Joint Optimization of Gain and Beaconing for First Order Consensus under Rayleigh Fading
Daniel M. Schneider, Hannes Frey · 2019
We consider distributed first order average consensus in wireless networked systems. Two optimization parameters (discrete gain and beaconing frequency) are considered in a mathematical framework. The framework is derived by extending the L2 joint spectral radius to a joint numerical radius and by deriving a closed form expression on beaconing success under Rayleigh fading. With our theoretical insights we derive a two-step heuristic which optimizes both parameters and requires significantly less information about the network compared to a global optimization. The heuristic is evaluated numerically and compared to the optimal solution. The simulated cases show that our heuristic can get close to the optimal solution.