Jin-Suk Ma

KVM-QEMU Virtualization with ARM64bit Server System

In the conventional x86 or x86-64 bit system, virtualization is commonly achieved with KVM or Xen which is widely adapted in many experimental or commercial server system. At now, It is common sense that high and mid volume servers are virtualized with Xen, mid and small volume servers are virtualized with KVM. Recently, the microserver system concept was introduced, which contains low power and multicore ARM64bit sever SoCs. We present the KVM-QEMU virtualization technique and process in a real ARM64bit microserver system with ubuntu 14.04 ARM64 bit edition root file system.

The Virtualization and Performance Comparison with LXC-LXD in ARM64bit Server

The microserver with low power and high efficiency ARM64bit cores is now preparing the future eco-friendly microserver market. Virtualization is a promoting software technology in commercial cloud application. Many experimental and enterprise servers have adapted with XEN, KVM-QEMU, LXC and etc to achieve a server virtualization. This paper deals with the realization and performance measure in ARM64bit server with LXC-LXD which is now fully promoted by Canonical co. We presented the essential steps of utilizing LXC-LXD in real ARM64bit server platform not emulator and measured the performance with XEN in view of sysbench benchmark. In ARM64bit, LXC-LXD virtualization could be directly achieved with kernel parameter adjusts and simple command operations over XEN hypervisor. But it showed the competitive performance results over the more sophisticated XEN.

Application of feedforward neural network for the deblocking of low bit rate coded images

In this paper, we propose a novel post-filtering algorithm to reduce the blocking artifacts in block-based coded images using block classification and feedforward neural network. This algorithm exploited the nonlinearity property of the neural network learning algorithm to reduce the blocking artifacts more accurately. At first, each block is classified into four classes; smooth, horizontal edge, vertical edge, and complex blocks, based on the characteristic of their discrete cosine transform (DCT) coefficients. Thereafter, according to the class information of the neighborhood block, adaptive feedforward neural network is then applied to the horizontal and vertical block boundaries. That is, for each class a different multi-layer perceptron (MLP) is used to remove the blocking artifacts. Experimental results show that the proposed algorithm produced better results than those of the conventional algorithms both subjective and objective viewpoints.

The LXC-LXD Virtualization in ARM64bit X-Gene2 Server

The server system with low power and high efficiency ARM64bit cores is now accelerating the versatility of usages such as the low power utility, cloud computing and etc. But before the real advent using ARM64 SoC in server market, ARM64bit based server system has to satisfy the stringent operation prerequisite of the server operator. Various server virtualization techniques such as XEN, LXC-LXD, KVM-QEMU are applicable in ARM64 server with the help of world-wide developers as intel x86 based case. LXC-LXD is fully promoted by Cannonical Co. which can house the multiple applications in a single container instance differentiating the docker container. In this paper we dealt with LXC-LXD 2.0 virtualization in a real ARM64bit server reference board APM X-Gene2. The detailed process and linux kernel configuration are given in main part of this paper. Finally LXC-LXD is compared with XEN hypervisor in view of sysbench and STREAM benchmark.

Picture quality improvement in low bit rate video coding using block boundary classification and simple adaptive filter

In this paper, we propose a novel post-filtering algorithm with low computational complexity that improves the visual quality of block-based coded images using block boundary classification and simple adaptive filter (SAF). In this algorithm, each block boundary is classified into smooth or complex sub-region, and the existence of blocking artifacts is determined using blocky strength for smooth-smooth sub-regions. Simple adaptive filtering is processed adaptively in each block boundary. That is, a nonlinear 1-D 8-tap filter is applied to smooth-smooth sub-regions with blocking artifacts, and for smooth-complex or complex-smooth sub-regions, a nonlinear 1-D variant filter is applied to block boundary pixels so as to reduce the blocking and ringing artifacts. And for complex-complex sub-regions, a nonlinear 1-D 2-tap filter is only applied to adjust two block boundary pixels so as to preserve the image details. Experimental results show that the proposed algorithm produced better results than those of the conventional algorithms both subjective and objective viewpoints.

A Robust Design, Method for a Long Dead-time System with an Integral Mode

In this paper, we present a robust controller design method that can not only deal with a constant time delay plant but a1so the uncertain time delay one. For a constant time delay plant, the proposed DTC can independently adjust the set point response and the disturbance rejection response without any stability constraint. And in the uncertain time delay case, one can process the control design step with uncertainty norm bound. To verify real effectiveness, theoretical analysis and simulation results are given.