Packet communications system for disaster with functions of multi-gateway

研究并开发了基于蚁群通信方式的大规模自然灾害监测技术。蚁群通信是一种并行分布函数。因此，我们设计了基于该算法的分组终端。在灾难发生时，每个通信终端都需要了解整个网络的拓扑结构，以选择发送消息的最佳路径。该分组终端具有GPS网络拓扑、多网关和GPS秘密共享等功能，能够识别救援队伍和遇险人员的队形。为了减少通信数据，我们测试了语音识别连接喉麦克风，语音合成作为人机界面，并将这些数据输入到分组终端。我们还对该系统进行了现场测试；富山湾浮标上装有8个原型包终端的搜索和救援系统。项目总结如下：1)语音识别连接喉式麦克风，可将数据减少1/ 10000左右，抗现场噪声能力强，在灾难发生时对救援人员非常有用。2)对多浮标上多接收的多数决定译码方法进行了测试，发现该算法在海上是非常有用的。3)利用GPS数据，通过分组终端识别网络拓扑。

We have studied and developed state of art technology based on communication style between ants for large scale natural disaster. The ant communications looks one of the parallel distributed functions. So we have designed packet terminal based on such algorithm. During disaster each communications terminal needs to be ware of the entire network topology to select the best path to send message. The packet terminal recognized formations of rescue team and victims with the functions of network topology from GPS, of multi- gateway and of secret sharing with GPS. In order to reduce communication data, we have tested voice recognition connected Laryngeal microphone, and voice synthesis as a man-machine interface and inputed these data to packet terminal. We have also performed a field test such a system; search and rescue system with 8 prototype packet terminal on buoys at Toyama bay. The project summary were as follows.1) Voice recognition connected Laryngeal microphone might reduce data about 1/10,000, and very powerful against field noise, so it looks very useful for rescue person during disaster.2) Majority-decision decoding method with multi-reception on the multi-buoy is tested and we find that might be very useful algorithm on the sea.3) With GPS data, we can recognized network topology through Packet terminals.