• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제12권11호
• /
• pp.5234-5251
• /
• 2018

It is well-known that virtualization technology can bring many benefits not only to users but also to service providers. From the view of system security and resource utility, higher resource sharing degree and higher system reliability can be obtained by the introduction of virtualization technology in distributed cloud. The small size time-sharing multiplexing technology which is based on virtual machine in distributed cloud platform can enhance the resource utilization effectively by server consolidation. In this paper, the concept of memory block and user satisfaction is redefined combined with user requirements. According to the unbalanced memory resource states and user preference requirements in multi-virtual machine environments, a model of proper memory resource allocation is proposed combined with memory block and user satisfaction, and at the same time a memory optimization allocation algorithm is proposed which is based on virtual memory block, makespan and user satisfaction under the premise of an orderly physical nodes states also. In the algorithm, a memory optimal problem can be transformed into a resource workload balance problem. All the virtual machine tasks are simulated in Cloudsim platform. And the experimental results show that the problem of virtual machine memory resource allocation can be solved flexibly and efficiently.

• 정보처리학회논문지A
• /
• 제11A권4호
• /
• pp.251-256
• /
• 2004

고성능 컴퓨터들의 대량 보급으로 인해 다양한 종류의 분산 자원이 주위에 산재해 있다. 이러한 분산 자원을 이용하는데 각 분야별 시스템들의 개념, 모듈, 운영 및 관리를 배제하고 사용할 수 있다면 최소한의 비용으로 분산 자원을 사용할 수 있을 젓이다. 이에 '분산환경에 적용될수 있는 DVM 및 DESPL의 제안'을 통해 제안된 내용중 그 일부를 수용하여 SDVM을 설계 및 구현한다. SDVM을 이용하는 사용자는 분산 환경의 개념없이 최소한의 비용으로 분산된 자원을 사용할 수 있다.

• Journal of Communications and Networks
• /
• 제12권5호
• /
• pp.449-458
• /
• 2010

In this paper, we present a distributed resource allocation algorithm for multi-cell uplink systems that increases the weighted sum of the average data rates over the entire network under the average transmit power constraint of each mobile station. For the distributed operation, we arrange each base station (BS) to allocate the resource such that its own utility gets maximized in a noncooperative way. We define the utility such that it incorporates both the weighted sum of the average rates in each cell and the induced interference to other cells, which helps to instigate implicit cooperation among the cells. Since the data rates of different cells are coupled through inter-cell interferences, the resource allocation taken by each BS evolves over iterations. We establish that the resource allocation converges to a unique fixed point under reasonable assumptions. We demonstrate through computer simulations that the proposed algorithm can improve the weighted sum of the average rates substantially without requiring any coordination among the base stations.

• 한국정보통신학회논문지
• /
• 제5권7호
• /
• pp.1250-1256
• /
• 2001

본 논문에서는 Ad Hoc 무선 LAN에서의 자원 관리방식을 제안하였다. 이 방식은 분산적인 자원 관리방식과 직교주파수분할 다중화-부호분할 다중화(Orthogonal Frequency Division Multiplexing-Code Division Multiple Access, OFDM-CDMA) 기술을 결합하여 QoS를 보장하는 멀티미디어 서비스를 지원한다. 이 방식에 대하여 호 거절률(Blocking Rate)과 QoS 손실확률(QoS Loss Probability)의 관점에서 성능을 분석하였다.

• 한국정보통신학회논문지
• /
• 제9권5호
• /
• pp.965-972
• /
• 2005

본 논문에서는 멀티미디어 Ad Hoc 무선 LAM(AWLAN)을 위한 새로운 자원 관리 방안으로 분산채널 할당 프로토롤(Distributed Channel Allocation Protocol, DCAP)을 제안한다 이 방안은 QoS (Quality-of-Service)를 보장하는 분산적 자원 관리를 OFDM-CDMA(Orthogonal Frequency Division Multiplexing-Code Division Multiple Access) 채널 구조 위에 구현한 것이다. MPEG 트래픽에 대한 성능평가 결과 DCAP는 광대역 무선 접근(Broadband Wireless Access, BWA) 형 물리구조 기반에서의 멀티미디어 서비스를 지원하는 AWLAN을 위한 좋은 자원 관리 방안임이 판명되었다.

• International journal of advanced smart convergence
• /
• 제13권3호
• /
• pp.143-149
• /
• 2024

As the size of data and models in machine learning training continues to grow, training on a single server is becoming increasingly challenging. Consequently, the importance of distributed machine learning, which distributes computational loads across multiple machines, is becoming more prominent. However, several unresolved issues remain regarding the performance enhancement of distributed machine learning, including communication overhead, inter-node synchronization challenges, data imbalance and bias, as well as resource management and scheduling. In this paper, we propose ParamHub, which utilizes orchestration to accelerate training speed. This system monitors the performance of each node after the first iteration and reallocates resources to slow nodes, thereby speeding up the training process. This approach ensures that resources are appropriately allocated to nodes in need, maximizing the overall efficiency of resource utilization and enabling all nodes to perform tasks uniformly, resulting in a faster training speed overall. Furthermore, this method enhances the system's scalability and flexibility, allowing for effective application in clusters of various sizes.

• Transactions on Control, Automation and Systems Engineering
• /
• 제3권3호
• /
• pp.181-189
• /
• 2001

The software components of embedded control systems get extremely complex as they are designed into distributed systems get extremely complex as they are designed into distributed systems consisting of a large number of inexpensive microcontrollers interconnected by low-bandwidth real-time networks such as the controller area network (CAN). While recently emerging middleware technologies such as CORBA and DCOM address the complexity of distributed programming, they cannot be directly applied to distributed control system design due to their excessive resource demand and inadequate communication models. In this paper, we propose a CORBA-based middleware design for CAN-based distributed embedded control systems. Our design goal is to minimize its resource need and make it support group communication without losing the IDL (interface definition language) level compliance to the OMG standards. To achieve this, we develop a transport protocol on the CAN and a group communication scheme based on the well-known publisher/subscriber model. The protocol effectively realizes subject-based addressing and supports anonymous publisher/subscriber communication. We also customize the method invocation and message passing protocol, referred to as the general inter-ORB protocol (GIOP), of CORBA so that CORBA method invocations are efficiently serviced on a low-bandwidth network such as the CAN. This customization includes packed data encoding and variable-length integer encoding for compact representation of IDL data types. We have implemented our CORBA-based middleware on the mArx real-time operating system we have developed at Seoul National University. Our experiments clearly demonstrate that it is feasible to use CORBA in developing distributed embedded control systems possessing severe resource limitations. Our design clearly demonstrates that it is feasible to use a CORBA-based middleware in developing distributed embedded systems on real-time networks possessing severe resource limitations.

• Journal of Communications and Networks
• /
• 제11권4호
• /
• pp.368-374
• /
• 2009

This paper focuses on a distributed multiple spectrum pricing scheme to maximize system capacity in next generation multiaccess systems, where multimode user equipments (MUEs) can connect simultaneously to multiple base stations (BSs) with multiple radio access technologies (RATs). The multi-price based scheme provides a distributed decision making for an optimal solution where radio resource allocations are determined by each MUE, unlike most centralized mechanisms where BS controls the whole radio resource. By the proposed optimal solution, MUEs can decide their share of spectrum bands and power allocation according to the spectrum price of each RAT, and at the same time the multiaccess system can achieve maximized total throughput. Numerical analysis shows that the proposed scheme achieves the maximal capacity by distributed resource allocation for the multiaccess system.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1995년도 추계학술대회 논문집 학회본부
• /
• pp.219-221
• /
• 1995

This paper proposes a model for simulation and performance evaluation of distributed computer systems(DCS). The model is composed of operating system(OS), resource, task, environment submodel. Task Flow Graph(TFG) is suggested to describe the relation between tasks. This paper considers task response time, the scheduler's ready queue length, utilization of each resource as performance indices. The distributed system of Continuous Annealing Line(CAL) in iron process is simulated with the proposed model.

• 한국멀티미디어학회논문지
• /
• 제15권10호
• /
• pp.1212-1220
• /
• 2012

In this paper, a collision-free resource reservation scheme for WUSB (Wireless Universal Serial Bus) networks with WiMedia D-MAC (Distributed Medium Access Control). Since distributed characteristic of the WiMedia D-MAC supporting DRP (Distributed Reservation Protocol) scheme may cause lots of conflicts, overall performances of the WUSB with WiMedia D-MAC can be deteriorated. In addition, when we consider multi-hop environment, "mobile" hidden node problem due to mobility of WUSB devices can be happened, and it is a critical problem to mobile WUSB devices transceiving real-time QoS (Quality of Service) traffic. To tackle the problem, we propose a new DRP reservation mechanism to prevent DRP conflicts for multi-hop mobility support in WUSB networks with WiMedia D-MAC and show its improved performance via simulation results.