Unfortunately, there's no completely reliable way to determine whether a paper is plagiarized, but I think you're right in this case.
The 2012 IJWMN paper and 2014 IJLTET paper are pretty clearly plagiarized. Of course I can't prove anything, and one of them could in principle be the original, but it would take quite a story to explain how the same text was stolen and published by someone else three to five years earlier. A one-year delay might be due to refereeing or a conference rejection, but three to five years is a very long delay for CS.
The 2013 AJER paper is a strange case. It does copy some text without attribution, both from the 2009 paper and from the Wikipedia article on software design, but it has worse problems than that: the paper seems to have virtually no content beyond mock ups of what the user interface might look like, with messages such as "Your Secret Key is: 1004". Even aside from the plagiarism, it's safe to ignore this article as having essentially no research content.
The 2009 ICNS paper looks to me like the original. There are several tests one can do to try to gauge this:
You can do web searches for sentence fragments, such as "vulnerable to a large spectrum of attacks" or "introducing IBC into P2P". If you do a reasonable job of guessing distinctive phrases, you'll find a short list of potential word-for-word plagiarized papers. (Annoyingly, search engines will sometimes miss matches due to not parsing PDF files well, so there's a random element to this.) In this case, searches show that the 2009 paper took some wording from a 2008 paper by other authors, such as the bit about vulnerability, but seems not to have copied the whole paper. The borrowed wording is objectionable, but it doesn't invalidate any originality in content.
You can investigate the prestige of the journal/conference. Plagiarists are less likely to get away with publishing in high-prestige venues, because their theft is more likely to be noticed. I don't think ICNS is particularly prestigious, though, so this criterion doesn't help. The papers from 2009 are archived by the IEEE, but the IEEE archives a lot of junk on behalf of other conference organizers, so the IEEE archiving shouldn't be viewed as indicating high quality. It's very different from journals or conferences run by the IEEE.
You can see whether there are a lot of citations. That wouldn't prove anything, but it would suggest that people citing the paper weren't aware of plagiarism allegations. In this case, Google Scholar finds only four citations, so we don't learn much.
So in summary, it looks likely that the 2009 paper is the original (although I've only looked briefly, so it would be worth doing a few more searches). That's a remarkable collection of plagiarism you've uncovered, and I can see how it would be unsettling, but I think this is an unusual case.
