Search

The amount of information traffic is on increase due to the increase of the number of electric components in vehicles. In this situation, IEEE Audio/Video Bridging (AVB) is promising technology in In-Vehicle Network (IVN) since it guarantees high level of Quality of Service (QoS) for real-time applications. Now, new AVB standards that add time-triggered traffic is discussed. Time-triggered traffic uses time division multiplexing access (TDMA) to guarantee the performance of time-critical traffic like control traffic. In other words, time-triggered traffic would be sent when their time window comes. However, it is inevitable that the time performance of AVB traffic is deteriorated by adding time-triggered traffic since the transmission of time-triggered traffic is prior to the transmission of AVB traffic. So, in this paper, a Dual-path method is proposed to improve the time performance of AVB traffic by using Redundancy path, which is presented in IEEE P802.1CB. In Dual-path method, the additional path, which is used as a backup path in IEEE P802.1 CB, is used as an independent path. Therefore, traffic is divided and transmitted through two of independent paths respectively. By distributing traffic, the tiem performance of AVB traffic could be enhanced. To verify the advantage of Dual-path method, three scenarios of traffic allocation are configured. These scenarios are mathematically analyzed. Based on results, the optimal traffic allocation is suggested.

CONceptual-BAsed image retrieval system (CONBA) supports the object recognition and retrieval based on object modeling and machine learning. The system, being independent of a domain, has an advantage in that the probability of the image recognition and the retrieval would be improved on the basis of a continuous learning theory of the image object. This system also includes a design for supporting a conceptual object on the basis of the set-based theory to extract the conceptual object and support it in user's enquires.