Power line communication (PLC) is an emerging technology for the realization of smart grid and home automation. It utilizes existing power line infrastructure for data communication in addition to the transmission of power. The PLC channel behaves significantly different from the wireless channel; and it is characterized by signal attenuation as well as by additive noise and multiplicative noise effects. The additive noise consists of background noise and impulsive noise; while the multiplicative noise results in fading of the received signal power. This paper focuses on the impact of the PLC channel characteristics on the outage and BER performance of a PLC system over Rayleigh fading channel with frequency- distance dependent attenuation and colored Nakagami-m distributed additive noise. We derive the exact closed-form expressions for the distribution of the instantaneous signal-to-noise ratio (SNR) and show the detector based the maximum Likelihood (ML) criterion as well as a simple but effecient suboptimal detector. Monte Carlo simulation results are used to verify the derived analytical expressions.