• ETRI Journal
• /
• v.40 no.3
• /
• pp.337-346
• /
• 2018

In this paper, we propose a grant-aware (GA) scheduling algorithm that can provide higher throughput and lower latency than a conventional dual round-robin matching (DRRM) method. In our proposed GA algorithm, when an output receives requests from different inputs, the output not only sends a grant to the selected input, but also sends a grant indicator to all the other inputs to share the grant information. This allows the inputs to skip the granted outputs in their input arbiters in the next iteration. Simulation results using OPNET show that the proposed algorithm provides a maximum 3% higher throughput with approximately 31% less queuing delay than DRRM.

• Journal of Communications and Networks
• /
• v.5 no.1
• /
• pp.82-89
• /
• 2003

In this paper, an efficient threshold-based filtering (TF) buffer management scheme is proposed. The TF is capable of minimizing the overall loss performance and improving the fairness of buffer usage in a shared buffer packet switch. The TF consists of two mechanisms. One mechanism is to classify the output ports as sctive or inactive by comparing their queue lengths with a dedicated buffer allocation factor. The other mechanism is to filter the arrival packets of inactive output ports when the total queue length exceeds a threshold value. A theoretical queuing model of TF is formulated and resolved for the overall packet loss probability. Computer simulations are used to compare the overall loss performance of TF, dynamic threshold (DT), static threshold (ST) and pushout (PO). We find that TF scheme is more robust against dynamic traffic variations than DT and ST. Also, although the over-all loss performance between TF and PO are close to each other, the implementation of TF is much simpler than the PO.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.1 s.39
• /
• pp.157-165
• /
• 2006

This paper proposes a new output-buffered multicast ATM switch with tandem crosspoints switching fabric, named the MTCOS(Multicast Tandem Crosspoint Output-buffered Switch). The MTCOS consists of multiple simple crosspoint switch fabrics, named TCSF(Tandem Crosspoint Switch Fabric) , and concentrated output buffers for efficient multicasting. The TCSF resolves the cell delay deviation problem which the self-routing crossbar switches inherently have. Further, it offers multiple concurrent pathes from one input to multiple output ports. It also provides multi-channel switching by easy software configuration and has several desirable characteristics such as scalability, high Performance, and modularity. A shared traffic concentration and output queuing strategies of the MTCOS results in lower cell loss as well as lower cell delay time over a wide range of multicast traffic. Furthermore, it has lower hardware complexity than that of the SCOQ and Knockout multicast switch to achieve the same Knockout concentration rate as the conventional switches. It is shown that the proposed switch can be easily applied to design high performance for any multicast traffic by analytic analysis and computer simulation.

• Journal of KIISE:Computing Practices and Letters
• /
• v.12 no.1
• /
• pp.78-89
• /
• 2006

In development of network-enabled consumer electronics, much of the time and effort is spent analyzing and solving network performance problems. In this paper, we define an instance of such problems discovered while developing a commercial home network gateway. We then analyze its cause and propose a solution mechanism. Our home network gateway uses art asymmetric link (ADSL) and suffers from an undesirable phenomenon where downlink traffic interferes with upload speed. We call this phenomenon the network performance interference problem. While this problem can easily be confused with receive livelock caused by packet contention at the input queue, we and that this is not the case. By performing extensive experiments and analysis, we reveal that our problem is caused by packet contention at the output queue and certain intrinsic characteristics of TCP. We devise an ACK-separation queuing mechanism for this problem and implement it in the home network gateway Our experiments show that it effectively solves the problem.

• Journal of Korean Society of Industrial and Systems Engineering
• /
• v.45 no.4
• /
• pp.53-60
• /
• 2022

In a group-testing method, instead of testing a sample, for example, blood individually, a batch of samples are pooled and tested simultaneously. If the pooled test is positive (or defective), each sample is tested individually. However, if negative (or good), the test is terminated at one pooled test because all samples in the batch are negative. This paper considers a queueing system with a two-stage group-testing policy. Samples arrive at the system according to a Poisson process. The system has a single server which starts a two-stage group test in a batch whenever the number of samples in the system reaches exactly a predetermined size. In the first stage, samples are pooled and tested simultaneously. If the pooled test is negative, the test is terminated. However, if positive, the samples are divided into two equally sized subgroups and each subgroup is applied to a group test in the second stage, respectively. The server performs pooled tests and individual tests sequentially. The testing time of a sample and a batch follow general distributions, respectively. In this paper, we derive the steady-state probability generating function of the system size at an arbitrary time, applying a bulk queuing model. In addition, we present queuing performance metrics such as the offered load, output rate, allowable input rate, and mean waiting time. In numerical examples with various prevalence rates, we show that the second-stage group-testing system can be more efficient than a one-stage group-testing system or an individual-testing system in terms of the allowable input rates and the waiting time. The two-stage group-testing system considered in this paper is very simple, so it is expected to be applicable in the field of COVID-19.

• The Transactions of the Korean Institute of Electrical Engineers D
• /
• v.52 no.6
• /
• pp.365-371
• /
• 2003

Input queued switch is useful for high bandwidth switches and routers because of lower complexity and fewer circuits than output queued. The input queued switch, however, suffers the HOL-Blocking, which limits its throughput to 58%. To overcome HOL-Blocking problem, many input-queued switch controlled by a scheduling algorithm. Most scheduling algorithms are implemented based on a centralized scheduler which restrict the design of the switch architecture. In this paper, we propose a simple scheduler called Pipelined Round Robin (PRR) which is intrinsically distributed by each input port. We presents to show the effectiveness of the proposed scheduler.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.28 no.8B
• /
• pp.683-690
• /
• 2003

In this paper we present a theoretical framework for the network design with delay QoS guarantee to a voice at the packet level. Especially, we propose a method for estimating the bandwidth at the ingress edge routers accommodating the voice connections and data sessions in the next-generation If network. First, we describe network architecture for VoIP (Voice over IP) services in the NGN (Next Generation Network). After that, we propose a procedure for dimensioning the bandwidth at the output port of a router that accommodates voice and data traffic using the non-preemptive queuing system with strict priority service scheme. Via numerical experiments we illustrate the implication of the proposition.

• Journal of Korean Society of Transportation
• /
• v.25 no.6
• /
• pp.209-230
• /
• 2007

This paper analyzed the social cost function of a congestion-prone service system, which is developed from the social cost minimization problem. The analysis focused on the following two issues that have not been explicitly explored in the previous studies: the effect of the heterogeneity of value-of-travel-times among customers on the structure of cost functions; and the structure of the supplier cost function constituting the social cost function. The analysis gave a number of findings that could be summarized as follows. First, the social marginal cost for one unit increase in system output having a certain value-of-travel-time is the sum of the service time cost for that value-of-travel-time and the marginal congestion cost for the average value-of-service-time of all the system outputs. Second, the marginal congestion cost equals the marginal supplier cost of system output under the condition that supplier compensates the customers for the changed service time costs which is incurred by the marginal capacity increase necessary for economically facilitating an additional system output. Third, the compensated marginal cost is the multiple of the marginal capacity cost and the inverse of system utilization ratio, if the service time function is homogeneous of degree zero in its inputs.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2003.10a
• /
• pp.373-376
• /
• 2003

This paper presents a new, three-dimensional round-robin scheduler that provides high throughput and fair across in an advanced input-queued packet switch using shared input buffers. We consider an architecture in which each input port group shares a common buffer and maintains a separate queue for each output, which is ratted the distributed common input buffer switch. In an NxN switch, our scheduler determines which queue in the total MxN input queues is served during each time slot where M is the number of common buffers. We suppose that each common buffer has K input ports and K output ports, and manages N output queues. The 3DRR scheduler determines MxK queues in every K(M) cycle when $K\geq$M (K$\leq$M), and provides massively parallel processing for the applications of high-speed switches with a large number of ports. The 3-DRR scheduler can be implemented using duplicated simple logic components allowing very high-speed implementation.

• Journal of Communications and Networks
• /
• v.17 no.3
• /
• pp.286-295
• /
• 2015

Avionic databuses fulfill a critical function in the connection and communication of aircraft components and functions such as flight-control, navigation, and monitoring. Ethernet-based avionic databuses have become the mainstream for large aircraft owning to their advantages of full-duplex communication with high bandwidth, low latency, low packet-loss, and low cost. As a new generation aviation network communication standard, avionics full-duplex switched ethernet (AFDX) adopted concepts from the telecom standard, asynchronous transfer mode (ATM). In this technology, the switches are the key devices influencing the overall performance. This paper reviews the avionic databus with emphasis on the switch architecture classifications. Based on a comparison, analysis, and discussion of the different switch architectures, we propose a new avionic switch design based on a time-division switch fabric for high flexibility and scalability. This also merges the design concept of space-partition switch fabric to achieve reliability and predictability. The new switch architecture, called space partitioned shared memory switch (SPSMS), isolates the memory space for each output port. This can reduce the competition for resources and avoid conflicts, decrease the packet forwarding latency through the switch, and reduce the packet loss rate. A simulation of the architecture with optimized network engineering tools (OPNET) confirms the efficiency and significant performance improvement over a classic shared memory switch, in terms of overall packet latency, queuing delay, and queue size.