You are here

Conference vs. Journal Papers

Updated: December 17, 2015

Background: There has been a lot of confusion in our community over the issue of conference papers appearing as journal papers (or more rarely, vice versa).  The goal of this policy and webpage is to reduce or ideally eliminate this confusion by articulating a clear policy.

The following is the policy at the IEEE Transactions on Wireless Communications, which consistent with the IEEE policy (see IEEE Publications and Services Products Board Operations Manual, section 8.1.7.E) and and the ComSoc Policy, which is to encourage authors to submit papers that are extensions of their conference papers, ComSoc does not have a quantitative threshold for determining if the submitted manuscript has enough material beyond that of the conference paper(s). Rather, they leave that judgment to the editorial board.  The goal of this document is to give specific guidance to the TWC editorial board to help ensure that such “judgments” are consistent.

  • TWC Editors and Reviewers are instructed to consider conference and journal papers as totally separate venues for publication. 
    • Special Case 1: a full-length conference paper, for example 10 or more pages as in Infocom or some ACM conferences, cannot simply be republished in TWC verbatim without “substantial” differences.  Indeed, those communities generally treat the full-length conference paper as the terminal publication for their best ideas, unlike most of the wireless communications research community, where journals serve as the terminal publication.
    • Special Case 2: sometimes two different conference papers are “combined” into a journal paper.  This is acceptable, as long as the resulting journal paper is coherent and of high quality.
  • The “old” policy was that there should be a “30% difference” between the journal and conference version(s).  Note that this is not IEEE policy, nor is it ComSoc policy.  If it ever was, it no longer is.  The IEEE policy simply requires “substantial” differences, with the meaning of “substantial” being at the discretion of the Editor. TWC Editors are instructed to follow the more explicit policy here.
  • Text and figures can in fact be exactly copied from the conference paper(s) into the journal paper.  This is not double publishing.  An at least slightly different title should be used for the journal paper, to describe the typically broader journal contribution and to avoid confusion in the archival record.
  • The conference version(s) should be cited when submitting the journal paper in two places: (i) in the Manuscript Central field requesting this information, (ii) in the journal paper itself.  There does not need to be a summary explaining the differences/similarities, since they are assumed and allowed to be quite similar, simply the existence of such a paper must be acknowledged.
  • It is also possible to submit the journal and conference version at the same time.  Or the journal version can even be submitted first. The journal version should not have appeared, however, at the time of submission of a conference version of that paper.
  • It is the responsibility of the Authors to ensure that any copyright commitments they have made outside the IEEE, however, are not violated by the eventual IEEE journal publication.

Overall rationale for this policy:  Authors should be encouraged to present their work to the community at IEEE conferences, which inevitably involves publishing a conference version.  A policy that prevented conference papers from being later published in a journal version would have the opposite effect and discourage researchers from presenting their best work at conferences.  Similarly, it would damage the quality of IEEE journals if innovative ideas are barred from publication in them, simply because they have been presented in a conference format.