No abstract

With the rapid development of the network, data fusion becomes an important research hotspot. Large amounts of data need to be preprocessed in data fusion; in practice, the features of datasets can be filtered to reduce the amount of data. The feature selection based on fuzzy rough sets can process a large number of continuous and discrete data to reduce the data dimension, making the selected feature subset highly correlated with the classification but less dependent on other features. In this paper, a new method of fuzzy rough feature selection is proposed which combines the membership function determination method of fuzzy c-means clustering and fuzzy equivalence to the original selection. Different from the existing research, our method takes full advantage of knowledge about the dataset itself and the differences between datasets, which makes the features selected have a higher correlation with the classification, improves the classification accuracy, and reduces the data dimension. Experimental results on the UCI machine learning repository datasets confirmed the performance and effectiveness of our method. Compared to the existing method, smaller subsets of features and an average of 1% higher classification accuracies were achieved.

Opportunistic networks take full advantage of opportunistic encounters among nodes to transfer packets. According to the characteristics of the limited energy of nodes and the frequent link variation in opportunistic networks, we introduce a novel routing metric that comprehensively takes into consideration the energy consumption of nodes, the probability of relay nodes meeting their destination, the time-to-live of the packet, and the approximate number of packet copies. Based on this metric, we propose a controllable multi-replica routing approach in which a single-branch diffusion strategy is applied to achieve the goal of dynamically controlling the number of replicas of packets. Our simulation results show that the proposed approach can avoid excessive load on individual nodes, guarantee the energy fairness among nodes, prolong the network lifetime, and effectively improve the delivery ratio of packets.Correspondingly, our design focuses on two challenges: (i) how to maintain the energy fairness among nodes in an opportunistic network to prolong the network lifetime while achieving a high delivery ratio, and (ii) how to dynamically control the number of copies of packets in an opportunistic network without affecting the packet delivery ratio. Our main contributions are summarized as follows:

In mobile opportunistic social networks (MOSNs), mobile users move around and conduct message transmission by using their portable devices without the support of communication infrastructures, in which routing is a challenge due to high connection intermittency and node mobility. The current routing schemes for MOSNs only focus on the improvement of cost‐effective delivery by using the characteristics of users' interests or mobility. In this paper, we propose a novel routing approach for MOSNs in which the probabilistic and social‐aware strategies are combined and the multiple‐copy policy is adopted. We first model an MOSN into some overlapping regions and then propose 2 region‐related centrality degrees related to users' social characteristics with the consideration of users' centrality, mobility, and interactions among users. According to the definition of centrality degrees, we design a social‐aware probabilistic routing approach in which a message replica control policy is applied to allow a user with higher encounter probability to the destination and the user could obtain more message replicas. The simulation results show that the proposed approach outperforms the previous routing algorithms in terms of delivery ratio and overhead ratio.