注册日期: Aug 2002
来自: 湖南
发帖数: 742

通讯#4：Multiple Detection Search

通讯 #4
2000年6月19日
多重检测搜索(// Multiple Detection Search)
作者：Stan Corroyer（来自法国Ecole Polytechnique的访问学者）
翻译：holyphoenix ([email protected])

自从SETI@home项目开始以来，SETI@home用户利用屏保客户程序已经发现了大约6千万个高斯信号。这些高斯信号的大部分是由于宇宙背景噪音所引起的。虽然如此，如果地外文明向我们发送强烈的信号的话，我们将能够通过在两次不同的时刻且在大致相同的位置上检测到它。

分析SETI@home高斯信号计算结果的第一步是，通过搜索整个数据库以搜寻成对的或者三元的、将能指示出类似地外文明信号的高斯信号。因为搜索超过6千万个高斯信号需要一定的时间，所以初始的搜索只包含最顶级的10000个信号，也就是拥有最佳高斯适合度(// gaussian fits)的最强烈的那些信号。初始搜索找寻在同一区域被发现的高斯信号——例如，天空坐标(// sky coordinates)（赤经和赤纬）两者最多有5弧分（这是望远镜束宽的一半）的偏差——而且相互之间的频率相差在100kHz以内(// frequencies within 100kHz of each other)。

三个令人感兴趣的“三元信号”在这次搜索中被找出来，它们被更加细致地进行分析，通过绘制数据库中所有其它结果的时间和频率图，并将我们所发现的这几个三元信号的频率居中。

下面的图片来自第一个令我们感兴趣的三元信号。你可以点击这些图片以放大它们。

根据这些图表，看来很多被检测到的信号是由于RFI（射电频率干扰）所引起的。这就是为什么在某个特定的频率上会有很多点（差不多形成了一条线）贯穿了整个时间标尺。在RFI页面上可以看到关于窄带干扰(// narrowband interference)得更详细的信息。在第一幅图片的右边，宽带干扰明显的是图上那两条从上到下的粗条纹。

在仔细的分析之后，其它两个“三元信号”被证明也是RFI，就像下图所显示的那样。注意在第二幅图右边的绿色矩形。这些都是由一个单独的workunit所产生的。我们还不能确定为什么从这个workunit返回了这些糟糕的结果，我们将进一步对之进行探测。另一个类似的区域在第一幅图的左上角可以朦胧的看到。在窄带信号周围的大块黑色区域是因为客户端程序由于射电频率干扰而导致达到了计算结果的极限所致(//the client reaching the result limit due to the radio frequency interference)，由这个workunit中的最高能量的信号所导致的结果(//resulting in only the highest powered signals being returned for that workunit)。在这个例子中，最高能量的信号在这幅图中被限于窄带RFI可视性(//the highest powered signals were limited to the narrowband RFI visible in the plot)。这也可以解释在上面一图图中第一幅图中的方形区域。这样，RFI在黑色区域中强烈到超过计算结果极限(// result limit)，但是在这区域上下去足够弱以允许其他信号能够被返回。

嗯……在我们的初始分析中，我们还没有找到外星人……我们将继续通告您关于我们所继续进行的分析的事情(to keep you informed as our analysis continues)！

万维网网址：http://setiathome.ssl.berkeley.edu/index.html
版权所有：Copyright ©2001 SETI@home


// 原文网址：http://setiathome.ssl.berkeley.edu/...ewsletter3.html
// 欢迎大家对我的翻译批评指正！！！
// 我在译文中用（//）表示的就是我所拿不准如何翻译的部分。
// 同时希望大家能对其他译文指教！


Newsletter #4
June 19, 2000
Multiple Detection Search
Stan Corroyer - Visiting Scholar from the Ecole Polytechnique (France)

Since the beginning of the SETI@home project, SETI@home users have found about 60 million gaussians using the screensaver client program. A great deal of these gaussians are the result of background noise in the universe. However, if extraterrestrials were sending us a strong signal, we should be able to detect it coming from roughly the same place in the sky on two different occasions.

The first step in analysing the SETI@home gaussian results is to search through the database for pairs or triplets of gaussians that would indicate such an extraterrestrial signal. Since searching over 60 millions gaussians would take a while, the initial search consisted of only the top 10000 ones, that is, the strongest ones with best gaussian fits. The initial search looked for gaussians that were found in the same area - i.e. both sky coordinates (right ascension and declination) at most 5 arcminutes apart (this is half a beam width) - and frequencies within 100kHz of each other.

Three interesting "triplets" came out of this search, and all of them were more carefully analysed by plotting a time and frequency map of all other results in the database, centered on the specified frequency at which the triplet was found.

The pictures below come from the first triplet that interested us. You can click on each of the images to enlarge them.

From the graphs, it appears the detected signals resulted from RFI (Radio Frequency Interference). This is why there are so many dots (almost making a line) at a particular frequency for the entire time scale. See the narrowband interference section on the RFI page for more details. On the right side of the first picture, broadband interference is evident from the two thick stripes running down the graph.

After careful analysis, it was determined that the two other "triplets" seemed to be RFI too, as displayed by the pictures below. Notice the green rectangle on the right portion of the second plot. These are the results from a single workunit. We are not yet sure why bad results were returned for this workuit and are exploring it further. Another such area can be seen faintly in the upper left portion of the first plot. The large black area surrounding the narrowband signal is caused by the client reaching the result limit due to the radio frequency interference, resulting in only the highest powered signals being returned for that workunit. In this instance, the highest powered signals were limited to the narrowband RFI visible in the plot. This also explains the square blank area in the first plot of the first set, above. In this case, the RFI was strong enough to overrun the result limit in the blank area but weak enough above and below that area to allow other signals to be returned.

Well... We didn't find E.T. on our initial analysis... We will continue to keep you informed as our analysis continues!

http://setiathome.ssl.berkeley.edu/index.html
Copyright ©2003 SETI@home

__________________
Are We Alone?

向版主反映这个帖子 | IP: 已记录