Rudin defined the integral of a nonnegative measurable function f as the supremum of a set of integrals of simple functions less than f.
With the assumption of f being bounded, how could I define the integral of f in terms of infimum of a set of integrals of simple functions greater than f. ?
What I got so far: Let f<=M, then integral of f is <= integral of M. M is a constant function hence is measurable and simple. Should I show the integral of M is the infimum of set?
Or should I start working on the integral of M-f, which is nonnegative measurable. The integral of M-f hence can be expressed as the sup of a set of integrals of simple functions less than M-f.
This definition appears in a paper which skips the proof. I would like a solid argument of it.