\(\psi\)-Li-Yau inequality for the \(p\)-Laplacian on weighted graphs with the \(CD_p^\psi (m, 0)\) curvature (Q6181154)

Gaussian heat kernel estimates, eigenvalue estimates, and others are some significant applications of the Harnack inequality. For the past ten years, an extensive effort has been undertaken to establish the related Li-Yau inequality [\textit{P. Li} and \textit{S. T. Yau}, Acta Math. 156, 154--201 (1986; Zbl 0611.58045)] and its applications on graphs. The chain rule of the Laplace operator, however, does not work on graphs, which is the primary issue. In this study, the author investigates the Li-Yau gradient estimate for the broad \(p\)-Laplacian for \(p\geq 2\) on graphs. His goal is to develop the idea of \textit{F. Münch} [J. Math. Pures Appl. (9) 120, 130--164 (2018; Zbl 1400.05219)] to the more general \(CD^{\psi} _{p} (m, 0)\) curvature situation within the \(p\)-Laplacian framework. In comparison to the linear case of \(p = 2\), studying the \(p\)-Laplacian for \(p > 2\) appears to be considerably more challenging, and the methods used are typically different. The Li-Yau inequality for the \(p\)-Laplacian on weighted graphs is also investigated. Under the condition \( p\geq 2\) for the \(CD^{\psi} _{p} (m, 0)\) curvature the author derives a more general type of \(\psi\)-Li-Yau inequality for positive solutions to the \(p\)-Laplace heat equation on finite graphs or locally finite graphs with bounded weighted vertex degree. The author applies the maximum principle and the semigroup method, which are two different strategies, respectively. The unified version of Münch [loc. cit.], which is connected to the log-Li-Yau inequality on manifolds and the \(\sqrt{\cdot}\)-Li-Yau inequality on graphs, is hence enlarged to the broader \(p\)-Laplacian setting. As an application of the main conclusions, the relevant Harnack inequality for the \(p\)-Laplacian is demonstrated.