• 제어로봇시스템학회:학술대회논문집
• /
• 2003.10a
• /
• pp.2131-2136
• /
• 2003

Since the number of user-requests increases dramatically on the Internet, the servers and networks can be swamped unexpectedly without any prior notice. Therefore, the end-users are waiting or refused for the responses of the contents from the originating servers. To solve this problem, it has been considered that a distributed web caching system efficiently utilizes structural elements of the network. Because a distributed web caching system uses the caches that are close to end-users on the network, it transmits the contents to users faster than the original network system. This paper proposes a simple client-based distributed web caching system(2HRCS) that client can directly perform object allocation and load balancing without an additional DNS for load balancing in CARP (Cache Array Routing Protocol) and GHS (Global Hosting System) that are the recent distributed web caching system protocol. The proposed system reduces the cost of setup and operation by removing DNS that needs to balance the load in the existing system. The system has clients with consistent hashing method, so it extends its environment to other distributed web caching system that has caches of different capacity. A distributed web caching system is composed and tested to evaluate the performance. As a result, it shows superior performance to consistent hashing system. Because this system can keep performance of the existing system and reduce costs, it has the advantage of constructing medium or small scale CDN (Contents Delivery Network).

• Journal of KIISE:Information Networking
• /
• v.31 no.6
• /
• pp.595-607
• /
• 2004

Networks often evolve to provide a host with multiple access points to the Internet. In this paper, we propose a transport layer load distribution mechanism utilizing the multiple network interfaces simultaneously. We specifically propose an extension of Stream Control Transmission Protoco1 (SCTP) to have load sharing over multiple network interfaces. We named the particular service provided by the Proposed load sharing mechanism to be LS (Load Sharing) mode service. LS mode service is based on the following four key elements: (i) the separation of flow control and congestion control, (ii) congestion window based striping, (iii) redundant packet retransmission for fast packet loss recovery, (iv) a novel mechanism to keep track of the receiver window size with the SACKS even if they arrive out-of-order. Through simulations, it is shown that the proposed LS mode service can aggregate the bandwidth of multiple paths almost ideally despite of the disparity in their bandwidth. When a path with a delay of 100% greater is utilized as the second path, the throughput is enhanced about 20%.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2007.05a
• /
• pp.1288-1290
• /
• 2007

Congestion in WSN increases energy dissipation rates of sensor nodes as well as loss of packets and thereby hinders fair and reliable event detections. We find that one of the key reasons of congestion in WSN is allowing sensing nodes to transfer as many packets as possible. This is due to the use of CSMA/CA that gives opportunistic media access control. In this paper, we propose an energy efficient congestion avoidance protocol that includes source count based hierarchical and load adaptive medium access control. Our proposed mechanism ensures load adaptive media access to the nodes and thus achieves fairness in event detection. The results of simulation show our scheme exhibits more than 90% delivery ratio with retry limit 1, even under bursty traffic condition which is good enough for reliable event perception.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.16 no.12
• /
• pp.1417-1422
• /
• 1991

In this paper we develop an analytic model for end-to-end communication protocols and study the window mechanism for flow control in store-and-forward computer networks. We develop a flow control model in which the parameters of the system are not dynamically adjusted to the stochastic fluctuation of the system load. Simulation results are presented and it is shown that the throughput-delay performance of a network can be improved by proper selection of the design parameter, such as timeout, the number of retransmission, etc.

• Proceedings of the Korean Information Science Society Conference
• /
• 2012.06d
• /
• pp.401-403
• /
• 2012

In this paper, we introduce a multipath congestion control algorithm for audio/video streaming, called MPSQRT. MPSQRT is derived from single-path Square-Root TCP (SQRT TCP) designed for multimedia streaming, where a lower variation of sending rate is important. Based on the fluid model of SQRT for single-path, we extend it towards spreading concurrently packets across multiple paths ensuring load-balancing and fairness to SQRT at shared bottleneck. Through simulations, we evaluate the proposed protocol under various network conditions.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• v.5 no.1
• /
• pp.123-140
• /
• 2011

In this paper, we propose an Energy Efficient Media Access Control (EE-MAC) protocol for wireless sensor networks. The proposed scheme is designed to save power consumption and guarantee quality-of-service for real-time traffic. EE-MAC uses the superframe structure which is bounded by the transmission of a beacon frame and can have an active and an inactive portion. The active period is divided into the contention free period (CFP) for real-time traffic transmission and the contention access period (CAP) for non-real-time traffic transmission. We propose the exclusively allocated backoff scheme which assigns a unique backoff time value to each real-time node based on bandwidth allocation and admission control. This scheme can avoid collision between real-time nodes by controlling distributed fashion and take effect a statistical time division multiple access. We also propose the algorithm to change the duty cycle adaptively according to channel utilization of media depending on network traffic load. This algorithm can prolong network lifetime by reducing the amount of energy wasted on idle listening.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.5 no.6
• /
• pp.1083-1088
• /
• 2001

In this paper, we propose a new CBT(core-based tree)multicast routing Mechanism by load balancing mechanism. CBT may result in traffic concentration and bottlenecks near the core routers since traffic from all sources nodes the same set of links as approaches the core. So the router have to load packet to new core router for such congestion. Congestion information is given by rtt(round trip time) between designed core router and the other router.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2001.10a
• /
• pp.154-157
• /
• 2001

In this paper, we propose a new CBT(core-based tree)multicast routing Mechanism by load balancing mechanism. CBT may result in traffice concentration and bottlenecks near the lore routers since traffice from all sources nodes the same set of links as approaches the core. So the router have to load packet to new core router for such congestion. Congestion information is given by rtt(round trip time) between designed core router and the other router.

• The KIPS Transactions:PartC
• /
• v.9C no.6
• /
• pp.865-874
• /
• 2002

HTTP (HyperText Transfer Protocol) is a transfer protocol used by the World Wide Web distributed hypermedia system to retrieve the objects. Because HTTP is a connection oriented protocol, it uses TCP (Transmission control Protocol) as a transport layer. But it is known that HTTP interacts with TCP badly. it is discussed about factors affecting the performance or HTTP over TCP, the transaction time obtained by the per-transaction TCP connections for HTTP access and the TCP slow-start overheads, and the transaction time for T-TCP (Transaction TCP) which is one or methods improving the performance or HTTP over TCP. Average waiting time and average turnaround time are important parameters to satisfy QoS (Quality of Service) of end users. Formulas for calculating two parameters are derived. Such formulas can be used for the environment in which each TCP or T-TCP transaction time is same or different. Some experiments and computational experiences indicate that the proposed formulas are well acted, can be applied to the environment which the extension of bandwidth is necessary, and time characteristics of T-TCP are superior to that of TCP. Also, the load distribution method of web server based on the combination of bandwidths is discussed to reduce average waiting time and average turnaround time.

• Journal of Institute of Control, Robotics and Systems
• /
• v.6 no.7
• /
• pp.602-610
• /
• 2000

In communication networks used in safety-critical systems such as control systems in nuclear power plants there exist three types of data traffic : urgent or asynchronous hard real-time data hard real-time periodic data and soft real-time periodic data. it is necessary to allocate a suitable bandwidth to each data traffic in order to meet their real-time constraints. This paper proposes a method to meet the real-time constraints for the three types of data traffic simultaneously under a timer-controlled token bus protocol or the IEEE 802.4 token bus protocol and verifies the validity of the presented method by an example. This paper derives the proper region of the high priority token hold time and the target token rotation time for each station within which the real-time constraints for the three types of data traffic are met, Since the scheduling of the data traffic may reduce the possibility of the abrupt increase of the network load this paper proposes a brief heuristic method to make a scheduling table to satisfy their real-time constraints.