How do I handle inaccessible DOIs when citing papers?

I'm from Crossref. The DOI should always resolve to some location, even if the content has moved location on the web or has changed publisher. Can you share the example so we can get it reported? It's possible it's not a Crossref DOI of course (there are several other DOI agencies) but I can find out from the DOI. The prefix does often denote a particular publisher but journals change hands frequently as societies negotiate different publishing agreements so that's not necessarily a failsafe way to solve the problem (which I agree should not be your problem!). Happy to help if you share the example(s) :-)

I'm also from Crossref. Broken DOI links aren't good and we do try and fix them. If you get the "DOI Not Found" error page -for example when following this link - https://doi.org/10.15680/IJIRSET.2019.0806081 - it means that the DOI hasn't been registered. If you fill in the form then this error is reported to the appropriate Registration Agency. In Crossref's case we notify the publisher of the error and ask them to fix the problem - it usually gets fixed but if the publisher has ceased operations or isn't a Crossref member anymore the link may not be able to be fixed. Where archiving arrangements are in place the DOI can be redirected to archived copies of the content.

While not as volatile as URLs, DOIs still can be revoked or otherwise become unavailable.

In your case, consider if there is a chance that the DOI could become available again (there might be technical difficulties with the host) or whether there is another available source (another DOI). Personally, I would include the DOI in the bibliography and send a message to the author(s) or the provider (this should not be your problem to solve).

All accessible DOIs can be searched with Crossref (search by title, author, DOI and other metadata).