Satoshi Yoshizawa

IMPACT: an interactive natural-motion-picture dedicated multimedia authoring system

A new approach to achieving a natural-motion-picture dedicated multi-media authoring system is proposed. The main point of this approach, discussed in this paper, is that the user’s environment or interface is improved to encourage user’s creativity, with image processing and recognition technology. According to the discussion, a prototype motion picture authoring system dtat has several image-processing functions is developed. The newly developed functions include object extraction of the picture, semi-automatic visualization of motion pictures structure, and certain descriptions of the scene. Result of using the prototype shows the appropriateness of the proposed approach.

Intelligent seamless vertical handoff algorithm for the next generation wireless networks

With the rapid development of wide variety of wireless technologies, it is envisioned that the future network will be a convergence of various wireless access technologies providing users with the best anywhere, anytime connections to the Internet. In such environment, it is a critical problem to enable a mobile node to intelligently select wireless access networks and achieve seamless handoff performance. This paper proposes a vertical handoff algorithm which enables a mobile node to intelligently select wireless access network among multiple access technologies, synthetically considering network characteristics and status, such as access bandwidth, response time, data loss, network congestion status, etc, along with access cost and received signal strength (RSS). With the proposed algorithm, wireless network utilization can be optimized according to users strategy and network conditions, and real-time applications can be maintained without packet loss during handoff process. A prototype system is implemented and experimental results are analyzed.

QoS-aware multicast for internet video applications

In this paper we extend our work in joint design of source rate control and congestion control for unicast video streaming, and propose a cross-layer design approach for Internet video multicast applications. By jointly designing source rate control at application layer and multicast congestion control at transport layer, our QoS-aware multicast approach can provide better QoS support for multimedia applications and significantly improve the playback quality, while maintaining good long-term TCP friendliness.

A repair mechanism for reliable multicast in Integrated Communication and Broadcast Networks

With the rapid growth of bandwidth requirement for mobile multimedia services, broadcast communication infrastructure is desired as one of the best solutions for avoiding communication congestion. Integrated Communication and Broadcast Network (ICBN), consisting of broadcast network and communication network, is one of the solutions. However, in reliable multicast service over ICBN, feedbacks from receivers cause feedback implosion problem. Also, because the broadcast network resources are shared by all the users in a system, data retransmission through broadcast network decreases system throughput. To avoid feedback implosion and congestion of broadcast network, we propose a local repair mechanism for reliable multicast in ICBN, which assures complete data reception by recovering the missing data from neighbor terminals. The experimental results show that our proposal avoids feedback implosion and improves the system throughput.

SIP-based Group Service Improvement in CDMA2000 Network

SIP-based multi-party conference service (MFCS) of IETF is adopted by 3GPP2 as conferencing solution. In MFCS, the service content received by multiple participants is the same. So, it is better to use multicast to support MFCS in 3G network, where air resources are limited. In 3GPP2, broadcast multicast service (BCMCS) is standardized allowing optimization of air resources for delivery of uniform content to multiple access terminals in CDMA2000 communication system. In this paper, a new solution is proposed, wherein service mapping is set up between BCMCS and MPCS in CDMA2000 network. By utilizing BCMCS, the radio and access networks resources of the CDMA systems could be saved for MPCS.

Performance prediction method for Web-access response time distribution using formula

Since there are many causes that can affect Web response time, it is difficult to predict. NetForecast Inc. has proposed a formula to calculate the response time of an application over the TCP/IP layer. This paper proposes a method to generate a probability function for the response time by using NetForecasts formula. This calculation method uses two input parameters: the file-size cumulative distribution function (CDF) and the delay probability density function (PDF). In the formula, the data sampled from the file-size CDF is input as the payload-length parameter, and the data sampled from the delay PDF is input as the round-trip delay parameter. Processing these parameters using the performance prediction method, we can generate the response-time CDF.

Simulation and Basic Experiment of Inter Radio System Handover for Cognitive Radio

Cognitive radio system consists of multiple wireless accesses that cover overlapping area, and cognitive terminals that use one or more of the wireless accesses simultaneously. In this paper, we propose the architecture of the cognitive radio system, and the inter-system handover protocols. In the proposed architecture, each cognitive terminal, that has multiple radio systems, operates with a single local IP address. The control sequence and packet format are designed to achieve fast handover among the radio systems. Based on the proposed architecture, we have developed both a simulator and a testbed system with WiMAX and wireless LAN. Through simulation and testbed, we have proved that stream data can be delivered continuously without any effects of the handover. Also, we present the evaluation result of the end-to-end latency on the testbed system.

Trick Play Function for VOD with SVC source

Networked video applications have become more and more popular. However, most commercially available video on demand (VOD) solutions still lack trick play features (fast forward, fast rewind etc.), because it usually needs high processing complexity and huge consumption of network bandwidth to support trick play function. On the other hand, H.264 Scalable Video Coding (SVC) will be adopted in more and more video contents, and have great impact on the video application. In this paper, we propose a mechanism to support trick play function for H.264 SVC video source. According to different fast speed, appropriate combination of temporal and SNR layers are selected and transmitted in server side. Performance evaluation results show that the proposed mechanism can effectively reduce network bandwidth consumption, while keeping satisfied quality of experience (QoE) for users.

Local repair mechaism for reliable multicast in Integrated Communication and Broadcast Networks with scalability

With the rapid growth of bandwidth requirement for mobile multimedia services, broadcast communication infrastructure is desired as one of the best solutions for avoiding communication congestion. Integrated Communication and Broadcast Network (ICBN), consisting of broadcast network and communication network, is one of the solutions. However, in reliable multicast service over ICBN, feedbacks from receivers cause feedback implosion problem. Also, because the broadcast network resources are shared by all the users in a system, data retransmission through broadcast network decreases system throughput. To avoid feedback implosion and congestion of broadcast network, we propose a local repair mechanism for reliable multicast in ICBN, which assures complete data reception by recovering the missing data from neighbor terminals. The experimental results show that our proposal avoids feedback implosion and improves the system throughput.

Proposal and Testbed System of Inter Radio System Switching for Cognitive Radio

Cognitive radio system consists of multiple wireless accesses that cover overlapping area, and cognitive terminals that use one or more of the wireless accesses simultaneously. In this paper, we propose the architecture of the cognitive radio system, and the inter-system handover protocols. In the proposed architecture, each cognitive terminal, that has multiple radio systems, operates with a single local IP address. The control sequence and packet format are designed to achieve fast handover among the radio systems. Based on the proposed architecture, we have developed a testbed system. On this system, we have proved that data can be delivered continuously and radio systems can be switched correctly without any packet loss. Also, we present the evaluation result of the end-to-end latency on the testbed system.