Ji-Seung Nam

Fast required bandwidth estimation technique for network adaptive streaming

Research into high quality media streaming is regarded as an important area of information technology with the rapid development of network technology and digital A/V home appliances. However, a scheme that can flexibly deal with extreme network dynamics is required to realize high quality media streaming service because the modern internet is designed for best effort services. In this paper, we design and implement a scheme that controls quality of service (QoS) based on whether the network can accommodate the required bandwidth of the media. The relative one-way delay (ROWD) trend is used to determine whether the required bandwidth is available and the temporal coding of scalable video coding (SVC), which can be easily applied to real-time systems, is introduced to control QoS. Experimental results conducted with a fiber-to-the-home (FTTH) network show that the proposed system appropriately adapted to variations in network bandwidth and the service quality improved.

Cluster and Location Based Overlay Multicast in Mobile Ad Hoc and Sensor Networks

We propose a novel overlay multicast protocol over wireless sensor network, CLOM (cluster and location based overlay multicast) which use both cluster based approach and location based approach. A cluster based approach is used to manage multicast group consisting of relatively large number of nodes while a location based approach is used to construct a multicast tree with rapidity and low control overhead by using geometric information taken from GPS device. Additionally, the information of damping area and the scheme for reducing the CH node discovery time and control overhead have been introduced into CLOM to improve the performance of the tree. The simulation results showed that CLOM works properly with low data distribution delay and low control overhead where the number of multicast group members is relatively high.

Image Pattern Recognition in Natural Environment Using Morphological Feature Extraction

The gray-scale morphological Hit-or-Miss transform is theoretically invariant to vertical translation of the input function, which is analogous to gray-value shift of the input images. Designing optimal structuring elements for the Hit-or-Miss transform operator is achieved by neural network learning methodology using a shared-weight neural network (SWNN) architecture. Early stage of the neural network system performs feature extraction using the operator, while the late stage does classification. In experimental studies, this morphological feature-based neural network (MFNN) system is applied to location of human face and automatic recognition of vehicle license plate to examine the property of the operator. The results of the experimental studies show that the gray-scale morphological Hit-or-Miss transform operator is reducing the effects of lighting variation.

Virtual and Pervasive Smart Home Services Using Tangible Mixed Reality

This paper proposes a new way of providing virtual smart home services using tangible mixed reality (MR), which provides more cost-effective and reliable visualization and simulation of the existing pervasive environment. One of the main characteristics of the proposed approach is capable of embedding virtual objects into the real environment such that it is easy to test the feasibility of the existing pervasive smart home. The paper also presents how to support those services by linking contexts to virtual objects and their tangible behaviors. We will show the effectiveness and advantage of the proposed approach by demonstrating a tangible smart home testbed. It has been also applied to the tangible virtual studio authoring and interaction.

Efficient and Automatic Faces Detection Based on Skin-Tone and Neural Network Model

In this paper, we consider the problem of detecting the faces without constrained input conditions such as backgrounds, luminance and different image quality. We have developed an efficient and automatic faces detection algorithm in color images. Both the skin-tone model and elliptical shape of faces are used to reduce the influence of environments. A pre-built skin color model is based on 2D Gaussian distribution and sample faces for the skin-tone model. Our face detection algorithm consists of three stages: skin-tone segmentation, candidate region extraction and face region decision. First, we scan entire input images to extract facial color-range pixels by pre-built skintone model from YCbCr color space. Second, we extract candidate face regions by using elliptical feature characteristic of the face. We apply the best-fit ellipse algorithm for each skin-tone region and extract candidate regions by applying required ellipse parameters. Finally, we use the neural network on each candidate region in order to decide real face regions. The proposed algorithm utilizes the momentum backpropagation model to train it for 20*20 pixel patterns.The performance of the proposed algorithm can be shown by examples. Experimental results show that the proposed algorithm efficiently detects the faces without constrained input conditions in color images.

A Fuzzy-Based Adaptive Streaming Algorithm for Reducing Entropy Rate of DASH Bitrate Fluctuation to Improve Mobile Quality of Service

Dynamic adaptive streaming over Hypertext Transfer Protocol (HTTP) is an advanced technology in video streaming to deal with the uncertainty of network states. However, this technology has one drawback as the network states frequently and continuously change. The quality of a video streaming fluctuates along with the network changes, and it might reduce the quality of service. In recent years, many researchers have proposed several adaptive streaming algorithms to reduce such changes. However, these algorithms only consider the current state of a network. Thus, these algorithms might result in inaccurate estimates of a video quality in the near term. Therefore, in this paper, we propose a method using fuzzy logic and a mathematics moving average technique, in order to reduce mobile video quality fluctuation in Dynamic Adaptive Streaming over HTTP (DASH). First, we calculate the moving average of the bandwidth and buffer values for a given period. On the basis of differences between real and average values, we propose a fuzzy logic system to deduce the value of the video quality representation for the next request. In addition, we use the entropy rate of a bandwidth measurement sequence to measure the predictable/stabilization of our method. The experiment results show that our proposed method reduces video quality fluctuation as well as improves 40% of bandwidth utilization compared to existing methods.

An efficient scheduling multimedia transcoding method for DASH streaming in cloud environment

As a result of technological evolution, streaming service providers have been dealing with the problem of delivery multimedia content to the diversity of devices with different resolutions. This issue can be solved by using dynamic adaptive streaming over hypertext (DASH) transfer protocol. However, a transcoding job in DASH requires a lot of computation resource which could lead to delaying the starting of multimedia streaming. Recently, new studies have addressed novel scheduling methods on video transcoding, but those research did not solve the problem entirely, such as the solution did not concern server performance or speed connection between a server and its requested users. Moreover, the load and speed connection status of the data servers is often unstable, leading to increasing the starting delay. So in this article, we solve such problem by modeling transcoding jobs in the form of an optimization problem and propose an algorithm to find an optimal schedule to transcode video source files. In which, we use moving average method to find average points for a short period to deal with server state changes. In the experiment, we implement our proposed method with DASH to demonstrate the effectiveness of the optimization scheduling method. In the system, we create several servers running on the Docker platform to simulate a cloud environment. Experimental results show that our methodology reduces the time of the transcoding process up to 30% compared to existing research.

Induced Packet Loss Reduction by Max-Heap Overlay Construction Based on the Bandwidth-Latency Product Measures

Constructing max-heap-form overlay governed by node gain scores (NGSs), earned from requested bandwidth discrepancy ratio (BDR) with respect to available bandwidth, and latency discrepancy ratio (LDR) between the nodes and the source, reduces induced packet loss caused by schemes neglecting these parameters in constructing the tree. This proposal positions nodes using NGSfunction governed by available bandwidth, requested bandwidth, proposed latency to prospective parent, and best latency as advised by the source. BDR and LDR are set with a and (1,000-alpha) respective weights, and arbitrarily chosen a between 0 and 1,000. NGSs, used as nodelDs, maintain uniqueness and balance between the critical factors. Max-heap-form tree is constructed assuming that all the nodes possess NGS less than the source. For load balancing, parent-node can not accept another child until all its able siblings acquire same number. This scheme out-performs the others, like BASE, in packet delivery, link stress, and end-to-end delay.

Design and Deployment of Low-Delay Hybrid CDN---P2P Architecture for Live Video Streaming Over the Web

In the recent years, with the development of high-speed and broadband networking, the content delivery service has been grown up widely. There are a lot of providers for online streaming via content delivery network (CDN) and peer to peer (P2P) network, each having its own set of advantages and disadvantages. CDN provide excellent quality to end-users when the work load is within the provisioning limits. However, CDN servers are expensive to deploy and maintain. While P2P network has better scalability and less deployment costs, but the instability of dynamic peers, and the low performance when participated peers are insufficient are a bottleneck. So, by combining advantages of two system we will take advantage of both. In this paper, we propose the hybrid CDN–P2P structure for live video streaming over the Web in order to decrease the number of requests to CDN servers that means reduce the cost of transmission. Also by set up short timeout for request to server and split video into small chunk, our system has smaller end-to-end delay than unconnected mesh system in Seyyedi and Akbari (Hybrid CDN–P2P architectures for live video streaming: comparative study of connected and unconnected meshes. In: International symposium on computer networks and distributed systems, 2011). By set up a reality, we compare two metrics: the number of requests to CDN server and end-to-end delay with previous researches. Through experiment, we will show that a hybrid CDN–P2P approach is very effective approach in providing a live streaming service to the public.

A New Prefetch Policy for Data Filter Cache in Energy-Aware Embedded Systems

As process technology scales down, energy consumption in embedded processors becomes a crucial issue. In embedded processors, data cache accounts for a considerable portion of total dynamic energy consumption. In this paper, we propose a novel energy-efficient Prefetch Data Filter cache (PDF-cache) technique that enables filtering cache accesses by using the access pattern to data cache when a loop instruction is executed. In the proposed architecture, the accesses to data cache are partly migrated to the PDF-cache, which has a very small size. According to our experimental results, the proposed cache architecture with PDF-cache can reduce the dynamic energy consumption compared to the baseline about 7.1% on average with little storage overheads.