Canonical URL tag is a page level meta tag that helps site owners rid self-created duplicate content by specifying the version of a URL that is considered the most relevant and telling search engines which duplicate content to keep out.
How it Works
Let’s say our preferred URL for the BFM blog is: http://www.bluefountainmedia.com/blog/
But because of parameters or session ids we have 2 pages with duplicate content on different URLs: http://www.bluefountainmedia.com/blog/?ref=609
To fix this problem simply add a <link> tag with the preferred URL version inside the section of the duplicate content page:
<link rel=”canonical” href=”http://www.bluefountainmedia.com/blog/” />
Rand Fishkin Explains Canonical Tags