p-adic "Laplacian"

[hung]

hello, do anyone know of a good analogue of the Laplacian over the the p-adic field? if so, what is its spectrum, say on an 'interval", a + (p^n)Zp?
thank you,
thl