• Proceedings of the Korea Information Processing Society Conference
• /
• 2009.04a
• /
• pp.1178-1181
• /
• 2009

Geographic forwarding algorithms을 사용하는 Georouting protocol에서는 route maintenance을 위해서 고정된 주기마다 beacon message을 이웃노드에 전송하는 fixed periodical beacon based route algorithm을 사용하여 mobile node 정보를 습득한다. Fixed periodical beacon scheme에서의 기존 연구 이슈는 이웃노드 정보의 신뢰성을 유지하기 위해서 다양한 mobility environment 환경에 맞도록 beacon 전송주기를 적절히 결정하는 연구가 진행되어 왔다. 이와 같은 연구에서는 long periodical interval 사용 시 이웃노드의 위치정보 error을 발생시킬 확률이 높으며 또한 short periodical interval 의 경우 네트워크 내에 high route overhead 을 유발시키는 단점을 가지고 있다. 따라서 고정된 주기 방법은 다양한 mobility environment에 잘 적용되지 못하며 또한 mobility environment 에서 발생하는 route maintenance 내에 이웃노드 정보에 대한 불확실성 문제가 빈번이 발생하여 경로 설정에 잘못된 정보를 제공하여 경로 신뢰성(path reliability)과 낮은 전송률 (transmission rate)을 야기한다. 본 논문에서는 이런 이웃노드 정보의 불확실성 문제를 극복하기 위해서 mobile node가 스스로 자신의 위치를 체크하여 routing table을 능동적으로 갱신하는 방법을 이용하여 이 문제를 극복할 수 있도록 하였다. 모의 실험은 ns2를 이용하여 실시하였으며 결과는 low/high mobility scenario에서는 기존 방법 보다 routing overhead 을 줄이면서 높은 전송률 (transmission rate)을 보인다.

• Smart Media Journal
• /
• v.12 no.5
• /
• pp.28-35
• /
• 2023

In this study, we propose a new light-weight model RoutingConvNet with fewer parameters to improve the applicability and practicality of speech emotion recognition. To reduce the number of learnable parameters, the proposed model connects bidirectional MFCCs on a channel-by-channel basis to learn long-term emotion dependence and extract contextual features. A light-weight deep CNN is constructed for low-level feature extraction, and self-attention is used to obtain information about channel and spatial signals in speech signals. In addition, we apply dynamic routing to improve the accuracy and construct a model that is robust to feature variations. The proposed model shows parameter reduction and accuracy improvement in the overall experiments of speech emotion datasets (EMO-DB, RAVDESS, and IEMOCAP), achieving 87.86%, 83.44%, and 66.06% accuracy respectively with about 156,000 parameters. In this study, we proposed a metric to calculate the trade-off between the number of parameters and accuracy for performance evaluation against light-weight.

• The KIPS Transactions:PartC
• /
• v.11C no.6 s.95
• /
• pp.799-806
• /
• 2004

A growing diversity of pervasive devices is gaming access to the Internet and other information. However, much of the rich multimedia contents cannot be easily handled by the client devices because of the limited communication, processing, storage and display capabilities. The in-tegration of voice, data and video service modified the target of networking technologies. Networks must have some the capabilities for in-tegration of various services and also for QoS support as required by each of those service. Because of these reasons, we developed EAGIS(Efficient ATM Gateway for real time Internet Service) to provide seamless multimedia service between the ATM network and the Internet. EAGIS consists of the interworking unit, content server, transcoding server, and the serveice broker to provide seamless multimedia service be-tween the ATM network and the Internet. In this paper, we design the architecture and transcoding service scenario of the EAGIS. When the RTP is used for the bi-directional communication, transcoding time is configured by the time-stamp of RTCP. When HTTP is used for unidirec-tional communication, self-timer is used. By using these reference time, standard transcoding method is applicable according to the frame trans-mission rate and network traffic load. And we can also assure the QoS of the multiple users` effective bandwidth by our algorithm.