Resource management and optimization in multi-user DSL systems

Digital subscriber line (DSL) technology is currently the most widely deployed broadband internet access technology and will continue to play an important role during the next decade. However, one of the major sources that limits the performance of current DSL systems is crosstalk, which is a channel distortion that is caused by the electromagnetic coupling among the different copper wires (DSL connections). Multi-user resource management is a very promising approach to prevent or even remove the impact of crosstalk, and that can significantly increase the performance of DSL systems. In this thesis, multiple efficient algorithms are proposed for multi-user resource management that only require a very low computational complexity and that can be applied to large-scale DSL systems. The application of these algorithms allows to significantly increase the data rates of DSL systems. It is furthermore shown that the proposed algorithms can be used to significantly reduce the energy consumption of DSL broadband internet. The proposed energy-aware technology is referred to as Green DSL, and it fits within the current emerging trend of green information and communication technology (ICT). For instance, it is demonstrated that the application of Green DSL can reduce the transmit powers of DSL modems by 50% with only a minor impact on the data rate performance.