• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.53 no.7
• /
• pp.517-524
• /
• 2004

For real world applications of reinforcement learning techniques, function approximation or generalization will be required to avoid curse of dimensionality. For this, an improved function approximation-based reinforcement teaming method is proposed to speed up convergence by using CA-CMAC(Credit-Assigned Cerebellar Model Articulation Controller). To show that our proposed CACRL(CA-CMAC-based Reinforcement Learning) performs better than the CRL(CMAC- based Reinforcement Learning), computer simulation and experiment results are illustrated, where a swing-up control Problem of an acrobot is considered.

• Proceedings of the KIEE Conference
• /
• 2003.11c
• /
• pp.621-624
• /
• 2003

For real world applications of reinforcement learning techniques, function approximation or generalization will be required to avoid curse of dimensionality. For this, an improved function approximation-based reinforcement learning method is proposed to speed up convergence by using CA-CMAC(Credit-Assigned Cerebellar Model Articulation Controller). To show that our proposed CACRL(CA-CMAC-based Reinforcement Learning) performs better than the CRL(CMAC-based Reinforcement Learning), computer simulation results are illustrated, where a swing-up control problem of an acrobot is considered.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.33 no.12B
• /
• pp.1050-1057
• /
• 2008

Wireless Mesh Networks (WMNs) have recently become a hot issue to support high link capacity in wireless access networks. The IEEE 802. I 1 standard which is mainly used for the network interface technology in WMNs supports up to 3 or 12 multiple channels according to the IEEE 802.11 specification. However, two important problems must be addressed when we design a channel assigmnent algorithm: channel dependency problem and channel scanning delay. The former occurs when the dynamic channel switching of an interface leads to the channel switching of other interfaces to maintain node connectivity. The latter happens per channel switching of the interface, and affects the network performance. Therefore, in this paper, we propose the Distance-Based Channel Assigmnent (DB-CA) scheme for multi-channel WMNs to solve such problems. In DB-CA, nodes just perform channel switching without channel scanning to communicate with neighboring nodes that operate on different channels. Furthermore, DB-CA minimizes the interference of channels being used by nodes near the gateway in WMNs. Our simulation results show that DB-CA achieves improved performance in WMNs.

• Journal of the Korean Society of Groundwater Environment
• /
• v.6 no.4
• /
• pp.159-170
• /
• 1999

The hydrogeochemical study on the $CO_2$-rich water in the Chojeong area was carried out. The $CO_2$-rich water of Ca-$HCO_3$type is characterized by low pH (5.0~5.8). high $CO_2$concentration ($Pco_2$＜$10^{0.31}$atm) and high TDS. The water chemistry indicates that the $CO_2$-rich water was probably evolved by the local suppy of deep seated $CO_2$gas resulting in the enhanced water/rock (granite) interaction under low pH conditions. High $NO_3$concentration indicates that the $CO_2$water was mixed and diluted with low $CO_2$groundwater in the vicinity of the area, in which the extensive groundwater abstraction occurred during the past years. The evoiution of the $CO_2$-rich water in the Chojeong area for the process of $CO_2$injection water/rock interaction and mixing processes was thermodynamically simulated by PHREEQC. Although the simulation was limited to water/plagioclase interaction, the results show the feasible explanation about the observed trend of pH and Ca and Na concentrations of the $CO_2$-rich water.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.37B no.9
• /
• pp.750-759
• /
• 2012

This paper addresses the collision resolution issue to enhance the performance of wireless sensor networks based on IEEE 802.15.4 slotted CSMA/CA. Some solutions in existing work try to solve this issue by adjusting Backoff Exponent (BE) value or Backoff Period (BP). In contrast to the existing solutions, the proposed scheme in this paper aims at providing high system throughput, but also achieving efficient energy consumption of sensor nodes by using Preamble Address (PA). For this, in the proposed scheme, only one sensor node begins data packet transmission by performing PA contention phase with other nodes before sending each data packet. Our simulation results show that the proposed scheme outperforms existing algorithms in terms of energy consumption and throughput.

• Proceedings of the Plant Resources Society of Korea Conference
• /
• 2003.10b
• /
• pp.43-43
• /
• 2003

Multiple regression equations (Rehman et al. 2000) have been developed to predict the salinity tolerance of Acacia seeds, expressed as the I$\sub$50/ (the concentration of NaCl required to reduce final germination to 50% of the control value in DW). Accurate predictions can be made using one or more chemical and biological seed parameters. In this study the theoretical effect of varying final germination percentage in distilled water, germination rate in distilled water. (Rate), Ca$\^$2+/ or K$\^$+/ contents and their ratios, as independent factors or related factors, on the predicted salinity tolerance (I$\sub$50/) of Acacia species was investigated. Simulation of the effects of changing final germination, rate, calcium and potassium suggest the possibility of practical application of these results to modify the salinity tolerance of seeds. The predicted I$\sub$50/ increased with increasing final germination. Similarly, the higher the rate of germination, the higher the predicted salt tolerance of Acacia species. The Ca$\^$2+/ content of seeds was found to be positively correlated with I$\sub$50/. Species with higher Ca$\^$2+/ contents had a higher I$\sub$50/. This suggests that I$\sub$50/ might be increased by increasing the Ca$\^$2+/ contents of seeds by pretreatment with calcium salts or by supplying these to the mother plants.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.9 no.1
• /
• pp.91-100
• /
• 2010

According to the recent increase in air traffic a more efficient air traffic flow control is required. This paper proposes the algorithm for efficient air traffic flow control. Aeronautical communication environment is different from the common communication environment since it needs a direct communication between stations (or airplanes) and requires a strict priority for safety message. Moreover, the wide area for basic service causes high propagation delay. In this paper, by using existing CSMA/CA Media Access Control (MAC) protocol, we set the message priorities according to air traffic condition (congestion airport area, approach control area, En route area, and ocean area). This prioritization scheme enables the data transmission with higher access probability. Simulation results confirm that our protocol shows better performance in terms of channel efficiency and success probability compared to exisiting CSMA/CA protocol.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.42 no.10 s.340
• /
• pp.39-50
• /
• 2005

In this paper, we propose a performance improving MAC algerian for the IEEE 802.11 DCF. WLAN based IEEE 802.11 uses two control methods called 'Distributed Coordination Function(UF)' and 'Point Coordination Function(PCF)'. The nF controls the Urnsmission based on carrier sense multiple access with collision detection(CSMA/CA), that decides a random backoff time with the range of contention window for each terminal. Normally, each terminal the CW double after collision, and reduces the CW to the minimum after successful transmission. This paper proposes an enhanced DCF algorithm that decreases the CW smoothly after successful transmission in order to reduce the collision Probability by utilizing the current status information of WLAN. We also analyze the throughput and delay performance for the unsaturated case mathematically. Simulation results show that our algorithm enhances the saturation throughput of WLAN. They also coincide well with the analytical results.

• Journal of the Korean Recycled Construction Resources Institute
• /
• v.11 no.1
• /
• pp.1-8
• /
• 2023

In this study, to reduce CO₂ emission in the cement manufacturing process, we evaluated the limestone that is used as a raw material for cement, substituted with steel slag by the various substituted levels. Based on the chemical composition of each raw materials including limestone, and blast furnace slow cooling slag, converter slag, and KR slag as an alternative raw material, we simulated the optimal cement raw mixture by the substitution levels of limestone. Test results indicated that the steel slags contain a certain level of CaO that can be used as alternative decarbonated raw materials, and it has enough to partially reduce the amount of limestonem. And we estimated the maximum usable levels of each raw material. In particular, it was confirmed that by using a mixture of these raw materials rather than using them one by one, the effect of reducing limestone was increased and CO₂ emission from the cement manufacturing process could be reduced.

• Journal of the Korean Association of Geographic Information Studies
• /
• v.19 no.2
• /
• pp.1-13
• /
• 2016

In an effort to establish adaptive measures for low carbon use and climate change, this study developed storylines for shared socio-economic reference pathways(SSP) and simulated change in land use for each storyline. First, cellular automata modeling was performed using past data, and a transition rule for the local characteristics of each planning area under study was derived by comparing with the results of the base year. Second, three storylines were formulated based on the hypothesized change in land use for the SSP. SSP1, the scenario for sustainability, assumed that the land was developed into a compact city, SSP2 assumed the development of a road through the middle of the land while maintaining the current situation, and SSP3 assumed unsustainable development into a fragmented world. Third, change in land use depending on planning area was predicted by integrating the SSP scenarios with cellular automata(CA) modeling. According to the results of analysis using the SSP scenarios, the urban area ratio increased slightly up to 2020 in SSP1 and up to 2030 in SSP2 and did not change any more subsequently, but it increased continuously until 2050 in SSP3 that assumed low level urban planning. These results on change in land use are expected to contribute towards making reasonable decisions and policies on climate change, and the outcomes of simulation derived from spatial downscaling, if applied to vulnerability assessment, will be useful to set the priority of policies on climate change adaptation.