On an integral inequality in n-independent variables (Q1058616)

A new generalization of a nonlinear integral inequality in n independent variables of the Bihari type due to \textit{B. G. Pachpatte} [J. Math. Anal. Appl. 79, 256-272 (1981; Zbl 0455.26009), Theorem 3] is proved here by induction argument. An application to pointwise estimates of solutions of a certain class of nonlinear hyperbolic partial differential equations is also introduced here to convey the usefulness of this new extension. The considered integral inequality is of the form \[ \quad (1)\quad \phi (x)\leq f(x)+q(x)\sum^{m}_{i=1}(\int^{x}_{x^ 0}g_ i(y)H_ i(\phi (y))dy)+w(x,\int^{x}_{x^ 0}K(x,y,\phi (y))dy),\quad x\in \Omega, \] where \(\Omega\) is an open bounded set in \(R^ n\)- the Euclidean n-space; \(\phi,f,q,g_ i,H_ i,K,w\) are all nonnegative continuous functions, and \(\int^{x}_{x^ 0}...dt\) denotes the n-fold integral \(\int^{x_ 1}_{x^ 0_ 1}\int^{x_ 2}_{x^ 0_ 2}...\int^{x_ n}_{x^ o_ n}...dt_ 1 dt_ 2...dt_ n.\) Assuming suitable conditions on the functions \(f,q,g_ i,H_ i,K\) and w, which are too complicated to be stated here and are similar to those used in that paper of Pachpatte mentioned above, the author established an explicit upper bound for solutions of the inequality (1). The known result obtained by Pachpatte can be derived by letting \(m=1\) in (1).