Jongho Won

GPU-based parallel genetic approach to large-scale travelling salesman problem

The travelling salesman problem (TSP) is a well-known NP-hard problem. It is difficult to efficiently find the solution of TSP even with the large number of gene instances. Evolutionary approaches such as genetic algorithm have been widely applied to explore the huge search space of TSP. However, the feasibility constraints of TSP make it difficult to devise an effective crossover method. In this paper, we propose an improved constructive crossover for TSP. As the performance of graphics processing units (GPUs) rapidly improves, GPU-based acceleration is increasingly required for complex computation problems. Unfortunately, the constructive crossover methods cannot be easily implemented in a parallel fashion because each gene element of offspring is dependent on the previous element in the gene string. In this paper, we propose a more effective method with which large number of genes can evolve effectively by exploiting the parallel computing power of GPUs and an effective parallel approach to genetic TSP where crossover methods cannot be easily implemented in parallel fashion.

Bidirectional Constructive Crossover for Evolutionary Approach to Travelling Salesman Problem

In this paper, we propose an improved crossover method for genetic approach to travelling salesman problem (TSP). Because any feasible solution of TSP must be an ordered permutation, the validity of an offspring generated by the simple crossover where corresponding parts of genes or chromosomes of parents are exchanged. Therefore, researchers have proposed special crossover methods, and so far it is known that SCX is superior to other methods in the aspect of convergence speed and fitness of the genes. In this paper, we extend the SCX to have bidirectional and circular search properties in the construction of offsprings. We also devised an simple and effective index management so that the search for candidate nodes during the offspring construction can be performed in an efficient way. The proposed BCSCX shows the better convergence speed and even better solution than those of SCX in the empirical experiments.

Partial Materialization for Data Integration in SQL-on-Hadoop Engines

SQL-on-Hadoop engines are considered as useful data integration tools for large-scale data. However, they may incur redundant network overhead by redistributing the intermediate results multiple times in the cases where a number of attributes are included in the query result. We propose an optimization method using partial materialization which avoids repetitively redistributing trivial attributes.

A collaboration middleware for service scalability in peer-to-peer systems

We introduce a novel mobile middleware which provides a collaboration service among associated apps in a symmetric fashion. This paper focuses on the challenge that how users can receive the seamless collaboration services regardless of the changes of physical device configurations in the multiscreen environment. In order to solve this problem, we propose a novel system architecture which supports primitive operations for collaborating among distributed applications, such as remote invocation, session join, session invitation, push migration, pull migration and synchronization. Our system can provide communication transparency, seamless collaboration services and scalability among heterogeneous distributed applications. The experimental results demonstrate that our system can be successfully applied to the collaboration services among multiple apps in the home network environment.

Flying KIWI: Design of Approximate Query Processing Engine for Interactive Data Analytics at Scale

This paper introduces the design of hybrid SQL-on-Hadoop system, which supports dual-mode (interactive and deep) analytics. We present an architecture of approximate query processing engine using horizontal and vertical sampling of the original database for interactive big data analytics. A key novelty of our approach is that we can support interactive analytics as well as deep analytics at scale in the unified manner.

Implementation of Real-time Data Stream Processing for Predictive Maintenance of Offshore Plants

In recent years, Big Data has been a topic of great interest for the production and operation work of offshore plants as well as for enterprise resource planning. The ability to predict future equipment performance based on historical results can be useful to shuttling assets to more productive areas. Specifically, a centrifugal compressor is one of the major piece of equipment in offshore plants. This machinery is very dangerous because it can explode due to failure, so it is necessary to monitor its performance in real time. In this paper, we present stream data processing architecture that can be used to compute the performance of the centrifugal compressor. Our system consists of two major components: a virtual tag stream generator and a real-time data stream manager. In order to provide scalability for our system, we exploit a parallel programming approach to use multi-core CPUs to process the massive amount of stream data. In addition, we provide experimental evidence that demonstrates improvements in the stream data processing for the centrifugal compressor.

Realistic woven fabric rendering using deformed microfacet distribution function

We present a new procedural method for photorealistic rendering of woven fabric material. The goal of our research is to provide a new procedural method that renders photorealistic woven fabric without any measured data. The proposed method models the reflectance properties of woven fabric with alternating anisotropy and yarn-level surface normal manipulation. The proposed method models the reflectance properties of woven fabric with alternating anisotropy and deformed microfacet distribution function (MDF). The deformed MDF effectively represents the perturbed reflectance caused by bumps on the woven fabric surface. The experimental results show the proposed method can be successfully applied to photorealistic rendering of diverse woven fabric materials.