What are some of the best research papers you have read?

On the Use of Windows for Harmonic Analysis with the Discrete Fourier Transform [1]. Why? I think its a good paper overall. Clear, concise, and a fundamental read in the world of signal processing.

But I like it for another reason. It proves you can have an insanely cited paper (~10k citations) by summarizing the work of others in a clear concise manner. It proves there is academic value in not just churning out "novel" research, but by presenting existing facts in a new and useful light. It doesnt leave the reader working out the the (perhaps sometimes ego-driven) equation jumps. It holds your hand and walks you through the subject in a really refreshing way.

[1]http://www.ece.uprm.edu/~domingo/teaching/inel5309/On%20the%...

One of the explanations of a fundamental encoding algorithm for data encryption is one I enjoy reading. There were other papers, but this is a good explanation of arithmetic coding - https://www.cs.cmu.edu/~aarti/Class/10704/Intro_Arith_coding...

[1] is an great research paper about supporting temporal data in SQL. It goes beyond the SQL:2011 standard to support temporal outer joins and some other things. Basically they define a temporal variant for each relational operator then show how to implement them all based on just two join-like additions to SQL. They even implemented a proof-of-concept in Postgres. I wrote some more about their paper in my temporal databases bibliography [2] (ctrl-f "Alignment"). The paper is easy to read & understand, and it solves problems people have been working on since the 90s, with the most elegant solution I've seen.

[1] https://files.ifi.uzh.ch/boehlen/Papers/modf174-dignoes.pdf

[2] https://illuminatedcomputing.com/posts/2017/12/temporal-data...

[1] Bitcoin: A Peer-to-Peer Electronic Cash System

Why I consider the paper to be the best? The imply, the authorship, the flawlessness.

[1] bitcoin.org/bitcoin.pdf