Distributed resource allocation with binary decisions via Newton-like Neural Network dynamics

This paper aims to solve a distributed resource allocation problem with binary local constraints. The is formulated as program cost function defined by the summation of agent costs plus global mismatch/penalty term. We propose modification Hopfield Neural Network (HNN) dynamics in order this while incorporating novel Newton-like weighting factor. addition lends itself fast avoidance saddle points, which gradient-like HNN susceptible to. Turning multi-agent setting, we reformulate and develop implementation dynamics. show that if solution reformulation obtained, it also centralized problem. A main contribution work probability converging point an appropriately energy both settings zero under light assumptions. Finally, enlarge our algorithm annealing technique gradually learns feasible solution. Simulation results demonstrate proposed methods are competitive greedy SDP relaxation approaches terms quality, advantage approach significant improvement runtime over method quality implementation.