• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.29 no.3A
• /
• pp.223-232
• /
• 2004

In this paper we propose the hierarchical-based dynamic bandwidth allocation algorithm for multi-class services in Ethernet-PON. The proposed algorithm consists of the high level scheduler in OLT and the low level scheduler in ONU. The hierarchical architecture is able to provide scalability and resource efficiency in Ethernet-PON which has the distributed nature of the scheduling domain, with queues and the scheduler located at a large distance from each other. We also propose three dynamic bandwidth allocation algorithms for the low level scheduler: Proportional Allocation algorithm, Maximum Request First Allocation (MRFA) algorithm and High Priority First Allocation (HPFA) algorithm. We implement the Ethernet-PON standardized in the IEEE 802,3ah using OPNET. We also evaluate and analyze the performance for the proposed algorithms in terms of channel utilization, queuing delay and the amount of remainder.

• Journal of KIISE:Information Networking
• /
• v.32 no.2
• /
• pp.254-260
• /
• 2005

In the guaranteed service, a real-time scheduling algorithm must achieve both high level of network utilization and scalable implementation. Here, network utilization indicates the number of admitted real-time sessions. Unfortunately, existing scheduling algorithms either are lack of scalable implementation or can achieve low network utilization. For example, scheduling algorithms based on time-stamps have the problem of O(log N) scheduling complexity where N is the number of sessions. On the contrary, round-robin algorithms require O(1) complexity. but can achieve just a low level of network utilization. In this paper, we propose a scheduling algorithm that can achieve high network utilization without losing scalability. The proposed algorithm is a Hierarchical Round-Robin (H-RR) algorithm that utilizes multiple rounds with different interval sizes. It provides latency bounds similar to those by Packet-by-Packet Generalized Processor Sharing (PGPS) algorithm using a sorted-Priority queue. However, H-RR requires a constant time for implementation.

• Journal of KIISE:Computing Practices and Letters
• /
• v.7 no.5
• /
• pp.539-547
• /
• 2001

Cluster systems employ high speed interconnection networks and use efficient communication layers to gain high performance and scalability. But the diversity in implementation mechanism among these communication layers causes lack of portability. A solution is to provide communication standard APIs such as MPI. This paper introduces MPI-VMMC: an MPI implementation on VMMC. Though the direct deposit transfer mechanism used in VMMC is not suitable for Send/Recv mechanism used in MPI, the proposed sub-layer laid between MPI and VMMC efficiently translates from one mechanism to the other. We also use the lazy pointer and selective zero-copy transfer technique to gain high performance. The peak performance of MPI-VMMC is 90.7Mbytes/sec, which is about 95% of the base communication layer\`s.

• Journal of the Korea Society of Computer and Information
• /
• v.15 no.1
• /
• pp.129-137
• /
• 2010

In this paper, we propose a packet prioritization scheme using preamble signal and backoff time which is designed to provide a differentiated channel occupation with a packet priority in wireless sensor networks. This scheme aims at enabling QoS such as fast delivery of high priority packets by reducing their backoff time as well as thus securing higher channel occupation. We expect that it could also improve the channel utilization of the entire network by avoiding unnecessary channel contention. For the purpose, we add new features of multiple queue and preamble modification to the TinyOS based B-MAC. This scheme achieves 82-88% reduction in delivery time of high priority packets, thus it enables realtime support for urgent applications.

• Journal of the Korean Institute of Illuminating and Electrical Installation Engineers
• /
• v.22 no.2
• /
• pp.134-141
• /
• 2008

This paper presents an innovative load management technique that can effectively lower the summer peak load by adjusting the aircondition loads through smoothe coordinations between utility companies and large customers. An intelligent hierarchical load management system composed of a Central Intelligent Load Management System(CIMS) and multiple Local Intelligent Management Systems(LIMS) is also proposed to implement the reposed technique. Upon receiving a load curtailment request from the utilities, CIMS issues tokens, which can be used by each LIMS as a right to turn on the airconditioner. CIMS creates and maintains a queue for fair allocation of the tokens among the LIMS demanding tokens. By adjusting the number tokens and queue management Policies, desired load factors can be achieved conveniently. The Markov Birth and Death Process and the Balance Equations are employed in estimating various queue performances. The proposed technique is tested using a summer load data of a large apartment complex and proved to be quite effective in load management while minimizing the customer inconveniences.

• Journal of KIISE:Information Networking
• /
• v.27 no.2
• /
• pp.206-218
• /
• 2000

Retransmissions on the DLC layer are essential to ABR service providing the low CLR (cell loss ratio) over the unreliable wireless channel with high bit error rate. In the wireless ATM, the DLC layer below ATM layer performs the retransmission and reordering of the cells to recover the cell loss over the wireless channel and by doing so, the effect of the wireless channel characteristics with high bit error rate can be minimized on the ATM layer which is designed under the assumption of the low bit error rate. We propose, in this paper, the schemes to reflect the changes of the transmission rate over the wireless channel on the ABR rate control. Proposed scheme can control the source rate to the changes of the transmission rate over the wireless channel and reduce the required buffer size in the AP (access point). In the simulation, we assume that the DLC layer can inform the ATM layer of the wireless channel quality as good or bad. Our simulation results show that the proposed schemes require the smaller buffer size compared with the existing scheme, enhanced dynamic max rate control algorithm (EDMRCA). It is also shown that the scheme with the intelligent DLC which adjusts the rate to the wireless channel quality not only provides the low CLR with smaller buffer requirement but also improves the throughput by utilizing the wireless bandwidth more efficiently.

• Proceedings of the Korean Information Science Society Conference
• /
• 2007.06d
• /
• pp.477-482
• /
• 2007

최근 다양한 P2P 프로그램을 많이 사용함에 따라 네트워크에서 생겨나는 트래픽의 상당 부분이 P2P가 발생시키는 트래픽으로 이미 HTTP, FTP의 양을 훨씬 뛰어넘고 있다. 현재 인터넷 환경에서 방화벽을 통과하기 위해 포트번호를 변경하여 통신을 하는 새로운 P2P응용들의 행동들은 전통적인 well-known port 기반의 응용프로그램을 구분하는 단순한 분석 방법만으로 신뢰하기가 어렵다. 새로운 P2P 응용들과 같은 트래픽 모니터링의 정확도를 높이기 위해서는 TCP/IP 헤더만이 아니라 패킷이 담고 있는 페이로드 내용에 대한 조사 차원의 모니터링 방법이 필요하다. 본 논문에서는 TCP/IP 헤더 정보와 더불어 패킷의 페이로드 내용을 조사하여 P2P 트래픽을 탐지하는 모니터링 기법을 제안한다. 이어 탐지되는 P2P 트래픽에 대하여 Linux Netfilter Framework의 Queuing Discipline에서 제공하는 계층적인 우선순위 큐를 사용하여 일정한 양의 대역폭을 할당하는 정책을 적용함으로써 안정적이면서 효율적인 네트워크 운용 방안을 제시한다.

• The Journal of the Institute of Internet, Broadcasting and Communication
• /
• v.18 no.1
• /
• pp.15-21
• /
• 2018

Per-packet scheduling is more suitable than per-flow scheduling to reduce the flow completion time by efficiently utilizing resources in data center networks. Recently, many per-packet scheduling schemes utilizing multiple paths have been proposed. However, to mitigate the negative effect of packet reordering on TCP performance, most of the schemes require supplemental measures such as putting packets in order at the lower layer. In this study, we investigate how well DCTCP, which is a representative TCP for data center networks, performs with per-packet scheduling through simulation. Simulation results show that DCTCP keeps the queue length short but that DCTCP shows low fairness due to the way of reducing the congestion window by ECN.

• Journal of Korea Multimedia Society
• /
• v.4 no.4
• /
• pp.339-348
• /
• 2001

In this paper, a cell scheduling algorithm which can be applied to all of the service class in ATM network is proposed. The proposed algorithm classifies the order of priority in each service class into 4 categories and generates the weight of each class service based on the traffic parameters which are negotiated in connection contract. The proposed algorithm guarantees QoS(Quality of Service) to the traffic which is sensitive to delay carrying out CBR and rt_VBR service. As it effectively manages the connection which has small bandwidth, it minimizes the cell delay in the queue. For verifying the effectiveness of the proposed algorithm the proposed algorithm is simulated with existing cell scheduling algorithm and the result is showed.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2009.04a
• /
• pp.344-347
• /
• 2009

데이터 스트림을 처리하기 위한 연속집계질의 수행 시 중복연산 및 메모리의 절약을 위하여 큐를 공유하는 자원공유기법이 연구되었다. 기존의 자원공유 기법들은 질의의 프리디킷이 일치할 때만 처리하기 때문에, 질의의 프리디킷이 차이가 나는 경우가 많은 다중공간 집계질의가 자주 요청되는 u-GIS 환경에서 효율적으로 중복영역을 처리할 수 있는 자원공유 기법이 요구된다. 본 논문에서는 공간영역을 효율적으로 그룹화하는 R-tree 의 특징을 이용하여 질의간의 중복영역을 그룹화하고 중복영역의 자원을 패인(Pane)구조를 이용하여 공유한다. 노드 수에 제한이 없고 레벨을 1로 하는 R-tree 로 유사한 위치의 질의들을 그룹화 한 후, 그 질의들의 영역이 겹쳐지는 부분을 패인을 이용해 집계 값을 공유하여 중복계산을 피하는 방법이다. 제안 기법은 공간 집계질의를 처리할 수 있고, 기존의 계층구조의 자원공유 기법을 사용할 때에 비해 자원을 적게 사용하고 질의 처리 시간을 단축시켰다. 성능평가를 통하여 제안기법이 메모리 사용량을 감소시키는 것을 보였으며, 질의 처리 속도가 증가하였다.