Green and robust optimal design of Single Frequency Networks by min-max regret and ACO-based learning.

Notwithstanding the introduction of brand new 5G-based wireless services, single frequency networks supporting digital television and radio broadcasting still represent a major source of telecommunications services in modern smart cities. In this work, we propose a robust optimization model for the green design of second generation single frequency networks based on the digital television DVB-T standard, whose ongoing adoption requires to reconfigure and redesign existing networks. Our robust model aims at protecting design solutions against the data uncertainty that naturally affect propagation of signals in a real environment. For reducing conservatism of solutions, we refer to a heuristic min-max regret paradigm and to solve the resulting problem we propose to adopt a hybrid exact-heuristic algorithm based on the combination of an Ant Colony Optimization-like learning procedure, exploiting tight formulations of the optimization model, with an exact large neighborhood search. Results of computational tests considering realistic instances show that the heuristic min-max regret approach can produce solutions characterized by a substantially lower price of robustness without sacrificing protection against data uncertainty.