Abstract-As the base station is usually placed above local clutter, the angular spectrum incident on the base is narrow, inducing correlation among base antenna signals, which reduces the capacity of a multiple transmit and receive antenna systems. In this work the general expression for link capacity is derived, when there is correlation among receive antennas and among transmit antennas. It is found that an antenna separation of 4 wavelengths between nearest neighbors in a linear base array of dually polarized antennas allows one to achieve 80% of the capacity attainable in the uncorrelated antenna case.
To defend against internal attacks in wireless sensor networks (WSNs), building a trust model between sensors nodes has been proved to be an effective way in this paper. The most current trust models only consider communication behavior when calculating direct trust, which is directly calculated based on the interactions between sensor nodes. However, this is not enough because of the various types of attacks. Furthermore, the adverse effect of poor-quality links on the trust value of normal nodes is not discussed in the current trust models. In this paper, we propose a beta and link quality indicator (LQI)-based trust model (BLTM) for the WSNs. First, communication trust, energy trust, and data trust are considered when calculating direct trust. Then, the weight of communication trust, energy trust, and data trust are discussed. Finally, an LQI analysis mechanism is proposed to maintain the accuracy and stability of the trust value of normal nodes in a network with poor-quality links. Compared with other models, e.g., beta-based trust and reputation evaluation system (BTRES), the simulation results show that the BLTM can defend against internal attacks, e.g., DoS attack and data tampering attack which the BTRES cannot resist and can reduce the adverse effect of poor-quality links on the trust value of normal nodes effectively. INDEX TERMS Wireless sensor networks, beta distribution, link quality indicator, trust model.
The intrusion detection system deals with huge amount of data containing redundant and noisy features and the poor classifier algorithm causing the degradation of detection accuracy, in this paper, we introduce the random forest feature selection algorithm and propose a method that multi-classifier ensemble based on deep learning for intrusion detection. It used the random forest feature selection algorithm to extract optimal feature subset that are used to train by support vector machine, decision tree, naïve bayes and k-nearest neighbor classification algorithm, then, applying the deep learning to stack the output of four classifiers. The experimental results show that the proposed method can effectively improve the accuracy of intrusion detection compared with the majoring voting algorithm.
Rapid change of the channels in high-speed mobile communications will lead to difficulties in channel estimation and tracking, but can also provide Doppler diversity. In this paper, the performance of multiple-input multiple-output system with pilot-assisted repetitive coding and spatial multiplexing has been studied. With minimum mean square error (MMSE) channel estimation, an equivalent channel model and the corresponding system model are presented. Based on the random matrix theory, the asymptotic expressions of the normalized achievable sum rate of linear receivers, such as maximal ratio combining (MRC) receiver, MMSE detection and MRC-like receiver are derived. In addition, according to the symbol error rate of the MRC-like receiver, the maximum normalized Doppler diversity order, the minimum coding gain loss can be achieved when the repetition number and the signal to noise ratio tend to infinity and corresponding conditions are also derived. Based on the theoretical results, the impacts of different system configurations and channel parameters on system performance have been demonstrated.Index Terms-Doppler diversity, MIMO, deterministic equivalent, high mobility wireless communication system.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.