• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.14 no.5
• /
• pp.1909-1928
• /
• 2020

Predicting human mobility has always been an important task in Location-based Social Network. Previous efforts fail to capture spatial dependence effectively, mainly reflected in weakening the location topology information. In this paper, we propose a neural network-based method which can capture spatial-temporal dependence to predict the next location of a person. Specifically, we involve a graph convolutional network (GCN) based on a seq2seq framework to capture the location topology information and temporal dependence, respectively. The encoder of the seq2seq framework first generates the hidden state and cell state of the historical trajectories. The GCN is then used to generate graph embeddings of the location topology graph. Finally, we predict future trajectories by aggregated temporal dependence and graph embeddings in the decoder. For evaluation, we leverage two real-world datasets, Foursquare and Gowalla. The experimental results demonstrate that our model has a better performance than the compared models.

• Journal of Power Electronics
• /
• v.15 no.5
• /
• pp.1168-1177
• /
• 2015

This paper presents a scheme to improve the line current distortion of power factor corrector (PFC) topology at the zero crossing point using a predictive control algorithm in both the continuous conduction mode (CCM) and discontinuous conduction mode (DCM). The line current in single-phase PFC topology is distorted at the zero crossing point of the input AC voltage because of the characteristic of the general proportional integral (PI) current controller. This distortion degrades the line current quality, such as the total harmonic distortion (THD) and the power factor (PF). Given the optimal duty cycle calculated by estimating the next state current in both the CCM and DCM, the proposed predictive control algorithm has a fast dynamic response and accuracy unlike the conventional PI current control method. These advantages of the proposed algorithm lower the line current distortion of PFC topology. The proposed method is verified through PSIM simulations and experimental results with 1.5 kW bridgeless PFC (BLPFC) topology.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.36 no.9B
• /
• pp.1039-1048
• /
• 2011

Carrier Ethernet, which extend The algorithm based on constructing the mixed topology and performing link stretching, MT/s, has been proposed for designing cost-efficient Carrier Ethernet in optical network with multi-line-rate. However, the MT/s algorithm has high blocking ratio because the wavelength capacity is fully allocated without considering the load balance of network. In this paper, we propose an efficient mixed topology configuration (EMTC) algorithm by modifying MT/s algorithm. In order to reduce blocking ratio, we adapt a threshold for each link to restrict the link utilization so that traffic load can be distributed over whole network. We also apply the EMTC algorithm into optical hybrid switched network to evaluate the availability of our algorithm for different applications. The performance of the EMTC algorithm is compared with that of MT/s algorithm through OPNET simulation. The simulation results show that our algorithm achieve lower blocking ratio than the MT/s algorithm. Moreover, in hybrid switched network, our algorithm performs better than MT/s algorithm in terms of packet loss ratio and end-to-end delay.

• Journal of the Korean Mathematical Society
• /
• v.47 no.5
• /
• pp.1031-1054
• /
• 2010

Let $\mathbb{Z}^n$ be the Cartesian product of the set of integers $\mathbb{Z}$ and let ($\mathbb{Z}$, T) and ($\mathbb{Z}^n$, $T^n$) be the Khalimsky line topology on $\mathbb{Z}$ and the Khalimsky product topology on $\mathbb{Z}^n$, respectively. Then for a set $X\;{\subset}\;\mathbb{Z}^n$, consider the subspace (X, $T^n_X$) induced from ($\mathbb{Z}^n$, $T^n$). Considering a k-adjacency on (X, $T^n_X$), we call it a (computer topological) space with k-adjacency and use the notation (X, k, $T^n_X$) := $X_{n,k}$. In this paper we introduce the notions of KD-($k_0$, $k_1$)-homotopy equivalence and KD-k-deformation retract and investigate a classification of (computer topological) spaces $X_{n,k}$ in terms of a KD-($k_0$, $k_1$)-homotopy equivalence.

• Journal of the Optical Society of Korea
• /
• v.19 no.1
• /
• pp.13-21
• /
• 2015

In recent years energy consumption has become a main concern for network development, due to the exponential increase of network traffic. Potential energy savings can be obtained from a load-adaptive scheme, in which a day can be divided into multiple time periods according to the variation of daily traffic patterns. The energy consumption of the network can be reduced by selectively turning off network components during the time periods with light traffic. However, the time segmentation of daily traffic patterns affects the energy savings when designing multiperiod logical topology in optical wavelength routed networks. In addition, turning network components on or off may increase the overhead of logical topology reconfiguration (LTR). In this paper, we propose two mixed integer linear programming (MILP) models to design the optimal logical topology for multiple periods in IP-over-WDM networks. First, we formulate the time-segmentation problem as an MILP model to optimally determine the boundaries for each period, with the objective to minimize total network energy consumption. Second, another MILP formulation is proposed to minimize both the overall power consumption (PC) and the reconfiguration overhead (RO). The proposed models are evaluated and compared to conventional schemes, in view of PC and RO, through case studies.

• Computational Structural Engineering : An International Journal
• /
• v.1 no.1
• /
• pp.21-29
• /
• 2001

The general topology optimization can be considered as optimal material distribution. Such an approach can be unstable, unless composite materials are introduced. In this research, a nodal volume fraction method is used to obtain the optimum topology of continuum structures. This method is conducted from a composite material model composed of isotropic matter and spherical void. Because the appearance of the chessboard patterns makes the interpretation of the optimal material layout very difficult, this method contains a chessboard prevention strategy. In this research, several topology optimization problems are presented to demonstrate the validity of the present method and the recursive quadratic programming algorithm is used to solve the topology optimization problems.

• Proceedings of the Korea Contents Association Conference
• /
• 2005.11a
• /
• pp.601-607
• /
• 2005

In this paper, we aggregate multi-links information between boundary nodes using the line segment scheme that aggregates topology information within PG referring bandwidth and delay parameter. To do this, we propose a modified line segment algorithm using two line segment method that represents two points which consist of delay-bandwidth pair to reduce topology information and provide a flexibility to the multiple-links aggregation. And we apply it to current star topology aggregation. Through the simulation result analysis, the proposed star topology aggregation scheme presents the better performance than existing scheme.

• Journal of Power Electronics
• /
• v.15 no.1
• /
• pp.96-105
• /
• 2015

In recent years, the multilevel converters have been given more attention due to their modularity, reliability, failure management and multi stepped output waveform with less total harmonic distortion. This paper presents a novel symmetric multilevel inverter topology with reduced switching components to generate a high quality stepped sinusoidal voltage waveform. The series and parallel combinations of switches in the proposed topology reduce the total number of conducting switches in each level of output voltages. In addition, a comparison between the proposed topology with another topology from the literature is presented. To verify the proposed topology, the computer based simulation model is developed using MATLAB/Simulink and experimentally with a prototype model results are then compared.

• Journal of Biomedical Engineering Research
• /
• v.26 no.2
• /
• pp.73-86
• /
• 2005

We need to flatten the brain cortex to smooth surface, sphere, or 2D plane in order to view the buried sulci. The rendered 3D surface of the segmented white matter and gray matter does not have the topology of a sphere due to the partial volume effect and segmentation error. A surface without correct topology may lead to incorrect interpretation of local structural relationships and prevent cortical unfolding. Although some algorithms try to correct topology, they require heavy computation and fail to follow the deep and narrow sulci. This paper proposes a method that corrects topology of the rendered surface fast, accurately, and automatically. The proposed method removes fractions beside the main surface, fills cavities in the inside of the main surface, and removes handles in the surface. The proposed method to remove handles has three-step approach. Step 1 performs smoothing operation on the rendered surface. In Step 2, vertices of sphere are gradually deformed to the smoothed surfaces and finally to the boundary of the segmented white matter and gray matter. The Step 2 uses multi-resolutional approach to prevent the deep sulci from geometrical intersection. In Step 3, 3D binary image is constructed from the deformed sphere of Step 2 and 3D surface is regenerated from the 3D binary image to remove intersection that may happen. The experimental results show that the topology is corrected while principle sulci and gyri are preserved and the computation amount is acceptable.

• Communications of the Korean Mathematical Society
• /
• v.18 no.4
• /
• pp.713-723
• /
• 2003

We study a limit problem of reflected solutions of parabolic stochastic partial differential equations driven by martingale measures. The existence of solutions is found in an extension of the work with respect to white noise by Donati-Martin and Pardoux [4]. We show that if a certain sequence of driving martingale measures converges, the corresponding solutions also converge in the Skorohod topology.