• 한국정보처리학회논문지
• /
• 제6권8호
• /
• pp.2196-2203
• /
• 1999

본 논문은 데이터패스 합성에서의 버스 스케줄링 문제와 레지스터의 최적화 방법을 다룬 것이다. 스케줄링은 DFG(Data Flow Graph)의 연산을 제어스텝(control step)에 할당하는 과정으로서, 주어진 조건을 만족하는 범위 내에서 비용함수(cost function)의 최소화에 목적을 둔다. 이를 위해 본 논문에서는, 연산자 배치를 위한 하드웨어 할당(hardware allocation) 과정에서의 설계비용을 최소화시키기 위해, 연산결과를 저장하는 레지스터(register)와 연산간의 이동 통로인 버스(bus)의 최적화 기법을 논하였다. 특히, 하드웨어 할당과정의 중요한 과제인 버스와 레지스터의 최소화 기법을 논하였으며, 레지스터의 최적화는 스케줄링이 완료된 후의 결과를 이용하였다. 실험대상으로는 벤치마크 모델인 5차 디지털 웨이브필터(5th-order digital wave filter)를 사용하였으며, 본 논문의 결과를 기존결과와 비교함으로써, 본 논문의 효용성을 입증하였다. 모든 실험결과는 구조형태의 선형정수계획법(ILP : Integer Linear Programming)을 이용함으로써, 모든 경우에 언제나 최적의 결과를 얻을 수 있도록 하였다.

• 대한전자공학회논문지TC
• /
• 제49권7호
• /
• pp.23-31
• /
• 2012

스케쥴링 및 데이터율의 결합 제어와 같은 자원할당 기술은 전파인지 네트워크에서는 매우 중요한 문제이다. 그러나 전파인지 네트워크에서는 주사용자 채널의 스토케스틱 특성으로 인하여 데이터율 및 스케쥴링을 결합하여 제어하는 것은 매우 어렵다. 본 논문에서는 전파인지 네트워크에서 신뢰성 제한 조건들을 고려한 비대칭 데이터율 및 스케쥴링 결합 제어 기법을 제안한다. 데이터율 및 스케쥴링 문제를 컨벡스 최적화 기법으로 공식화하고 쌍대성 분해 기법을 사용하여 부분 문제로 변환하여 분산화 하였다. 본 논문에서는 전체 시스템의 효용함수를 최대화 하도록 분산 노드들의 데이터율을 분산적으로 제어하는 알고리즘을 제안 하였다. 반면, 스케줄링은 기지국이 최적화하는 비대칭 기법을 제안하였다. 본 논문에서 제안한 비대칭 결합 제어 알고리즘은 전체 최적화 해로 수렴하는 것을 수치해석 기법으로 검증하였다.

• 경영과학
• /
• 제29권3호
• /
• pp.183-192
• /
• 2012

For the past several decades, personnel scheduling and rostering problem has been one of the most popular research topics in optimization area. Among the numerous applications, airline (aviation) industry has been given most attention due to the economic scale and impact. Most of the literatures about the staff scheduling problem in airline industry are dealing with the air crew, pilots and flight attendances, and the rest of the literatures are about the ground staff, by whom cleaning, maintenance, fueling of aircraft and handling luggage are done from landing to taking off. None of the literatures found by the authors are dealing with the airline ground crew. In this paper roster of airline ground crew, who is responsible for issuing boarding pass, checking baggage, etc, is introduced, formulated and solved using CPLEX. Some expressions of the mathematical formulations, which are not suitable input format of the CPLEX, were transformed. Numerical examples are presented for the validation of proposed scheduling system.

• Journal of applied mathematics & informatics
• /
• 제18권1_2호
• /
• pp.383-391
• /
• 2005

The paper is devoted to some flow shop scheduling problems, where job processing times are defined by functions dependent on their positions in the schedule. An example is constructed to show that the classical Johnson's rule is not the optimal solution for two different models of the two-machine flow shop scheduling to minimize makespan. In order to solve the makespan minimization problem in the two-machine flow shop scheduling, we suggest Johnson's rule as a heuristic algorithm, for which the worst-case bound is calculated. We find polynomial time solutions to some special cases of the considered problems for the following optimization criteria: the weighted sum of completion times and maximum lateness. Some furthermore extensions of the problems are also shown.

• 한국철도학회논문집
• /
• 제9권1호
• /
• pp.50-56
• /
• 2006

This paper presents a mathematical model of the long-term track tamping scheduling problem in the Korean highspeed railway system. The presented model encompasses various operational field constraints, moreover improves a state-of-the-art model in extending the feasible space. We show the model is sized up to intractable scale, then propose another approximation model that is possible to handle with the present computer system and commercial optimization package, directly. The aggregated index, lot, is selected, considering the resolution of the planning horizon as well as scheduling purpose. Lastly, this paper presents two test results for the approximation model. The results expose the approximation model to quite promising in deploying it into an operational software program for the long-term track tamping scheduling problem.

• ETRI Journal
• /
• 제37권5호
• /
• pp.888-897
• /
• 2015

Group scheduling is an operation whereby control packets arriving in a time slot schedule their bursts simultaneously. Normally, those bursts that are of the same wavelength are scheduled on the same channel. In cases where the support of full wavelength converters is available, such scheduling can be performed on multiple channels for those bursts that are of an arbitrary wavelength. This paper presents a new algorithm for group scheduling on multiple channels. In our approach, to reach a near-optimal schedule, a maximum-weight clique needs to be determined; thus, we propose an additional algorithm for this purpose. Analysis and simulation results indicate that an optimal schedule is almost attainable, while the complexity of computation and that of implementation are reduced.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1996년도 하계학술대회 논문집 B
• /
• pp.829-831
• /
• 1996

This paper is concerning on a study on the object-oriented programming and its application to maintenance scheduling. The concept of object-oriented programming enables us to modify and reuse software with much ease. By introducing object-oriented programming to maintenance scheduling, we can develop a hierarchical and reusable software in maintenance scheduling. The maintenance scheduling problem becoming more and more large and complex can be dealt with the concept of object-oriented technique and we hope this concept will give a reasonable solution. And evolutionary computation will be developed as a optimization technique.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 1998년도 추계학술대회 논문집 학회본부A
• /
• pp.341-343
• /
• 1998

Hitherto, train schedules are made by several human experts and the scheduling is very long and tedious job. Moreover those results are not accepted as the optimal solution. The purpose of this research is the finding of optimal methodology and useful constraints for locomotive scheduling. For these purposes, the interactive simulation program for train linking schedule is developed. Some constraints and technique for train linking scheduling is able to be edited or modified by various interactive windows. The constraints, rules and methodology for scheduling can be analyzed and also obtained useful schedule results by using this program.

• International Journal of Computer Science & Network Security
• /
• 제22권3호
• /
• pp.1-8
• /
• 2022

Supporting real-time flows with delay and throughput constraints is an important challenge for future wireless networks. In this paper, we develop an optimal scheduling scheme to optimally choose the packets to transmit. The optimal transmission strategy is based on an observable Markov decision process. The novelty of the work focuses on a priority-based probabilistic packet scheduling strategy for efficient packet transmission. This helps in providing guaranteed services to real time traffic in Heterogeneous Wireless Networks. The proposed scheduling mechanism is able to optimize the desired performance. The proposed scheduler improves the overall end-to-end delay, decreases the packet loss ratio, and reduces blocking probability even in the case of congested network.

• 한국지능시스템학회:학술대회논문집
• /
• 한국퍼지및지능시스템학회 2003년도 ISIS 2003
• /
• pp.187-190
• /
• 2003

Many manufacturing companies consider the integrated and concurrent scheduling because they need the global optimization technology that could manufacture various products more responsive to customer needs. In this paper, we propose an advanced scheduling model to generate the schedules considering resource constraints and precedence constraints in make-to-order (MTO) manufacturing environments. Precedence of work- in-process(WIP) and resources constraints have recently emerged as one of the main constraints in advanced scheduling problems. The advanced scheduling problems is formulated as a multiobjective mathematical model for generating operation schedules which are obeyed resources constraints, alternative workstations of operations and the precedence constraints of WIP in MTO manufacturing. For effectively solving the advanced scheduling problem, the multi-objective hybrid genetic algorithm (m-hGA) is proposed in this paper. The m-hGA is to minimize the makespan, total flow time of order, and maximum tardiness for each order, simultaneously. The m-hGA approach with local search-based mutation through swap mutation is developed to solve the advanced scheduling problem. Numerical example is tested and presented for advanced scheduling problems with various orders to describe the performance of the proposed m-hGA.