• ETRI Journal
• /
• 제15권3_4호
• /
• pp.35-45
• /
• 1994

In this paper we develop an approximation formalism on the queue length distribution for general queueing models. Our formalism is based on two steps of approximation; the first step is to find a lower bound on the exact formula, and subsequently the Chernoff upper bound technique is applied to this lower bound. We demonstrate that for the M/M/1 model our formula is equivalent to the exact solution. For the D/M/1 queue, we find an extremely tight lower bound below the exact formula. On the other hand, our approach shows a tight upper bound on the exact distribution for both the ND/D/1 and M/D/1 queues. We also consider the $M+{\Sigma}N_jD/D/1$ queue and compare our formula with other formalisms for the $M+{\Sigma}N_jD/D/1$ and M+D/D/1 queues.

• 전자공학회논문지A
• /
• 제31A권2호
• /
• pp.9-17
• /
• 1994

This paper proposes Double queue threshold QLT(Queue Length Threshold) algorithm and Hysteresis effect QLT algorithm. as being DPS(Dynamic Priority Scheduling) techniques. in order to advance the processing of multiple class traffics. Also, the performance of the proposed algorithms is analyzed through computer simulations,and the priority scheduling is analyzed using a retrial queue with two types of calls. Our simulation results show that the performance of the proposed Double queue length threshold QLT algorithm is superior to that of the conventinal QLT algorithm for 2 or more classes delay sensitive traffics. Also we find that Hysteresis effecT QLT algorithm is better mechanism than that of the existing QLT for real time and non-real time traffics.

• 디지털산업정보학회논문지
• /
• 제6권2호
• /
• pp.153-167
• /
• 2010

Traffic conditions, in particular number of active TCP flows, change dramatically over time. The main goal of this paper is an adaptive queue management algorithm that can maintain network state of high-throughput and low-delay under changing traffic conditions In this paper, we devise Probability Adaptive RED(PARED) that combines the more effective elements of recent algorithms with a RED core. It automatically adjusts its adaptive marking function to account for changes in traffic load and to keep queue length within the desired target queue length. We simulate that PARED algorithm results in under changes in traffic load and mixed traffic load. The simulation test confirm this stability, and indicate that overall performances of PARED are substantially better than the RED and ARED algorithms.

• Journal of the Korean Statistical Society
• /
• 제12권2호
• /
• pp.95-99
• /
• 1983

Limiting law for maximum queue length in G/M/1 queue system is derived. Furthermore, a simple approach using mixing sequence is also discussed.

• Industrial Engineering and Management Systems
• /
• 제7권2호
• /
• pp.143-149
• /
• 2008

There have been some approximation analysis methods for a GI/G/1 queueing system. As one of them, an approximation technique for the steady-state probability in the GI/G/1 queueing system based on the iteration numerical calculation has been proposed. As another one, an approximation formula of the average queue length in the GI/G/1 queueing system by using the diffusion approximation or the heuristics extended diffusion approximation has been developed. In this article, an approximation technique in order to analyze the GI/G/1 queueing system is considered and then the formulae of both the steady-state probability and the average queue length in the GI/G/1 queueing system are proposed. Through some numerical examples by the proposed technique, the existing approximation methods, and the Monte Carlo simulation, the effectiveness of the proposed approximation technique is verified.

• 대한수학회보
• /
• 제43권3호
• /
• pp.657-669
• /
• 2006

In this paper, we analyze a leaky bucket (LB) scheme with queue length dependent arrival rates. In other words, if the queue length exceeds an appropriate threshold value on buffer, the arrivals need to be controlled. In ATM networks, if the congestion occurs, the input traffics must be controlled (reduced) for congestion resolution. By the bursty and correlated properties of traffics, the arrivals are assumed to follow a Markov-modulated Poisson process (MMPP). We derive the loss probability and the waiting time distribution for arbitrary cell. The analysis is done by using the embedded Markov chain and supplementary variable method. We also present some numerical examples to show the effects of our proposed LB scheme.

• 한국통신학회논문지
• /
• 제22권5호
• /
• pp.1018-1025
• /
• 1997

In this paper, we propose the interworking and emulation methods for the interoperability of ABR/ABT capability and algo propose the BCR Re-negotition with State-Dependent (BRSD) mechanism based on the common buffer with two thresholds and the EBCN method when public network hs the ABT capability. As BRSD mechansism has one common buffer for all ABR connectiona and BCR re-negotiation request according to the threshold, it can reduce the queue length and the frequency of ABT BCR re-negotiation. Using EBCN method, BRSD mechanism can prevent the continuous increasing of queue length in the emulation method with the CI bit setting of the ABR performance of the BRSD mechanism in the view poits of the masimum queue length, bandwidth utilization and ABT-BCR re-negotialtion frequency trough the simulation.

• 대한산업공학회지
• /
• 제28권1호
• /
• pp.36-43
• /
• 2002

We analyze bulk service $M/G^{b}/1$ queues using the arrival time approach of Chae et al. (2001). As a result, the decomposition property of the M/G/1 queue with generalized vacations is extended to the $M/G^{b}/1$ queue in which the batch size is exactly a constant b. We also demonstrate that the arrival time approach is useful for relating the time-average queue length PGF to that of the departure time, both for the $M/G^{b}/1$queue in which the batch size is as big as possible but up to the maximum of constant b. The case that the batch size is a random variable is also briefly mentioned.

• 한국ITS학회 논문지
• /
• 제4권2호
• /
• pp.13-22
• /
• 2005

본 연구는 대기행렬길이를 보다 정확하게 측정하는데 목적을 두고 있다. 현재 서울시 실시간 신호제어시스템(COSMOS)에서 적용되는 대기행렬산출 알고리즘은 평균차량길이와 점유시간을 이용하여 속도를 산출하고 이를 다시 밀도 개념의 정체도로 환산하여 대기행렬길이를 구하고 있다. 평균차량길이에 의한 속도산출로 인하여 오차가 발생되고, 운영자가 지정해 주어야하는 값이 산재되어 그 값을 최적화하기가 어려워 정확한 대기행렬길이를 구하기가 어렵다. 따라서 본 연구에서는 정지선으로부터의 거리와 대기검지기에서부터 수집되는 점유율 간의 관계를 밝혀, 이를 대기행렬길이 산출 알고리즘에 적용하였다. 대기행렬길이 추정 알고리즘을 사용할 경우 루프 검지기에서 많은 가공단계가 필요하지 않은 기본 데이터인 점유율을 이용하여 대기행렬길이를 예측할 수 있다고 할 수가 있다. 그리고 운영자 지정값이 없어 현장 적용에 있어서 그 값에 대한 최적화 작업이 필요가 없어 쉽게 적용할 수 있는 장점이 있다. 개발된 대기행렬길이 추정알고리즘을 COSMOS시스템이 운영중인 사이트(Site)에 적용할 결과 현재 사용 중인 대기행렬길이 추정 알고리즘 보다 우수한 것을 확인 할 수가 있었다. 하지만, 본 연구에서 개발된 대기행렬길이 추정 알고리즘을 일반적으로 적용하기 위해서는 다양한 현장 및 경우에 대하여 적용하여 검증을 하여야 할 것이다. 이에 관련한 연구가 향후 진행될 경우 현재 대기행렬길이 추정 알고리즘 보다 적용이 쉽고, 정확한 값을 얻는 알고리즘으로 완성될 것으로 기대한다.

• 한국통신학회논문지
• /
• 제18권2호
• /
• pp.279-287
• /
• 1993

본 논문은 다중등급의 통화유량에 대해 효율적인 처리를 위하여 동적 우선순위 스케쥴링이 가능한 2중이완성 한계치를 갖는 최소이완성한계 (Minimum Laxity Threshold:MLT)알고리즘과 2중 대기행열 한계치를 갖는 대기 행열간격한계(Queue Length Threshold QLT) 알고리즘을 제안하였다. 그리고 제안한 알고리즘의 성능을 컴퓨터 시뮬레이션을 통하여 분석하였다. 시뮬레이션 결과, 실시간서비스 중 지연민감도가 다른 2개 등급 이상의 통화유량 처리시2중 이완성 한계치를 갖는 MLT알고리즘을 적용함으로써 MLT알고리즘 자체보다 처리 효율성이 개선됨을 확인하고, 또2개 등급 이상의 비실시간 통화유량에 대해서는 2중 대기행열 한계치를 갖는 QLT알고리즘이 QLT보다 우수함을 알 수 있었다.