implemented in an effective way to prevent essential resources over-supplying on the server side. ...
In recent years, there has been an important growth of online gaming. Today's Massively Multiplayer Online Games (MMOGs) can contain millions of synchronous players spread across the world and participating with each other within a single shared game. The increasing number of players in Massively Multiplayer Online Games has led to some issues with the demand on server which generates a significant increase in costs for game industry and impacts the quality of service suffered by players. With the number of players gradually increasing, servers still need to work efficiently under heavy load and, new research is required to improve the established MMOG system architectures. In this paper, we survey load balancing techniques based on both client-server and peer-to-peer system to evaluate the current state of the art in MMOG scalability and to discuss the open issues. In addition, we proposed a new technique for dynamic load balancing based on hybrid peer-to-peer overlay architecture to deploy the players among the region, managing nodes joining and leaving operations, managing the nodes migration from a region to another, and managing the creation of the region in the virtual environment.
... The work in [3] is also evaluate system performance in peer to peer network, and define Variable Chunk based switching scheme in which the file to be downloaded is divided into many chunks, but in this scheme a user downloads the chunks sequentially one at a time. Early work [5], [6] are study on clustering technique for average download time in peer to peer networks. The approach of using the average service capacity to analyze the average download time has been a common practice in the literature [7], [8], [9],[10]. ...
... But it is very complex to organize the original data distribution since every user has the liberty to split any data with additional users. Cluster based imitation architecture [2] is offered for load-balancing in peer-to-peer contented distribution systems. As well as an intellectual replica position technique, it also comprises of an efficient load balancing technique. ...
To distribute the computer resources like storage, content, CPU cycles openly without employing an intermediary system or a central server, dispersed computer architecture, called "peer-to peer" are considered. Preserving load balance on heterogeneous peer to peer networks is a challenging purpose and several existing works presented state balancing algorithm for load balance demand factor. The previous work built Time Variant-Peer Node Heterogeneous Data Processing Scheme for efficient load balancing in distributed heterogeneous peer networks. The degree of load-demand balance factors of peer servers is calculated by using Dynamic Time Warping (DTW) algorithm. The previous work described only about the liveliness of the peer node in the heterogeneous network. To improve the previous study, in this work, we are going to present with the resultant load-demand factor obtained from DTW, peer server giving out cycle requirements are identified using Duty Cycle Data Appropriation (DCDA) Technique. During the active status of peer node server, the processing cycle requirements of the servers are analyzed and it share the data with other nodes based on its duty cycle. If load imbalance occurs, Load diversions are made with peer node server with Node Selection Strategy based on DCDA rank representation. An experimental evaluation is conducted using the Heterogeneous peer networks data extracted from the large internet service providers to estimate the performance of the proposed heterogeneous peer node server cycle analyzer using DCDA technique and Node selection strategy [HPSC-DCDA & NSS] with an existing time variant P2P network using Dynamic Time Warping algorithm in terms of number of duty cycles, energy consumption, load diversity.
... We discussed the fault tolerance and algorithmic extensions for the proposed architecture. Our work is ongoing and we are planning to implement other distributed system middleware functions such as load balancing [45][46][47] on top of this layered architecture. ...
We propose a distributed mutual exclusion algorithm for mobile ad hoc networks. This algorithm requires a ring of cluster coordinators as the underlying topology. The topology is built by first providing clusters of mobile nodes in the first step and then forming a backbone consisting of the cluster heads in a ring as the second step. The modified version of the Ricart-Agrawala Algorithm on top of this topologyprovides analytically and experimentally an order of decrease in message complexity with respect to the original algorithm. We analyze the algorithm, provide performance results of the implementation, discuss the fault tolerance and the other algorithmic extensions, and show that this architecture can be used for other middleware functions in mobile networks.
... But it is very complex to organize the original data distribution since every user has the liberty to split any data with additional users. Cluster based imitation architecture [2] is offered for load-balancing in peer-to-peer contented distribution systems. As well as an intellectual replica position technique, it also comprises of an efficient load balancing technique. ...
To distribute the computer resources like storage, content, CPU cycles openly without employing an intermediary system or a central server, dispersed computer architecture, called " peer-to peer " are considered. Preserving load balance on heterogeneous peer to peer networks is a challenging purpose and several existing works presented state balancing algorithm for load balance demand factor. The previous work built Time Variant-Peer Node Heterogeneous Data Processing Scheme for efficient load balancing in distributed heterogeneous peer networks. The degree of load-demand balance factors of peer servers is calculated by using Dynamic Time Warping (DTW) algorithm. The previous work described only about the liveliness of the peer node in the heterogeneous network. To improve the previous study, in this work, we are going to present with the resultant load-demand factor obtained from DTW, peer server giving out cycle requirements are identified using Duty Cycle Data Appropriation (DCDA) Technique. During the active status of peer node server, the processing cycle requirements of the servers are analyzed and it share the data with other nodes based on its duty cycle. If load imbalance occurs, Load diversions are made with peer node server with Node Selection Strategy based on DCDA rank representation. An experimental evaluation is conducted using the Heterogeneous peer networks data extracted from the large internet service providers to estimate the performance of the proposed heterogeneous peer node server cycle analyzer using DCDA technique and Node selection strategy [HPSC-DCDA & Heterogeneous Peer Server Node Cycle Analyzer using DCDA Technique with Load Diversion Strategy 120 NSS] with an existing time variant P2P network using Dynamic Time Warping algorithm in terms of number of duty cycles, energy consumption, load diversity.
... In [1], is presented the cluster based replication architecture for load-balancing in peer-to-peer content distribution systems using an intelligent replica placement technique. ...
The allocation planning of resources in a web server cluster is accomplished nowadays by the administrator. Once the internet is quite dynamic, as far as the use of resources is concerned, such a task may be considered critical and inefficient if accomplished manually. Our approach benefits from the use of agents to learn from the environment and adjust automatically the behavior of the system to make a better use of the available resources. With this approach it is possible to help the administrator by minimizing your stress in moments of work overload. The conception, specification, adopted allocation planning strategy, modeling in Petri nets, implementation of this platform in the Java language are presented. Experimentations and simulations which prove the efficiency of the proposal are presented.
