The stability of d'Alembert's functional equation (Q1779970): Difference between revisions

Let \((G,+)\) be an Abelian group. \textit{J.A. Baker} [Proc. Am. Math. Soc. 80, 411--416 (1980; Zbl 0448.39003)] showed the superstability of the d'Alembert's functional equation \[ f(x+y)+f(x-y)=2f(x)f(y).\tag{1} \] Namely, he proved that if \(f:G\to\mathbb{C}\) satisfies (with \(\delta\geq 0\)) \[ \| f(x+y)+f(x-y)-2f(x)f(y)\| \leq\delta \tag{2} \] (with \text {\(\| \cdot\| =| \cdot| \)}) for all \(x,y\in G\), then \(f\) is bounded or satisfies (1). This result is now extended to the case where \(\mathbb{C}\) is replaced by a normed algebra \(X\) satisfying some additional assumptions (but not for an arbitrary one, as it is shown). In particular, the Baker's result holds for \(X=M_n(\mathbb{C})\) (the normed algebra of \(n\times n\) matrices with complex entries) provided that the subalgebra generated by \(f(G)\cup\{I\}\) coincides with \(M_n(\mathbb{C})\). The other result states that if \(f:G\to M_n(\mathbb{C})\) satisfies (2), then there exists a mapping \(h:G\to M_n(\mathbb{C})\) satisfying \[ \left(h(x+y)+h(x-y)-2h(x)h(y)\right)^2=0,\quad x,y\in G \] and such that \(f-h\) is bounded on \(G\).