Lowest-degree preference random walks on complex networks

Abstract We introduce lowest-degree preference random walks on complex networks for which the walker concurrently adopts random walk strategy and lowest-degree search strategy controlled by a tuning parameter. Mean first passage time is adopted to quantify its search efficiency. We find that on a majority of real networks, our navigation strategy can significantly reduce the search time in comparison with random walks. The corresponding relative differences are over 15%. Moreover, we show that an optimal tuning parameter for which the minimal search time is achieved presents a strong positive correlation with entropy of degree sequence. Furthermore, we find that this optimal tuning parameter can indicate how much the search time can be reduced. Our work opens a new path for designing an efficient search strategy when only local information is available.