On a nonconvolution Volterra resolvent (Q1069094)

The author studies the system of Volterra equations \[ x(t)=\int^{t}_{0}a(t,s)x(s)ds+f(t),\quad t\geq 0. \] The solution can be given in the form \(x(t)=f(t)+\int^{t}_{0}r(t,s)f(s)ds\) provided that the resolvent kernel r(t,s) exists as the solution of the equation \[ r(t,s)=\int^{t}_{s}a(t,v)r(v,s)dv+a(t,s). \] The author gives some sufficient conditions for the existence of the resolvent kernel and proceeds to study under what conditions on the kernel a(t,s) it is true that the operator \(f\to \int^{t}_{0}r(t,s)f(s)ds\) maps some weighted \(L^ 1\)- or \(L^{\infty}\)-space into itself. All of these results require that the kernel a(t,s) should be small in some precisely defined sense. Then he gives some examples to show that in spite of this restriction, these results may be quite useful.