Road Centrality and Landscape Spatial Patterns in Wuhan Metropolitan Area,China

Road network is a corridor system that interacts with surrounding landscapes,and understanding their interaction helps to develop an optimal plan for sustainable transportation and land use.This study investigates the relationships between road centrality and landscape patterns in the Wuhan Metropolitan Area,China.The densities of centrality measures,including closeness,betweenness,and straightness,are calculated by kernel density estimation(KDE).The landscape patterns are characterized by four landscape metrics,including percentage of landscape(PLAND),Shannon′s diversity index(SHDI),mean patch size(MPS),and mean shape index(MSI).Spearman rank correlation analysis is then used to quantify their relationships at both landscape and class levels.The results show that the centrality measures can reflect the hierarchy of road network as they associate with road grade.Further analysis exhibit that as centrality densities increase,the whole landscape becomes more fragmented and regular.At the class level,the forest gradually decreases and becomes fragmented,while the construction land increases and turns to more compact.Therefore,these findings indicate that the ability and potential applications of centrality densities estimated by KDE in quantifying the relationships between roads and landscapes,can provide detailed information and valuable guidance for transportation and land-use planning as well as a new insight into ecological effects of roads.

Non-orthogonal pilot pattern for sparse channel estimation in large-scale MIMO-OFDM system

From the perspective of compressed sensing（CS） theory, the channel estimation problem in large-scale multiple input multiple output（MIMO）-orthogonal frequency division multiplexing（OFDM） system is investigated. According to the theory, the smaller mutual coherence the reconstruction matrix has, the higher success probability the estimation can obtain. Aiming to design a pilot that can make the system reconstruction matrix having the smallest mutual coherence, this paper proposes a low complexity joint algorithm and obtains a kind of non-orthogonal pilot pattern. Simulation results show that compared with the conventional orthogonal pilot pattern, applying the proposed pattern in the CS channel estimation can obtain the better normalized mean square error performance. Moreover, the bit error rate performance of the large-scale MIMO-OFDM system is also improved.