This is an interesting problem, however I am not sure the answers above rigorously prove it, as an integral is defined to be the limit between the upper and lower Riemann sums, and here we are technically asked to evaluate an infinite sum not an integral. Maybe the best we can do without...