Application of Internet of Things (IoT) in health, defense, banking, and other confidential information transfer urges the need for secure IoT. As most of the IoT devices are resource-limited (antennas, bandwidth, energy), securing the information transfer has always been a challenge. Looking at a solution for enhancing the security of single antenna, single carrier, energy efficient devices, we propose a novel scheme, channel-based mapping diversity. This scheme uses the inherent randomness of the wireless channel and multiple mappings available for an M -ary phase shift keying constellation in confusing an eavesdropper. When the legitimate and the eavesdropper channels are independent of each other, it is shown that a symbol error rate (SER) of \frac {(M-1)}{M} is induced at the eavesdropper. Whereas, when the channels are correlated, optimal and suboptimal strategies at source and eavesdropper are derived for their respective optimal performances. Further, a closed-form expression for a lower-bound on the SER at the eavesdropper is derived. Simulation results show that for the correlated case, as SNR at the eavesdropper increases, SER initially decreases, later saturates to a relatively high SER, hence making the job of the eavesdropper difficult in getting the legitimate data. Furthermore, the effect of the correlation is more pronounced on SER at higher levels of correlation. This indicates that for practical correlation scenarios, SER is high enough to confuse the eavesdropper. © 2014 IEEE.