ABSTRACT
This chapter deals with another type of optimization problem called geometric
program (GP). The GP has been found useful in some resource allocation prob-
lems in communication and networking, where all the unknowns are positive in
nature, such as powers, achievable transmission rates, secrecy rates, etc. The
GP itself is a nonconvex problem that, however, can be converted into a convex
problem via the change of variables and the function transformation introduced
in Chapter 4, thereby providing an exemplar of convex problem reformulation
from a nonconvex problem (as detailed in Chapter 4). Then we introduce the
condensation method that is effectively implemented using the GP via some con-
servative approximations to the constraints of a specific type problem, followed
by its application in physical layer secret communications.