Wireless networks traffic is doubling every year. Small cells are part of the 5G paradigm and are advocated as the solution to spatially fragment the network and offer high capacity in densely populated areas. This is also due to the availability of spectrum at high frequencies (mmWave and visible light, for instance) which inherently offers large bandwidth but forces the cells to become smaller due to the small coverage. Although this can solve the last meter connectivity problem, the back/front-haul technologies that have to bridge the radio network with the core network are challenged. In this context, power lines for communications (PLC) comes into play. Through its recent technical developments that pushed performance much higher than expected, PLC can have a role in front-hauling for small cell networks. This talk will discuss recent methodologies to assess PLC network performance, which includes stochastic network and channel modeling. It will also show what machine learning (being one of the major tools and trends in several application domains) can do in PLC network planning and resource management.