Mike Myung-Ok Lee

High-Speed Networks and Multimedia Communications

Various Internet applications involve multiple parties and usually adopt a one-to-many communication paradigm (multicast). The presence of multiple receivers in a multicast session rises the problem of inter-receiver fairness. Transmitting with a rate which matches the slowest receiver will limit the throughput of other receivers and thus their satisfaction. A multi-rate mechanism where the receivers are distributed into subgroups with similar capacities, can improve the inter-receiver fairness for multicast sessions. In this paper, we deal with the problem of receivers partitioning and propose a simple algorithm based on the receivers RTT variations where an explicit estimation of the receivers capacities is avoided. Our partitioning algorithm, although simple, performs an on-the-fly partitioning depending on the receivers’ feedback. We show that our partitioning algorithm approximates and in many cases, achieves the optimal solution with a minimum computation effort.

A new tuning maximum congestion window for improving TCP performance in MANET

In terms of improving TCP performance in Mobile Ad Hoc-NETwork (MANET), it is well known that small congestion window size of TCP connection improves TCP performance. We propose a new TCP algorithm to improve TCP performance in MANET. The basic idea of our approach is that TCP receiver estimates the optimum window size and then sets congestion window limit to an optimal value by the advertised window. It is found that our algorithm, using ns-2 simulator, improves TCP performance up to 60% in various topologies in MANET.

AN EFFICIENT LOAD-BALANCING ALGORITHM FOR SUPPORTING QOS IN MANET

A Mobile Ad hoc Network can be considered as an autonomous distributed system that consists of a set of identical mobile nodes that move arbitrarily and use wireless links to communicate with other nodes that reside within its transmission range. In our work, we analyze the performance of the original IEEE 802.1 1 b wireless local area networks and we present an efficient load-balancing scheme in MAC for support QoS over MANET. Simulation results aimed to highlight the capability of our algorithm to reduce packet lose rate and to increase throughput in the MAC layer are provided.

Design of a New IFFT/FFT for IEEE 802.11a WLAN Based on the Statistics Distribution of the Input Data

This paper presents a new design method of IFFT/FFT minimizing the quantization errors based on the statistical analysis of the input data distribution in 802.11a WLAN. The proposed algorithm decides the twiddle coefficients by applying the statistically calculated spreading coefficient from the data distributions which is a ratio of the data amplitude of the ancestor and the present stage. The data was leaned toward to one at each input stage, and is expanded by a new twiddle coefficient. We implemented this algorithm with radix-22 SDF(single-path delay feedback) architecture [1]. Both IFFT and FFT processor can share the circuit modules by the symmetric nature of them. We designed a base band processor including IFFT for TX and FFT for RX with I/O of 10 bits on a platform. It showed that the maximum quantization error is 0.0021. This is same as that the proposed algorithm improved the quantization errors with the linear bit extension by 3 bits. It yields reducing quantization error 8dB and smaller hardware size.

New MPLS Switch Architecture Supporting Diffserv for High_Speed Switching and QoS

We propose a new architecture of the MPLS switch supporting Differentiated Services in the MPLS-based network. The traffic conditioner consists of a classifier, a meter, and a marker. The VOQ-PHB module which combines input queue with each PHB queue is implemented to utilize the resources more efficiently, employing the Priority-iSLIP scheduling algorithm to support high-speed switching. The proposed MPLS switch architecture is modeled by C++ and synthesized by Very High Speed Integrated Circuits Hardware Description Language(VHDL), verified and then implemented by commercialized CAD tools to justify the validity of the proposed hardware architecture.

Leaky Bucket Based Buffer Management Scheme for TCP/IP Traffic over GFR Service

The ATM Forum recently introduced the Guaranteed Frame Rate (GFR) service category. GFR service has been designed to support classical best effort traffic such as TCP/IP. The GFR service not only guarantees a minimum throughput, but also supports fair distribution of available bandwidth to competing VCs. In this paper, we propose a new buffer management algorithm based on leaky bucket to provide a minimum cell rate guarantee and improve fairness. The proposed algorithm reduces complexity and the processing overhead of the leaky bucket algorithm to allow its easy implementation in hardware

Area Efficient and Low Power Pipelined IIR Filter Design for Intelligent Integrated Photonic System

An angle and radius constrained IIR filter design approach using modified Remez exchange algorithm and least squares algorithm is designed to avoid tight pole-crowding in pipelined filters, resulting in improved frequency responses and reduced coefficient sensitivities. Experimental results demonstrate that our proposed method leads to chip area reduction by 33% and low power by 45% against the conventional method.

A new user-assisted segmentation and tracking technique for an object-based video editing system

This paper presents a semi-automatic segmentation method which can be used to generate video object plane (VOP) for object based coding scheme and multimedia authoring environment. Semi-automatic segmentation can be considered as a user-assisted segmentation technique. A user can initially mark objects of interest around the object boundaries and then the user-guided and selected objects are continuously separated from the unselected areas through time evolution in the image sequences. The proposed segmentation method consists of two processing steps: partially manual intra-frame segmentation and fully automatic inter-frame segmentation. The intra-frame segmentation incorporates user-assistance to define the meaningful complete visual object of interest to be segmentation and decides precise object boundary. The inter-frame segmentation involves boundary and region tracking to obtain temporal coherence of moving object based on the object boundary information of previous frame. The proposed method shows stable efficient results that could be suitable for many digital video applications such as multimedia contents authoring, content based coding and indexing. Based on these results, we have developed objects based video editing system with several convenient editing functions.

MPLS switch architecture supporting Diffserv for high-speed switching and QoS

In this paper, we propose the architecture of the MPLS switch supporting Differentiated Services in the MPLS-based network. The traffic conditioner consists of a classifier, a meter, and a marker. The VOQ-PHB module which combines input queue with each PHB queue is implemented to utilize the resources more efficiently, employing the Priority-iSLIP scheduling algorithm to support high-speed switching. The proposed MPLS switch architecture is modeled and synthesized by Very High Speed Integrated Circuits Hardware Description Language (VHDL), verified and then implemented by commercialized CAD tools to justify the validity of the proposed hardware architecture.

Leaky Bucket Based Buffer Management Scheme to Support Differentiated Service in Packet-Switched Networks

The ATM Forum recently introduced the Guaranteed Frame Rate (GFR) service category. GFR service has been designed to support classical best effort traffic such as TCP/IP. The GFR service not only guarantees a mini-mum throughput, but also supports fair distribution of available bandwidth to competing VCs. In this paper, we propose a new buffer management algorithm based on leaky bucket to provide a minimum cell rate guarantee and improve fairness. The proposed algorithm reduces complexity and the processing overhead of the leaky bucket algorithm to allow its easy implementation in hardware.