An Edge-Turning Strategy Based on Greedy Algorithm to Optimize the Network Controllability

Structural optimization is a significant problem in the research of complex network controllability. Conventional approaches based on global optimization may often not distinguish the role of local network in structural optimization. This paper suggests an edge-turning strategy based on a greedy algorithm (ESGA) to optimize the network structure, which achieves global optimization by searching and turning the candidate edges for local networks. Moreover, a theoretical boundary is provided based on an upper limit formula for the optimized total controllability index (OTCI), which validates the effectiveness and optimization impact of the strategy. Note that the optimized results from the numerical fitting of model networks closely approximate the results obtained by the formula when the number of optimized cactuses is within a certain range. Furthermore, Actual computations with various networks show that without significant increments in computation cost, the ESGA achieves higher optimization degrees in sparse networks and uniform networks.