• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.18 no.3
• /
• pp.551-569
• /
• 2024

With the rapid development of electric vehicles (EVs) industry, EV charging service becomes more and more important. Especially, in the case of suddenly drop of air temperature or open holidays that large-scale EVs seeking for charging devices (CDs) in a short time. In such scenario, inefficient EV charging scheduling algorithm might lead to a bad service quality, for example, long queueing times for EVs and unreasonable idling time for charging devices. To deal with this issue, this paper propose a Deep-Q-Network (DQN) based two-stage scheduling method for the large-scale EVs charging service. Fine-grained states with two delicate neural networks are proposed to optimize the sequencing of EVs and charging station (CS) arrangement. Two efficient algorithms are presented to obtain the optimal EVs charging scheduling scheme for large-scale EVs charging demand. Three case studies show the superiority of our proposal, in terms of a high service quality (minimized average queuing time of EVs and maximized charging performance at both EV and CS sides) and achieve greater scheduling efficiency. The code and data are available at THE CODE AND DATA.

• Journal of Korea Society of Industrial Information Systems
• /
• v.29 no.5
• /
• pp.11-19
• /
• 2024

As task offloading from user devices transitions from the cloud to the edge, the demand for efficient resource management techniques has emerged. While numerous studies have employed reinforcement learning to address this challenge, many fail to adequately consider the overhead associated with real-world offloading tasks. This paper proposes a reinforcement learning-based distributed offloading policy generation method that incorporates task overhead. A simulation environment is constructed to validate the proposed approach. Experimental results demonstrate that the proposed method reduces edge queueing time, achieving up to 46.3% performance improvement over existing approaches.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.39 no.1
• /
• pp.20-33
• /
• 2002

Usage of ATM AAL2 packets becomes dominant to increase transmission efficiency of voice traffic in the backbone network. In case of voice service that uses AAL2 mechanism, if resources of network are enough, connection of new call is accepted. However, due to packets generated by the new call, transmission delay of packets from old calls can increase sharply. To control this behavior, in this paper we present an AAL2 buffer management scheme that allocates a virtual buffer to each call and after calculating its propagation delay variation(PDV), decides to drop packets coming from each call according to the PDV value. We show that this packet dropping algorithm can effectively prevent abrupt QoS degradation of old calls. To do this, we analyze AAL2 packet composition process to find a critical factor in the process that influences the end-to-end delay behavior and model the process by K-policy M/D/1 queueing system and MIN(K, Tc)-policy M/D/1 queueing system. From the mathematical model, we derive the probability generating function of AAL2 packets in the buffer and mean waiting time of packets in the AAL2 buffer. Analytical results show that the AAL2 packet dropping algorithm can provide stable AAL2 packetization delay and ATM cell generation time even if the number of voice sources increases dramatically. Finally we compare the analytical result to simulation data obtained by using the COMNET Ⅲ package.

• Journal of the Korea Society for Simulation
• /
• v.24 no.3
• /
• pp.27-35
• /
• 2015

We analyze an M/G/1/K queueing system with queue-length dependent service and arrival rates. There are a single server and a buffer with finite capacity K including a customer in service. The customers are served by a first-come-first-service basis. We put two thresholds $L_1$ and $L_2$($${\geq_-}L_1$$ ) on the buffer. If the queue length at the service initiation epoch is less than the threshold $L_1$, the service time of customers follows $S_1$ with a mean of ${\mu}_1$ and the arrival of customers follows a Poisson process with a rate of ${\lambda}_1$. When the queue length at the service initiation epoch is equal to or greater than $L_1$ and less than $L_2$, the service time is changed to $S_2$ with a mean of $${\mu}_2{\geq_-}{\mu}_1$$. The arrival rate is still ${\lambda}_1$. Finally, if the queue length at the service initiation epoch is greater than $L_2$, the arrival rate of customers are also changed to a value of $${\lambda}_2({\leq_-}{\lambda}_1)$$ and the mean of the service times is ${\mu}_2$. By using the embedded Markov chain method, we derive queue length distribution at departure epochs. We also obtain the queue length distribution at an arbitrary time by the supplementary variable method. Finally, performance measures such as loss probability and mean waiting time are presented.

• Journal of the Society of Naval Architects of Korea
• /
• v.61 no.4
• /
• pp.267-277
• /
• 2024

The Sortie Generation Rate (SGR) is a critical performance indicator for carrier-based aircraft and is a key factor for the carrier design process. This study aims to analyze the factors that affect SGR and establish a representative Sortie Generation Process (SGP) along with simulation results to calculate SGR for a naval ship equipped to carry aircraft. Detailed SGR factors are identified from the perspectives of the aircraft, aviation personnel, and aircraft carrier during the flight preparation stage, and the SGP is established accordingly. As a representative, Korean Navy's CVX basic design is chosen for detailed analysis. The physical dimension and spots for the deck design with time and probabilistic data of SGP are considered to develop a queueing network model for SGR calculation. To consider the specific probabilistic features, the model was solved with discrete event simulation tools(SimPy and AnyLogic) where the results show great agreement. Such findings on SGR factors and calculation are expected to be incorporated in the future development of SGR calculation algorithms and also present guidelines for proper design of aircraft carrier based on concrete operation concept.

• Journal of KIISE:Information Networking
• /
• v.28 no.4
• /
• pp.602-610
• /
• 2001

In the MPLS(Multiprotocol Label Switching) forwarding paradigm, once a packet is assigned to a FEC(Forwarding Equivalence Class), no further header analysis is done by subsequent routers; all forwarding is driven by the labels. This has a number of advantages over conventional network layer forwarding. The MPLS LER(Label Edge Router) is located at the boundary of MPLS domain and plays a role in connecting with the existing Internet as an ingress or an egress router. That is, the MPLS LER as an ingress router assigns a label to a packet which enters the MPLS network from the Internet by analyzing its header and forwards to a corresponding next router in MPLS domain. As an egress router, the MPLS LER turns the packets out of the MPLS network by performing the reverse operation. In this paper, we analyze the traffic performance of an MPLS LER system and estimate the IP(Internet Protocol) packet processing capacity of the system using queueing model and simulation. It is found that the maximum IP packet processing capacity of the system is estimated by 420,000 through 460,000 packets/sec.

• Journal of the Institute of Electronics Engineers of Korea TC
• /
• v.43 no.3 s.345
• /
• pp.126-134
• /
• 2006

Traditional TCP implementations have the under-utilization problem in large bandwidth delay product networks especially during the startup phase. In this paper, we propose a delay-based congestion control(DCC) mechanism to solve the problem. DCC is subdivided into linear and exponential growth phases. When there is no queueing delay, the congestion window grows exponentially during the congestion avoidance period. Otherwise, it maintains linear increase of congestion window similar to the legacy TCP congestion avoidance algorithm. The exponential increase phase such as the slow-start period in the legacy TCP can cause serious performance degradation by packet losses in case the buffer size is insufficient for the bandwidth-delay product, even though there is sufficient bandwidth. Thus, the DCC uses the RTT(Round Trip Time) status and the estimated queue size to prevent packet losses due to excessive transmission during the exponential growth phase. The simulation results show that the DCC algorithm significantly improves the TCP startup time and the throughput performance of TCP in large bandwidth delay product networks.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2013.05a
• /
• pp.401-403
• /
• 2013

Myopic method which is one of dynamic spectrum access method is an advantage that it maximizes immediate throughput is not complicated, and the simple and effective. The disadvantage of myopic method is that there may become less efficient because of conflicts between users secondary secondary user becomes many. Between the secondary if the secondary all users, select a channel to maximize the throughput of their own without considering the secondary other users, but it will try transfer secondary most users select the same channel collision will be more. Myopic / CA method has been proposed to solve this problem. Instead of selecting one channel to maximize their performance, select a candidate channels plurality of their advantage, the secondary user transfers to randomly selected channel that one in this way is a scheme. In this paper, we analyze whether it can be defined in a number of the channel number of candidates, depending on the network environment, maximizes the performance of the entire network.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.23 no.6
• /
• pp.1514-1525
• /
• 1998

Two main characteristics of VBR MPEG video traffic are different statistics according to different picture types and the periodic traffic pattern due to GOP structure. Especially, the I-pictures at the beginning of each GOP generate $$\mu$h more traffic than other pictures. When several VBR Mpeg video sources are superposed, the I-picture starting times of these sources may significantly affect the cell loss characteristics of ATM $$\mu$tiplexers. In this paper, we propse a performance model for ATM $$\mu$tiplexers with VBR MPEG video sources whose starting times are arbitrary given. For analysis, both single and superposed source traffic are modeled as NDPPs (non-deterministic periodic processes), and the ATM $$\mu$tiplexer is modeled as a U-state NDPPD/D/1/B queueing system. It is hown that the numerical results are very close to the si$$\mu$ation results. From the relationships between the starting tiem distributions and the corresponding $$\mu$tiplexer performances, some considerations for designing a scheduling policy in order to obtain the maxi$$\mu$ $$\mu$tiplexing gain are presented.

• Journal of the Korea Society for Simulation
• /
• v.20 no.1
• /
• pp.1-8
• /
• 2011

Cloud computing emerges as a new paradigm for deploying, managing and offering IT resources as a service anytime, anywhere on any devices. Cloud computing data center stores many IT resources through resource integration. So cloud computing system has to be designed by technology and policy to make effective use of IT resources. In other words, cloud vendor has to provide high quality services to all user and mitigate the dissipation of IT resources. However, vendors need to predict the performance of cloud services and the use of IT resources before releasing cloud service. For solving the problem, this research presents cloud service modeling on network environment and evaluation index for availability optimization of cloud service. We also study how to optimize an amount of requested cloud service and performance of datacenter using CloudSim toolkit.