We study a hierarchical heterogeneous Rayleigh fading wireless sensor network (WSN) in which sensor nodes surveil a region of interest (RoI) and use access points (APs) as relays to transmit their sensed information to base stations (BSs). By considering both large-scale path-loss signal attenuation and small-scale signal variation due to Rayleigh fading, we formulate the node deployment problem as an optimization problem intended to minimize the network's wireless communication power consumption. Given ergodic capacity constraints on all wireless links, we study the necessary conditions for an optimal AP and BS deployment. These necessary conditions are then assembled in the form of an iterative algorithm to deploy nodes. Finally, we establish the efficacy and superiority of our proposed node deployment algorithm against similar methods in the literature.