.
In this paper, we propose a new routing strategy on the basis of the so-called next-nearest-neighbor search strategy by introducing a preferential delivering exponent α. It is found that by tuning the parameter α, the scale-free network capacity measured by the order parameter is considerably enhanced compared to the normal next-nearest-neighbor strategy. Traffic dynamics both near and far away from the critical generating rate Rc are discussed, and it is found that the behavior of 1/f-like noise of the load time series not only depends on the generating rate R but also on the parameter α. We also investigate Rc as functions of C (capacity of nodes), m (connectivity density) and N (network size). Due to the low cost of acquiring next-nearest-neighbor information and the strongly improved network capacity, our strategy may be useful for the protocol designing of modern communication networks.