On convolution of entire probability densities

Abstract
Let t be a non-negative function of C 1[0, ∞) such that tt(0) = 0 and t′( r) ↑ ∞ as r ↑ ∞ We prove that there exists an entire function f non-negative on R and satisfying the following conditions: 1)0 < limsupr→∞ M(r, f) exp (−t(r)) < ∞, where 2)fL 1 (R); 3)sup{f(x): x ∈ R} = ∞; 4)ess sup{(f * f)( x) : x ∈ (α, β)} = ∞ for all non-empty intervals (α, β) ⊂ R.

This publication has 0 references indexed in Scilit: