In this paper we examine the effect of relay (R) placement on the energy efficiency of a wireless sensor network (WSN) operating over a η-μ fading channel. We consider η-μ statistics for modeling the small scale fading affected wireless channel between the sensor nodes. The η-μ model is quite general in nature and it encompasses other popular fading models such as Rayleigh, Nakagami etc. In order to address the demand for high data rate WSN applications, spectrally efficient M-ary quadrature amplitude modulation (MQAM) is adopted. Transmission between a pair of sensor nodes is benefitted from a third node, placed in a collinear manner (i.e. on the same straight line, joining source and destination) between source (S) and destination (D), and acting as a relay (R). The relay operates in a selective decode and forward (DF) mode. It has been seen that it is more beneficial to place the R closer to S than D. We have also investigated the effective relay placement region and how different design parameters such as constellation size (M), path loss exponent (n) and fading parameters (η and μ) affect the placement region.