Connected k-domination no-kernel at twin-width 4
Open problem: no-kernel lower bound at twin-width 4 for Connected/Total k-Dominating Set · arXiv:2107.02882
Status open high confidence
The source paper (arXiv:2107.02882) establishes that Connected k-Dominating Set and Total k-Dominating Set have no polynomial kernel on graphs of bounded twin-width via local gadget modifications of their main reduction, but with a worse upper bound on the twin-width than 4. The open problem asks whether the twin-width bound can be brought down to exactly 4 (matching the result for plain k-Dominating Set in Theorem 1). No follow-up paper resolving this specific question was found in five targeted web searches covering the period 2021–2026.
Reviewer notes. No follow-up paper resolving or partially addressing this open problem was found. The MFCS 2025 paper on dominating set variants and twin-width (LIPIcs.MFCS.2025.13) addresses FPT algorithms rather than kernelization lower bounds and is unrelated. The conjecture is open with high confidence.
Context
The authors establish via local gadget modifications of their main reduction (Theorem 1) that Connected $k$-Dominating Set and Total $k$-Dominating Set have no polynomial kernel on graphs of bounded twin-width, but with a worse upper bound on the twin-width than 4. Bringing the twin-width bound down to 4—as achieved for plain $k$-Dominating Set in Theorem 1—would require additional work.
Notes. Implicit open problem stated in prose: 'More work would be necessary to get the lower bound for twin-width at most 4.' Confirmed by Table 1, which lists no 'no PK' entry for Connected k-Dominating Set at twin-width at most 4. PDF source — math notation may be partially garbled.
Source paper
Twin-width and polynomial kernels
Édouard Bonnet, Eun Jung Kim, Amadeus Reinald, Stéphan Thomassé, Rémi Watrigant · 2021-09-14
https://arxiv.org/abs/2107.02882
PDF source