Abstract
Different architectures for achieving ultra-high aggregate data rate PON are compared and optimized. An ultra-dense scheme is used to set the limits for the technology and a strategy for minimizing inter-channel nonlinear crosstalk is proposed.