• 산업경영시스템학회지
• /
• 제31권4호
• /
• pp.86-92
• /
• 2008

The expected busy period for the controllable M/G/1 queueing model operating under the triadic Max (N, T, D) policy is derived by using a new concept so called "the pseudo probability density function." In order to justify the proposed approaches for the triadic policy, well-known expected busy periods for the dyadic policies are recovered from the obtained result as special cases.

• 한국시뮬레이션학회논문지
• /
• 제27권3호
• /
• pp.45-52
• /
• 2018

대기행렬에 고객 (또는 패킷 등)이 몰리는 오버로드(overload)가 발생하는 경우 긴 대기열이 발생하여 서비스 품질에 좋지 않은 영향을 줄 수 있다. 오버로드 상황에서 혼잡을 완화하기 위해 대기하는 고객숫자에 기반한 다양한 오버로드 제어 정책들이 고안, 적용되고 있다. 본 연구는 대기 중인 고객 숫자에 한계점 (threshold)을 두고, 한계점을 넘으면 서비스 속도를 빠르게 하거나 고객의 도착 간격(시간)을 증가시키는 제어정책을 대상으로 한다. 이러한 정책을 갖는 M/G/1 대기행렬에 대해 바쁜 기간(busy period)을 분석하는데, 연구결과는 비용구조가 주어졌을 때 최적 시스템 제어 정책을 찾는데 필수적이다.

• Communications for Statistical Applications and Methods
• /
• 제7권2호
• /
• pp.489-497
• /
• 2000

The M/M/1 queue with impatient customers is studied. Impatient customers wait for service only for limited time K/0 and leave the system if their services do not start during that time. Notice that in the analysis of virtual waiting time, the impatient customer can be considered as the customer who enters the system only when his/her waiting time does not exceed K. In this paper, we apply martingale methods to the virtual waiting time and obtain the expected period from origin to the point where the virtual waiting time crosses over K or reaches 0, and the variance of this period. With this results, we obtain the expected busy period of the queue, the distribution, expectation and variance of the number of times the virtual waiting time exceeding level K during a busy period, and the probability of there being no impatient customers in a busy period.

• 산업경영시스템학회지
• /
• 제34권1호
• /
• pp.67-73
• /
• 2011

Using the known result of the expected busy period for a controllable M/G/1 queueing model operating under the triadic Max (N, T, D) policy, its upper and lower bounds are derived to approximate its corresponding actual value. Both bounds are represented in terms of the expected busy periods for the dyadic Min (N, T), Min (N, D) and Min (T, D) and simple N, T and D operating policies. All three input variables N, T and D are equally contributed to construct such bounds for better estimation.

• 산업경영시스템학회지
• /
• 제33권2호
• /
• pp.97-104
• /
• 2010

Using the known result of the expected busy period for the triadic Min (N, T, D) operating policy applied to a controllable M/G/1 queueing model, its upper and lower bounds are derived to approximate its corresponding actual value. Both bounds are represented in terms of the expected busy periods for the dyadic Min (N, T), Min (N, D) and Min (T, D) and simple N, T and D operating policies. All three input variables N, T and D are equally contributed to construct such bounds for better approximations.

• 한국경영과학회지
• /
• 제17권2호
• /
• pp.123-130
• /
• 1992

The model of k-out-of n : G repairable system with identical components is extended to a repairable system with n different components. The objective is to analytically derive the mean time of the busy period for a k-out-of-n : G system with unrestricted repair. Then, the lower and upper bounds on the average time of the busy period of the n-component system with restricted repair are also shown.

• 산업경영시스템학회지
• /
• 제34권4호
• /
• pp.162-168
• /
• 2011

The most generalized form of the triadic operating policy for a controllable M/G/I queueing model is analyzed to obtain fundamental relations among the other forms of operating policies based on its corresponding expected busy period. Since it consists of three decision variables N, T and D, it could be possible to decompose into the simple, the dyadic and other forms of the triadic operating policies. The procedures to decompose the most generalized triadic policy into other forms of operating policies could provide a general methodology to identify each element associated with it.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제8권7호
• /
• pp.2512-2531
• /
• 2014

We propose a new Distributed Denial of Service (DDoS) defense mechanism that protects http web servers from application-level DDoS attacks based on the two methodologies: whitelist-based admission control and busy period-based attack flow detection. The attack flow detection mechanism detects attach flows based on the symptom or stress at the server, since it is getting more difficult to identify bad flows only based on the incoming traffic patterns. The stress is measured by the time interval during which a given client makes the server busy, referred to as a client-induced server busy period (CSBP). We also need to protect the servers from a sudden surge of attack flows even before the malicious flows are identified by the attack flow detection mechanism. Thus, we use whitelist-based admission control mechanism additionally to control the load on the servers. We evaluate the performance of the proposed scheme via simulation and experiment. The simulation results show that our defense system can mitigate DDoS attacks effectively even under a large number of attack flows, on the order of thousands, and the experiment results show that our defense system deployed on a linux machine is sufficiently lightweight to handle packets arriving at a rate close to the link rate.

• 산업경영시스템학회지
• /
• 제33권3호
• /
• pp.63-70
• /
• 2010

Based on the known results of the expected busy periods for the triadic Min (N, T, D) and Max (N, T, D) operating policies applied to a controllable M/G/1 queueing model, a relation between them is constructed. Such relation is represented in terms of the expected busy periods for the simple N, T and D, and the dyadic Min (N, T), Min (T, D) and Min (N, D) operating policies. Hence, if any system characteristics for one of the two triadic operating policies are known, unknown corresponding system characteristics for the other triadic operating policy could be obtained easily from the constructed relation.

• 산업경영시스템학회지
• /
• 제38권1호
• /
• pp.21-29
• /
• 2015

A steady-state controllable M/G/1 queueing model operating under the {T:Min(T,N)} policy is considered where the {T:Min(T,N)} policy is defined as the next busy period will be initiated either after T time units elapsed from the end of the previous busy period if at least one customer arrives at the system during that time period, or after T time units elapsed without a customer' arrival, the time instant when Nth customer arrives at the system or T time units elapsed with at least one customer arrives at the system whichever comes first. After deriving the necessary system characteristics including the expected number of customers in the system, the expected length of busy period and so on, the total expected cost function per unit time for the system operation is constructed to determine the optimal operating policy. To do so, the cost elements associated with such system characteristics including the customers' waiting cost in the system and the server's removal and activating cost are defined. Then, procedures to determine the optimal values of the decision variables included in the operating policy are provided based on minimizing the total expected cost function per unit time to operate the queueing system under considerations.