Earticle

본 논문은 무선 센서망에서 이동 싱크의 동적 브랜치를 통한 데이터 수집 방안을 제안한다. 데이터 전달에 이용되는 트리는 싱크 노드를 루트 노드로 하고 복수개의 브랜치를 가지며, 싱크의 이동과 함께 노드의 부모가 동적으로 변경된다. 또한, 홉 기반의 필터와 플러딩의 효과적인 제한을 통해 제어 트래픽을 줄이고, 효율적인 데이터 전달을 이루는 트리구성 방안을 다룬다. 시뮬레이션 결과를 통해, 제안된 데이터 수집 방안이 기존의 방안과 비교하여, 높은 데이터 도착율과 낮은 전달 지연, 그리고 효과적인 에너지 절약을 보이는 것은 확인할 수 있었다.

This paper suggests a data gathering scheme using dynamic branch tree in wireless sensor networks. A mobile sink gathers data from each sensor node using a dynamic data gathering tree rooted at the mobile sink node. As the sink moves, a tree that has multiple branch is formed and changed dynamically as with the position of the sink node. A hop- based scope filter and a restricted flooding scheme of the tree are also suggested. Simulation results show that the proposed data gathering scheme has better results in data arrival rate, the end-to-end delay and energy saving characteristics compared with the previous scheme.

Park, Bum-Sik & Oh, Sei-Rang. 2015. Left Branch Extraction in Multiple Fragments and Its Implications for Island Violations in Korean. Korean Journal of Linguistics, 41-1, 1-18. In this paper, we investigate the restricted distribution of fragments that undergo Left Branch Extraction (LBE). As observed by Park and Kim (2015a), in multiple fragments only the final fragment can be LBEed. We argue that the restricted distribution is observed because LBE induces a PF-effect on the intervening element between the LBEed fragment and the Left Branch Island. In multiple fragments, the non-final LBEed fragment always involves the configuration where the final fragment counts as a surviving intervenor after ellipsis takes place in PF. By contrast, appearing in the final position, the final fragment can be LBEed with the potential intervenors eliminated under ellipsis, giving rise to a repair effect. The analysis partially supports Fox and Lasnik (2003), but not Merchant (2008) or Chomsky (1972), in the sense that certain illegitimate movement has a PF effect on the intervening element. Examining interactions between LBE and relative clause island, we also provide an account for a certain asymmetry regarding the island-(in)sensitivity. (Dongguk University & Gyeongsang National University)

Park, Bum-Sik & Oh, Sei-Rang. 2015. Left Branch Extraction in Multiple Fragments and Its Implications for Island Violations in Korean. Korean Journal of Linguistics, 41-1, 1-18. In this paper, we investigate the restricted distribution of fragments that undergo Left Branch Extraction (LBE). As observed by Park and Kim (2015a), in multiple fragments only the final fragment can be LBEed. We argue that the restricted distribution is observed because LBE induces a PF-effect on the intervening element between the LBEed fragment and the Left Branch Island. In multiple fragments, the non-final LBEed fragment always involves the configuration where the final fragment counts as a surviving intervenor after ellipsis takes place in PF. By contrast, appearing in the final position, the final fragment can be LBEed with the potential intervenors eliminated under ellipsis, giving rise to a repair effect. The analysis partially supports Fox and Lasnik (2003), but not Merchant (2008) or Chomsky (1972), in the sense that certain illegitimate movement has a PF effect on the intervening element. Examining interactions between LBE and relative clause island, we also provide an account for a certain asymmetry regarding the island-(in)sensitivity. (Dongguk University & Gyeongsang National University)

This paper presents a multiple branch predictor with perceptrons. We describe our design and evaluate it with the SPEC 2000 benchmarks. Our predictor achieves increased accuracy than the previous multiple branch predictors.

An analytical performance model that can predict the performance of a superscalar processor employing multiple branch prediction is introduced. The model is based on the conditional independence probability and the basic block size of instructions, with the degree of multiple branch prediction, the fetch rate, and the window size of a superscalar architecture. Trace driven simulation is performed for the subset of SPEC integer benchmarks, and the measured IPCs are compared with the results derived from the model. As the result, our analytic model could predict the performance of the superscalar processor using multiple branch prediction within 6.6 percent on the average.

This paper deals with the single-hoist and multiple-products scheduling problem. Although a mixed integer linear programming model for the problem was developed earlier, a branch-and-bound based heuristic algorithm is proposed in this paper to solve the big-size problems in real situation. The algorithm is capable of handling problems incorporating different product types, jobs in the process, and tank capacities. Using a small example problem the procedure of the heuristic algorithm is explained. To assess the performance of the heuristic we generate a bigger example problem and compare the results of the algorithm proposed in this paper with the optimal solutions derived from the mathematical model of earlier research. The comparison shows that the heuristic has very good performance and the computation time is sufficiently short to use the algorithm in real situation.

본 연구의 목적은 다목적최적화(multiple objective optimization)를 위한 방법론을 개발하고 적용하는데 있다. 많은이론과 접근법이 존재함에도 불구하고, 이해하기 쉬우며 실제문제로의 적용이 용이한 방법론은 적다. 적용의 용이성을위해 무엇보다 중요한 것은 의사결정자와의 상호교호작용(즉 의사결정자의 입장에서 정보제공)이 용이해야 한다. 본 요구사항을 충족시키기 위해 상호교호적 방법론(interactive method)을 개발한다. 이는 가장 잘 알려진 기존의 두 가지상호교호기법들의 단점을 보충하고 장점을 취하는 상호 보완된 방법론이다. 또한 개발된 방법론이 관련된 몇몇 다른 기법들과 비교하여 주목할 만한 향상을 이루었다는 점을 강조한다. 개발된 방법론을 다수의 복잡한 비선형 목적함수들을갖는 사례문제로 적용한다. 본 사례는 즉석식품 회사의 신규개설지점에 대한 운영설계(operational design)와 생산목표설정(production target setting)에 관한 문제이다. 적용에 있어서 비선형 다목적문제의 설정부터 해 도출에 이르기까지의 모든 과정을 상세히 보여준다.

Suppose a methodology has what it takes to succeed in terms of its application to real problems. It might then be obvious that the method is a simple, understandable, and usable approach to the problem setting in question. This study elaborates on this highly abstract but obvious and important requirement in the context of multiple objective optimization (MOO) endeavors. We develop an enhanced interactive method to meet the requirement, which crossfertilizes the two most well-known interactive techniques to compensate for the drawbacks of the two methods and capture their positive aspects. Special emphases are also placed on what the current development makes a considerable improvement in comparison to some other relevant methods. The developed method is then applied to a real-world case problem involving several highly nonlinear objective functions, the problem of operational design and production target setting for the opening branches of a fast-food company. We demonstrate in detail the entire process of the application from modeling a nonlinear MOO problem to generating its solution, in order to guide the practical use of our method to the other potential applications. During the last five decades, a great deal of theories and methods have been developed to resolve the MOO problems. Interactive approach is one of the most widely used families of the MOO techniques. This is based upon a human-computer interaction process in that the computer algorithm generates and presents a solution to the human decision maker, and the decision maker then provides the algorithm with information, so this process repeats until the final solution satisfies the decision maker. The show-and-tell approach articulates the decision maker’s preferences gradually and, hence, helps her to provide her preferences in a step-by -step manner, unlike other families of the MOO techniques requiring all such information at a given moment. Furthermore, the decision maker can learn about the changing pattern of the generated solutions and anticipate the next solution during the solution process, which is also of great help in both information supply and desirable solution choice. Many different interactive methods constitute the family, interactive approach, and they each have different computational algorithms and require different types of information from the decision maker. The success of an interactive method usually lies in how easy and comfortable it is in terms of not only its computational aspect but also information requirement. We thus delve deeply into the characteristics of several representative interactive methods and take advantage of their positive features. As a result, a significantly improved interactive method is developed, thus minimizing both computational and informational burdens on the decision maker. This paper also explores an important managerial problem of operational design and production target setting for opening branches in a fast-food company. All the branches utilize multiple inputs, such as manpower and operating costs, to produce multiple outputs like revenue and customer satisfaction. It is therefore needed to plan as to how much inputs should be used and how much outputs should be produced, before the new branch’s operation. The input setting is referred to as operational design and the output setting as production target setting. To accomplish this, we model a nonlinear MOO problem and apply our method to that problem. It is worth mentioning that there are various cases of opening new branches in many different kinds of industries. Examples are fast-food franchise restaurants, telecommunication service offices, and bank branches. Possible areas of the application of our modeling idea and method are therefore numerous.