Until now, optical access networks have used basic intensity modulation direct detection transmission methods, using elementary signal recovery techniques (clock and data recovery). Now that baud rates are increasing, this simple approach is no longer viable if the system is to maintain the sizable loss budget (30 dB) that is required. At the same time, the cost of high-speed digital signal processing (DSP) has declined enough that it can be applied to an optical access system. The ITU-T has been developing the “High Speed Passive Optical Network” (HS-PON) since 2018, and that work has now culminated with the consent of the first complete set of recommendations that describe a 50 Gb/s (over a single wavelength) PON. In addition, several prototype systems have been developed and demonstrated in laboratory trials with several operators. It is expected that significant deployment of 50 Gb/s PON will begin in 2024.1 At present, there are many groups dedicated to DSP research, but most of these are targeted to wireless technologies or to coherent optical transmission. This means that the optical access field is presently underserved by the DSP community, and is fertile ground for research. Our presentation will provide the necessary technical background to understand the fiber access network, and then present the various ways in which DSP can be applied to solve the major issues that come from increasing the speed of those networks. One issue we will address is the use of bandwidth limited components to enable lower cost systems, where DSP can enable a doubling of baud rate with only 2dB worse optical sensitivity. We will also describe the problem of resolution in burst mode transmission, the classical challenge of passive optical networks. We show that DSP can do far better than the traditional analog techniques, both in terms of sophistication of signal recovery algorithms and in adaptability of reception parameters. Most engineers with optical access experience lack an in-depth knowledge of DSP technologies. Conversely, most DSP engineers lack the knowledge of the idiosyncrasies that arise from burst mode transmission at very high baud rate. The goal of this presentation is to bridge the gap between the PON and DSP engineers, planting a seen from which some new applications of existing techniques can be found.