sigcomm-related-work

SIGCOMM Related Work

Use this to audit novelty and placement. SIGCOMM is the broad networking flagship, so a paper

is read against the strongest prior work in its exact subarea — and against results scattered

across sibling venues. Reopen the current CFP for concurrent-submission and anonymity rules

before advising.

Positioning checks

• Separate the mechanism you contribute from engineering polish: a new protocol, transport

signal, forwarding architecture, measurement-driven design, or programmable-data-plane

technique.

• Sweep the whole networking literature, not just SIGCOMM: the nearest prior work may sit in

NSDI, MobiCom, CoNEXT, IMC, SIGMETRICS, INFOCOM, or a journal like IEEE/ACM Transactions on

Networking. A bibliography that cites only SIGCOMM signals a narrow read.

• Prove each placement. Verify the venue and year on dblp and the ACM DL before you assert

it; do not rely on memory for where a famous result appeared.

• Distinguish main-track proceedings from CCR editorial notes — OpenFlow's famous CCR

note is not a SIGCOMM conference paper, and citing it as one is a checkable error.

• Handle concurrent and prior versions honestly, and keep every self-citation phrased so

double-blind review survives.

Venue-coverage table

| Literature lane | Typical venues | What SIGCOMM reviewers check |

|---|---|---|

| Networking design flagships | SIGCOMM, NSDI, CoNEXT | Whether the nearest deployed mechanism is compared or distinguished |

| Wireless / mobile | MobiCom, SIGMOBILE venues | Whether relevant PHY/mobility work is acknowledged when adjacent |

| Measurement | IMC | Whether the motivating measurement rests on the right prior characterizations |

| Performance analysis | SIGMETRICS | Whether known models/bounds are credited |

| Archival journals | IEEE/ACM ToN, JSAC | Whether classical results are not being rediscovered |

A missing state-of-the-art comparison from the right venue is a recognizable SIGCOMM reject

pattern that benchmark strength alone does not repair.

The misattribution guard

Before citing "X appeared at SIGCOMM YYYY":
  1. Find it on dblp.org/db/conf/sigcomm/  -> confirm the edition and page range
  2. Confirm ACM DL lists it in the SIGCOMM conference proceedings, not CCR editorial
  3. Rule out the sibling: Hedera is NSDI, RED is ToN, BBR is CACM
  4. If unresolved -> mark 待核实 and do not assert the venue

Positioning vignette

A paper proposes a data-center transport that schedules flows in the fabric. Its nearest

neighbors: an earlier SIGCOMM congestion-signal paper (DCTCP), a fabric-scheduling paper in the

same lineage (pFabric), and an NSDI flow-scheduling system that is often confused with it. The

novelty sentence should name all three contrasts — a different signal than the first, a

different scheduling granularity than the second, and a design rather than the deployed-system

framing of the NSDI neighbor — with each venue verified.

Concurrent-work judgment calls

• Independently concurrent work at another venue: cite neutrally, state the technical

difference, and avoid priority claims reviewers cannot verify under blinding.

• Your own prior workshop or preprint version: cite in a blind-safe third person and confirm

the current CFP's stance on overlap before relying on it.

• When a venue's archival status is unclear, declare the overlap in the submission rather than

gambling on a chair's reading.

Output format

[Eligibility] clear / needs declaration / risky
[Closest venues] <SIGCOMM / NSDI / MobiCom / CoNEXT / IMC / journal>
[Nearest 3 works] <work -> distinction, each venue verified>
[Misattribution risk] <CCR-vs-main-track / sibling-venue traps checked>
[Novelty sentence] <SIGCOMM-ready mechanism contrast>