Fault-tolerant dynamic planning for wireless mesh networks based on real load profiles

An optimization algorithm for dynamic and fault-tolerant mesh network planning is proposed. The algorithm is based on instant user data demand prediction and optimizes the network by reducing the energy cost and guaranteeing a dynamic bandwidth management. The resulting mesh infrastructure becomes dynamic as nodes join and leave the system according to load profiles. Parameters, such as instantaneous load and SINR, are integrated into the system. Thus, a piece-wise linear approximation is proposed to handle non-linearity issues in order to resolve the complex optimization problem.