brycewang-stanford MobiCom Supplementary Material
MobiCom gives you a hard 12 double-column pages including figures and tables, with references and optional appendices (placed after the references) outside the cap. The skill is placement: what must a reviewer read to judge the contribution, and what can wait in an appendix without hollowing out the body.
The placement contract
| Content | Where it goes | Why |
|---|---|---|
| Mechanism, headline result, main figures | 12-page body | reviewers judge the contribution here |
| Channel-model derivation, proofs | appendix (after references) | needed for correctness, not for the argument's flow |
| Full protocol / state machine | appendix | body states the idea; appendix has the detail |
| Extended sweeps, extra conditions | appendix | supports, does not carry, the claim |
| Instrument/measurement setup detail | appendix | provenance, not narrative |
| All citations | references (outside cap) | never trimmed for space or anonymity |
The controlling rule: the body must be self-contained enough to judge novelty and
correctness without the appendix. An appendix strengthens; it must not be load-bearing.
Reviewers are not obligated to read appendices, so the strongest result cannot live only
there (mobicom-writing-style).
What must stay in the body
- The mechanism, stated with its operating regime.
- At least one headline over-the-air result with its condition and uncertainty.
- The baseline comparison that supports the main claim.
- Enough method that a reviewer can see why the mechanism works.
If moving something to the appendix would make the body's argument incomplete, it stays — find the page elsewhere by triaging a figure or tightening prose, not by shrinking geometry.
HotCRP fields are part of the submission
The form is not an afterthought; reviewers meet it first:
HotCRP submission fields to treat as content:
title/abstract -> verbatim-match the PDF (drift reads as sloppiness)
topics -> chosen for the reviewer sub-community you want (routing)
authors/conflicts -> full list at abstract time for conflict computation
optional fields -> artifact/ethics declarations per the current CFP (待核实 wording)
Anonymity crosses the body/appendix line
An appendix is still part of the double-blind submission. A channel-model derivation that
names your testbed, a protocol listing with your repository URL, or a measurement appendix
with a site identifier de-anonymizes exactly as a body leak would. Run the blindness sweep
over the appendices too (mobicom-submission).
Rebuttal and revision packets are separate
Do not confuse post-review material with submission supplementary material:
- The rebuttal is a bounded response entered after stage-2 reviews; it is not extra
paper pages and has its own length limit (
mobicom-author-response). - A one-shot revision returns with a change-highlighted PDF and a change summary as
auxiliary material — again outside the 12-page count and governed by the revision
instructions, not the original CFP page rules (
mobicom-review-process).
Neither is a place to smuggle content that should have been in the body; both are governed by their own limits.
Audit checklist
- Body self-contained: novelty and correctness judgeable without the appendix.
- Strongest result in the body, not only supplementary.
- Derivations/protocols/extended results in appendices after references.
- References complete and outside the cap; none trimmed for anonymity.
- Appendices passed through the blindness sweep.
- HotCRP title/abstract/topics/conflicts audited as content.
- Rebuttal/revision material kept off the page count and to its own limit.
Output format
[Body completeness] can the contribution be judged from the 12 pages alone? y/n
[Placement] content mis-shelved (body<->appendix) with the move
[References] complete + outside cap? y/n
[Anonymity] appendix leaks found
[HotCRP] fields needing attention