How to Write an IEEE Transactions on Signal Processing Cover Letter

The order matters. TSP editors triage for signal-processing signal density, then confirm the IEEE paperwork is clean. A letter that names the principle, justifies the EDICS, and closes the declarations in that sequence is faster to route to the right Senior Area Editor.

Every IEEE submission needs an explicit originality statement. IEEE SPS policy is plain: submissions cannot be under consideration elsewhere nor be previously published. State both required declarations in one sentence:

This manuscript is original, has not been published previously, and is not under consideration for publication elsewhere, and all authors have read and approved the submission.

That single sentence carries both required declarations: not under review elsewhere, and all authors approved. Keep it in the letter even if the ScholarOne form also asks the same question. Editors read the absence of either line as a process gap, and process gaps invite a closer look at everything else.

EDICS classifications are not a minor metadata step. They tell the editorial system which technical community evaluates the paper, and a broad or fashionable EDICS choice can route a sharp contribution to reviewers who read it as off-center. The practical rule: choose EDICS that describe the core signal-processing method, not the application data.

If the real contribution is a distributed estimator, do not let the EDICS make it look like a biomedical or radar application just because those supplied the test signals. Justify the choice in one sentence in the cover letter so the editor can confirm the routing matches the claim.

TSP publishes a small set of submission types, and the editor screens each against a different page budget. Name the type implicitly by matching length and depth.

Source: IEEE TSP information for authors, IEEE Signal Processing Society (accessed June 2026).

Submissions go through the SPS ScholarOne portal at mc.manuscriptcentral.com/sps-ieee. Final manuscripts longer than 10 printed pages incur mandatory overlength charges of $220 per page above the first ten, on top of the voluntary $110-per-page charge for the first ten pages of a Regular Paper, and a submission with no general signal-processing contribution is a candidate for immediate rejection.

A Regular Paper that earns its 13 pages with a clean theorem, algorithm, and honest experiments beats one that stretches a thin result across the full budget. For a focused single result, IEEE Signal Processing Letters with its 1-to-2-month first decision is often the more honest target.

Three declarations are non-optional at TSP, and getting them wrong stalls the desk screen.

Conference-version disclosure. If any part of the work appeared at a conference such as ICASSP, cite it explicitly and state the delta. IEEE SPS policy allows a conference paper of up to six double-column pages to be the basis for a fuller journal version, but the journal submission must add novel aspects and the prior work must be cited.

Self-plagiarism that overlaps more than 25 percent with another journal manuscript, or uncredited reuse above 35 percent of the submitted paper, is subject to sanctions. Name the concrete additions: expanded theory, new proofs, additional experiments, or a new analysis. "Extended journal version" with no stated delta is a flag, not a disclosure. A preprint posted to arXiv should also be linked and disclosed so the editor knows the public-record status.

Suggested and opposed reviewers. Suggest 3 reviewers with EDICS-matched expertise and no conflict of interest, which IEEE defines to include co-authors, recent collaborators, and institutional colleagues. You may also choose to exclude reviewers you are opposed to, with a stated reason, and the editor generally grants reasonable requests. Put both lists in the letter; a conflicted suggestion makes the editor discount the whole list.

Competing interests. State competing interests plainly, or declare that there are none. For a resubmission of a previously rejected TSP paper, reference the prior manuscript ID and summarize what changed.

A TSP cover letter and EDICS-fit check scores these against your manuscript before you commit to submission.

Speaking from the editor's side of the desk: when we read a TSP-bound cover letter during triage, we are not asking whether the math is sophisticated. We assume it is. We are running one question on the first paragraph: would this contribution survive if you swapped out the application data?

If the answer is no, because the result only holds for the authors' wireless or biomedical example, the routing decision is usually made before we open the EDICS, because the paper is a better fit for T-COM, TASLP, or TIP. The second pass is mechanical: does the EDICS describe the method, is the conference disclosure honest, are the suggested reviewers free of obvious conflicts? A clean letter on both passes earns a fast routing decision.

The letters that earn a full read are the ones where the signal-processing principle is obviously the discovery, not the wrapper around an application win.

In our pre-submission review work with IEEE Transactions on Signal Processing manuscripts, four cover-letter patterns predict a scope redirect or a slow desk screen more reliably than anything in the manuscript body. Across the N=100 Manusights pre-submission reviews and TSP-style manuscript reads behind this guide, these patterns recurred more often than any single technical flaw. Each is testable against your own letter before you upload, and almost all of them are fixable in an afternoon.

The application paper with a thin theoretical contribution. This is the single most common failure we see in TSP cover letters. The letter walks through a dataset, a model, and an accuracy number, but never states the signal-processing principle the work establishes. The Senior Area Editor reads that as an application paper and redirects it to TIP, TASLP, or T-COM.

We apply a blunt test to the letter: cross out every sentence that names the specific application or dataset. If the remaining claim is no longer a general signal-processing result, the theory is too thin for TSP. The fix is to rewrite the opener so the first sentence names the estimator, detector, or guarantee and why it generalizes, then introduce the application as the test, not the contribution.

A missing or vague conference-extension disclosure. A large share of TSP submissions extend a prior ICASSP or workshop paper, yet the cover letter either omits the disclosure or writes "extended version" with no delta. Because IEEE SPS sanctions self-plagiarism above the 25 percent journal-overlap threshold and requires the journal version to add novel aspects, a vague disclosure invites exactly the scrutiny authors are hoping to avoid.

The high-leverage fix is to cite the conference paper and name the new theory, proofs, or experiments in one sentence, then state the overlap is below 25 percent. This is a manuscript-component claim the editor can verify, so it earns trust.

The wrong EDICS, or scope drift to TIP or T-COM. Across TSP manuscripts coming through pre-submission review, the letters that get mis-routed are the ones whose EDICS describes the application surface rather than the method. An image-reconstruction contribution tagged with an imaging EDICS reads as a TIP paper; a channel-estimation result tagged with a communications EDICS reads as a T-COM paper.

The fix is to choose the EDICS that names the core signal-processing method and to defend that choice in one cover-letter sentence, so the editor routes the paper to reviewers who treat the contribution as on-center.

Experiments that hide the failure modes. Many otherwise strong TSP letters promise reproducibility but describe experiments that only show success cases. TSP's editorial culture rewards synthetic studies that expose where the method breaks and real-data examples that test it outside the authors' ideal model assumptions. We flag letters that claim broad applicability without naming the regime where the estimator or detector degrades.

The fix is to state, in the cover letter, that the experiments include the failure regime and that the proof assumptions and sample-complexity conditions are stated explicitly in the manuscript, which signals a screen-ready package.

These four are all testable against your own draft. Before you upload, read your cover letter and ask whether an editor outside your exact subarea could tell, from the first paragraph alone, that this is a general signal-processing contribution and not an application paper. If not, the theory framing is the highest-leverage thing to fix, and a TSP cover letter framing check evaluates it against the manuscript before you commit.

If you cannot write the contribution sentence without leaning on the dataset name, that is useful information.

It may mean the work is genuinely an application paper, in which case a sister venue is the more honest target: IEEE Transactions on Image Processing for image and video work, IEEE Signal Processing Letters for a single focused result, IEEE Transactions on Communications for physical-layer results, IEEE Transactions on Information Theory for information-theoretic foundations, or the IEEE Open Journal of Signal Processing for open-access theory work.

For scope and mechanics before you write the letter, the IEEE Transactions on Signal Processing submission guide covers EDICS, page limits, and the fit line, and the IEEE Transactions on Signal Processing journal profile positions TSP against TIP, TASLP, TCI, and Signal Processing Letters.

How this page was created: this guide combines the IEEE Signal Processing Society information-for-authors page, the IEEE Author Center publishing-ethics and overlap policies, the TSP EDICS and reproducible-research pages, Clarivate JCR context, and Manusights pre-submission review patterns from signal-processing manuscripts. We did not access a private IEEE editorial account; the cover-letter guidance is built from public IEEE SPS materials and the editorial triage pattern we see across pre-submission reviews.

The named failure patterns above are drawn from our review data, not from any single submission, and no specific editor or reviewer is named because editorial rosters change and per-submission attribution is not something this page can verify.