Today I finished reviewing the last of 51 papers submitted to the 1998 MTT International Microwave Symposium (IMS) sessions on Nonlinear Modeling and Analysis. Like my nine colleagues on my subcommittee and the other couple hundred reviewers on other subcommittees, I'm exhausted. Anyone who thinks that this is fun is kidding himself.

A majority of these papers will be rejected. This should be no surprise; the competition is stiff. Of our 51 papers, we will accept about eight for regular sessions and perhaps as many more for the open forum. Our acceptance rate will be about 30%. The symposium's overall acceptance rate probably will be about 50%. With this kind of competition, papers with even modest flaws--to say nothing of major ones--don't stand a chance.

The frustration of seeing many potentially good papers rejected has inspired me to write this article. I should start with a disclaimer: the following are just my opinions and observations; they are not official MTT or IMS policy. Still, after several years on the TPC, I think I have a good idea of what is accepted and why.

So, here are my suggestions:

First, before submitting a paper, consider whether your paper really belongs in the IMS. The IMS deals with RF and microwave technology. Many subjects, which arguably fit within our range of interests, actually belong in other symposia. For example, we no longer have sessions on solid-state devices; such papers belong in the Electron Device Meeting. Although we obviously deal with aspects of circuit theory, papers on general circuit theory should be submitted to the International Symposium on Circuits and Systems. Similarly, most antenna papers are best submitted to the Antennas and Propagation Society Symposium.

Many types of papers that do fall within our sphere of interest still are not well suited to an IMS presentation; these should be sent to a print journal, perhaps the IEEE Transactions on Microwave Theory and Techniques. For example, a paper that includes a lot of industrial-strength mathematics probably should not go to the IMS. Unless the results can be summarized without a lot of Greek letters and other funny symbols, the paper will not do well in a presentation. Be honest: how many of these papers, consisting of slide after slide of heavy-duty mathematics, do you really understand at first sight? Shouldn't these really be in print, where there is plenty of time to study them?

We sometimes divert these turgid works to an open-forum session. This is a somewhat better venue for heavy subjects, but still not as good as a print journal. Unfortunately, many people view the open forum as a dumping ground for mediocre papers and can't imagine their shining creations in such lowly surroundings. If you have this problem, get help. Meanwhile, submit your gem to a print journal.
 
If your paper really is better suited to the open forum, submit it to the open forum. The open forum sessions are not for marginal papers; the standards of novelty and quality are the same as for the presentation sessions. The open forum is simply for a different type of paper: one that requires demonstrations, hardware exhibition, or extensive discussions.
 
Second, a point I repeat until I'm breathless, yet it still needs to be said: before submitting it, have your paper reviewed by someone in your organization who publishes regularly. This is especially important for beginning grad students. Go to your advisor (that's the guy who is always too busy to see you) and demand that he help you with the paper. That's his job, and, after all, he is a coauthor. Nothing is more irritating than to get a paper full of sophomoric drivel written by Joe Student, Jane Student, and Great God Professor, and it's obvious that G. G. P. never even looked at the paper. This is an insult to the reviewers: we supposedly have time to read and evaluate "his" paper, but he doesn't. What, do you suppose, are the chances that a reviewer will accept a paper that offends him in this way?
 
Here's a technique I perfected in grad school for getting a piece of a busy advisor's time: wait until he goes to the restroom, follow him in, and stand in front of the door so he can't escape until he talks to you. This presupposes, of course, that you are the same sex as your advisor. But if you're not, the trick might work even better.
 
Third, think long and hard about what you should put into your submission. An IMS submission is different from other kinds of publications. Especially, it is not an MTT Transactions paper. Unlike a Transactions paper, an IMS paper is accepted on the basis of a summary of the paper, not the paper itself. Unfortunately, most authors submit a complete paper, not a summary, forcing the reviewers to slog through a lot of unnecessary details and to exhume the paper's main idea from a mound of nonessentials. Believe me, this does not endear you to a reviewer who has a pile of 50 more papers in front of him!
 
On the other hand, a few authors take the other extreme. Recognizing that the reviewer has a lot of work to do, they submit a one-page summary that is far too brief to be evaluated. Both types of papers are likely to be rejected, even though the work may be very good.
 
So, what should a summary contain? First, it must have an abstract (more on that later.) Second, it must explain, as clearly, directly, and succinctly as possible, the nature of the problem, the methods used, and the results. These must be described in a way that shows the novelty of the results and how they advance the state of the art. Statements like "Excellent results were achieved" are marketing claims that tell the reviewer nothing. A long Transactions-style introduction with lofty statements about the importance of the technology and a description of all previous work (along with its shortcomings) is a waste of the reviewer's time. Similarly, lots of mathematical derivations are a waste of space: I can guarantee that no reviewer (except, perhaps, for a few with far too much spare time) will go through them in detail. Above all, don't add a lot of superfluous mathematics to show how smart you are and to make the paper a seem "high-level." This ploy will backfire, for sure. IMS reviewers have very sensitive "BS" detectors.

A good summary rarely requires more than two pages of text and a few figures. As a reviewer and editor, I have always maintained that any paper can be shortened to any length. Most unimaginative authors think that shortening a paper requires discarding basic information. Nonsense! To shorten a paper, discard layers of detail, not basic information. After all, what is an abstract, but a paper shortened to a paragraph or two?

Fourth, spend some time developing a good abstract. The abstract helps the reviewer a lot. Unfortunately, few authors write acceptable abstracts. I often feel that a paper's abstract sounds like conclusions, and its conclusions sound like an abstract. (In this case, merely reverse the two!)

How do you write a good abstract? Here's what I suggest: imagine that you're walking down the hallway and a friend stops you. "Hey," he says, "I hear that you submitted a paper to the IMS. What's it about?" Your answer is an abstract. You're likely to respond, "We looked into the problem of giving enemas to elephants as part of our management-consulting research. We tried various hoses and found that a six-foot-long, five-inch-diameter one worked best, because it was most similar to the convolutions in the elephant's digestive tract. This increased the productivity of the enema by 47%." 56 words, and it tells the whole story.

Fifth, be sure you know the rules for publications. An unethical paper has a 100% probability of rejection. One common lapse is to submit a paper before the results are in. Another is to submit two papers covering only slightly different territory in a transparent attempt to inflate publication lists. This sort of thing should be rare, but unfortunately, because of pressure on academics to publish large numbers of papers, it is becoming more common.

Sixth, learn how the papers are evaluated and evaluate your own in this light. IMS papers are evaluated according to four criteria:

1.  Originality: Is this significant, new research of technical or commercial significance?

2.  Qualitative Content: Is the paper an explicit summary with all necessary supporting data?

3.  Quality: Is the work correct and nontrivial? Does it have value that extends beyond previous work?

4.  Interest to MTT Membership: Is this something thing that IMS attendees will be interested in?

I well know that I've prescribed a lot of work. But it's essential. Musicians, who face even stiffer competition for acceptance, have a saying: Somewhere, right now, someone is practicing. Somewhere, right now, someone is preparing a first-rate symposium paper. Is it you?