On the relationship between exponential dichotomies and the Fredholm alternative (Q1107697)

A problem on the existence of bounded solutions of the equation \[ \dot x=A(t)x+f(t) \tag{1} \] where \(A(t)\) is \(n\times n\), continuous in \(R\) and \(f(t)\) is \(n\times 1\), is considered. One says that the Fredholm alternative holds if for each bounded and continuous function \(f(t)\), (1) has at least one bounded solution in \(R\) if and only if \(\int^{\infty}_{- \infty}y(t)f(t)\,dt=0\) for every bounded solution in \(R\), of the equation \(\dot y=-yA(t)\). It is known that if the equation \(\dot x=A(t)x\) has exponential dichotomies in \(R_+\) and \(R_-\) then the Fredholm alternative holds. The authors present conditions under which the converse of the above statement is true.