大型传感器网络中高速连续数据流的缓存策略  

Buffer Strategy of High Speed and Continuous Data Stream in Huge Sensor Networks

作　　者：汪海山[1] 康建斌[1] 耿立中[1] 马骋[1] 熊剑平[1] 贾惠波[1] 

机构地区：[1]清华大学精密测试技术及仪器国家重点实验室,北京100084

出　　处：《计算机科学》2009年第5期36-38,共3页Computer Science

摘　　要：大型传感器网络是指一种规模巨大的传感器网络,其传感器节点多达上万个(设为n),如果每个传感器的采样率为1ms,每个采样点为24bits,则整个网络产生的数据总流量可高达24nkbps(因为n可达上万,所以流量可达几百Mbps)。如何实现这个巨大网络中传感器节点到基站的高速连续数据流的缓存,是保证整个网络顺利运行的关键所在。采用FPGA和异步FIFO的结合,完美地解决了这个问题。Huge sensor network is a large sensor network in scale. In this system, total sensor nodes can reach up to several ten thousands（suppose the numher is n）. If the sample of each sensor node is 1 millisecond, and every sample point has 24 bits, so the total speed of whole network is 24n kbps （because n may be up to several ten thousands, the speed can reach up to several hundred Mbps）. How to realize the high speed and continuous data stream from sensor nodes to sink nodes in this huge network,is the key to ensure the normal running of the whole network. This paper adoptd a combination of FPGA and asynchronous FIFO,which solves the problem perfectly.

关 键 词：大型传感器网络 高速连续数据流 二级缓存 双缓存 

分 类 号：TP212[自动化与计算机技术—检测技术与自动化装置] TP274.2[自动化与计算机技术—控制科学与工程]