In this paper, we study the fundamentals of the broadband powerline communication (BBPLC) for internet-of-things (IoT) applications, operating in the range of 1.2-300 MHz. First, we consider the BBPLC channel model by considering the joint statistics of the magnitude and phase of the channel frequency response. Then, we describe the widely used noise models in BBPLC, and study the detection performance of BPSK symbols over the BBPLC channel impaired with the asynchronous impulsive noise and the simplified Middleton's class A noise, over fading. Later, we consider the impact of the simplified Middleton's noise model on the BBPLC channel by evaluating the normalized capacity, by modeling the channel as an irreducible, stationary, aperiodic Markov chain with two states. Through numerical results, we compare the capacity obtained for this channel with that of the typical Middleton's class A channel, and show that the approximation due the simplified model is negligible when the ratio of th