Writing papers requires a lot of trial and error. Generally, a paper's chances to be accepted for a given venue depend on three major points:

• The paper addresses an important problem and has a good point. This is obviously the most important thing. Some papers propose a solution to an important problem, while others simply investigate a problem offering better understanding of it. The second kind of a paper is usually easier to write than the first one, since it takes less time to analyze already gathered data, than to design and implement a solution to a known problem.
• The paper is well written. This is extremely important. You may have the greatest idea, but if it is not well presented the paper will be rejected. How to learn to write good papers? Read a lot of papers, discuss them in your research group, ask different people to read your writeups, and take a technical writing class. Generally organizing your presentation well, and using simple vocabulary and a lot of examples and illustrations works much better than usign a complicated jargon. People like nice, simple, tangible explanations, not boring, complex rambling.

  I found outlining before writing critical to writing clearly and well. "Outlining" here means writing a few sentences in a free style for each section you plan to write, then expanding these into a few paragraphs, and reorganizing. You can then expand each paragraph into nicely written text. Bear in mind that it will likely take many rewrites to get the text write and plan for them.

  There are many good books on how to write technical papers well. Read them!

  Also very important is to do your homework and write a good "Related work" section. Be sure to mention all the related work (people hate if their work is left out) and to present an objective and well documented critique of the related work (people hate if you criticize their work without any supporting arguments).

• The chosen venue welcomes the kind and the topic of the paper. Sometimes an excellent papers will be rejected if you send them to ill-suited venue. Look at the papers published in this venue in the previous years to get the feeling whether your paper is the right match.

If your paper is rejected, read the reviewers' comments and learn from them. Then fix the paper and send it to another venue. Having a paper rejected is definitely unpleasant but it happens quite often. Don't get discouraged. In time you will learn how to write to minimize rejections. Sometimes it also happens that you wrote a perfectly good paper but the reviewer's didn't spend enought time reading it and they misunderstood your claims. These things happen and, although they fill you with righteous anger, there is nothing much to be done about it. Fix the paper and send it to another venue.

Send your papers to good enough conferences. I know it is tempting to send to mediocre venues in interesting places (Hawaii, Carribean, Australia) so you get to visit those places and you also greatly increase acceptance chances. However, once you graduate and start looking for a job, those publications will not look very impressive.

Short point summary

• Select good topic
• Write well. Learn from other authors, welcome comments, outline, rewrite, rewrite, rewrite.
• Select a suitable and well known venue
• Accept that some papers will be rejected, and learn from the experience