Activation of nomadic relay nodes in dynamic interference environment for energy saving

This paper presents an optimization framework for energy savings in nomadic relay networks, where interference and load are changing over time due to varying assignments. We prove the existence of an explicit load function taking assignments as arguments and we show the function is continuously differentiable. Based on the properties of the load function, we design iterative relay and user association algorithm for energy savings, where the non-convex load constraints are linearly approximated such that each sub-problem is a linear program and hence can be solved efficiently. Simulation results confirm that our proposed algorithmic approach leads to a significant reduction of the energy consumption when compared with algorithms considering the worst-case interference.