• Journal of Korea Multimedia Society
• /
• v.13 no.12
• /
• pp.1798-1804
• /
• 2010

TCP with AIMD mechanism, one of the most popular protocols in internet, can solve congestion control in wired networks. This protocol, however, is not efficient in wireless networks. This paper proposes a new mechanism namely General AIMD with Congestion Window Upper Bound in which congestion window is limited by an upper bound. By applying optimization theory, we find an optimal policy for congestion window upper bound to maximize network throughput.

• Proceedings of the IEEK Conference
• /
• 2003.07a
• /
• pp.477-480
• /
• 2003

Packet losses in the Internet can dramatically degrade quality of multimedia streams. Forward Error Correction (FEC) is one of the best methods that can protect data from packet erasures by means of sending additional redundant information. Proposed control algorithm provides the possibility of receiving real-time multimedia streams of given quality wifth minimal traffic overhead. The traffic optimization is reached by adjusting packet size as well as block code parameters. Calculations and simulation results show that for non-bursty network conditions traffic optimization can lead to more than 50% bandwidth reduction.

• Journal of Communications and Networks
• /
• v.13 no.1
• /
• pp.26-31
• /
• 2011

In this paper, we formulate the minimum power multicast problem in wireless networks as a mixed integer linear programming problem and then propose a Lagrangean relaxation based algorithm to solve this problem. By leveraging on the information from the Lagrangean multiplier, we could construct more power efficient routing paths. Numerical results demonstrate that the proposed approach outperforms the existing approaches for broadcast, multicast, and unicast communications.

• Proceedings of the Korean Institute of Intelligent Systems Conference
• /
• 1993.06a
• /
• pp.1009-1012
• /
• 1993

An association memory, solving an optimization problem, a Boltzmann machine scheme learning and a back propagation learning in our chaos neuron models are reviewed and some new results are presented. In each model its microscopicrule (a parameter of a chaos system in a neuron) is subject to its macroscopic state. This feedback and chaos dynamics are essential mechanisms of our model and their roles are briefly discussed.

• Annual Conference of KIPS
• /
• 2004.05a
• /
• pp.1453-1456
• /
• 2004

PARO, a power-aware routing optimization mechanism, is proposed in [1] to minimize the transmission power needed to forward packets between wireless devices in ad hoc network. The mechanism works by redirecting the route to pass through one or more intermediate nodes on behalf on source-destination pairs, then reducing the end-to-end transmission power. This paper will show an extension of this model and provide an analysis of the geometrical area lying between source and destination in which the intermediate node elects to perform redirection. The duration the intermediate node stays in that area is also computed.

• Proceedings of the Korean Association of Geographic Inforamtion Studies Conference
• /
• 1999.12a
• /
• pp.21-25
• /
• 1999

Planning support systems(PSS) add more advanced spatial analysis functions than Geographic information systems(GIS) and intertemporal functions to the functions of spatial decision support systems(SDSS). This paper reports the continuing development of a PSS providing a framework that facilitates urban planners and civil engineers in conducting coherent deliberations about planning, design and operation & maintenance(O&M) of water-distribution networks for urban growth management. The PSS using dynamic optimization model, modeling-to-generate-alternatives, value engineering(VE) and life-cycle cost(LCC) can generate network alternatives in consideration of initial cost and O&H cost. Users can define alternatives by the direct manipulation of networks or by the manipulation of parameters in the models. The water-distribution network analysis model evaluates the performance of the user-defined alternatives. The PSS can be extended to include the functions of generating sewer network alternatives, combining water-distribution and sewer networks, eventually the function of planning, design and O&H of housing sites. Capacity expansion by the dynamic water-distribution network optimization model using MINLP includes three advantages over capacity expansion using optimal control theory(Kim and Hopkins 1996): 1) finds expansion alternatives including future capacity expansion times, sizes, locations, and pipe types of a water-distribution network provided, 2) has the capabilities to do the capacity expansion of each link spatially and intertemporally, and 3) requires less interaction between models. The modeling using MINLP is limited in addressing the relationship between cost, price, and demand, which the optimal control approach can consider. Strictly speaking, the construction and O&M costs of water-distribution networks influence the price charged for the served water, which in turn influence the. This limitation can be justified in rather small area because price per unit water in the area must be same as that of neighboring area, i.e., the price is determined administratively. Planners and engineers can put emphasis on capacity expansion without consideration of the relationship between cost, price, and demand.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.11 no.3
• /
• pp.215-222
• /
• 2001

Neural network based models of semiconductor manufacturing processes have been shown to offer advantages in both accuracy and generalization over traditional methods. However, model development is often complicated by the fact that back-propagation neural networks contain several adjustable parameters whose optimal values unknown during training. These include learning rate, momentum, training tolerance, and the number of hidden layer neurOnS. This paper presents an investigation of the use of genetic algorithms (GAs) to determine the optimal neural network parameters for the modeling of plasma-enhanced chemical vapor deposition (PECVD) of silicon dioxide films. To find an optimal parameter set for the neural network PECVD models, a performance index was defined and used in the GA objective function. This index was designed to account for network prediction error as well as training error, with a higher emphasis on reducing prediction error. The results of the genetic search were compared with the results of a similar search using the simplex algorithm.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.5 s.43
• /
• pp.165-173
• /
• 2006

Network Mobility Basic Support protocol enables mobile network to change their point of attachment to the Internet, but causes some problems such as suboptimal muting and multiple encapsulations. The proposed scheme, combining Prefix Delegation protocol with HMIPv6 concept can provide more effective route optimization and reduce the amount of packet losses and the burden of location registration for handoff. It also uses hierarchical mobile network prefix (HMNP) assignment and provides tree-based routing mechanism to allocate the location address of mobile network nodes (MNNs) and support micro-mobility. In this scheme, Mobility Management Router (MMR) not only maintains the binding informations for all MNNs in nested mobile networks, but also supports binding procedures to reduce the volume of handoff signals oyer the mobile network. The performance is evaluated using NS-2.

• Transportation Technology and Policy
• /
• v.12 no.6
• /
• pp.16-24
• /
• 2015

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.10 no.1
• /
• pp.206-211
• /
• 2006

The quantification analysis problem is that how the m entities that have n characteristics can be linked to p-dimension space to reflect the similarity of each entity In this paper, the optimization approach for the quantification analysis problem using neural networks is suggested, and the performance is analyzed The computation of average variation volume by mean field theory that is analytical approximated mobility of a molecule system and the annealed mean field neural network approach are applied in this paper for solving the quantification analysis problem. As a result, the suggested approach by a mean field annealing neural network can obtain more optimal solution than the eigen value analysis approach in processing costs.