Search

In this paper, we investigate the problem of bandwidth allocation in wireless sensor network (WSN) under signal to noise plus interference ratio interference model, which aims at finding a tradeoff between fairness and network throughput. Specifically, we propose an approximate algorithm to solve this problem since it has been proved to be NP-hard. Different from similar algorithms proposed in previous issues, we maximize two utility functions, which are the newly defined bandwidth utility function for fairness and network throughput, by jointly considering sensor nodes association and resource allocation in WSN. In addition, we formulate a new utility function with respect to bandwidth allocation, with the method of weighted sum of two objectives as one objective function, which will find a tradeoff between fairness and throughput. Consequently, the problem is decomposed into two sub-problems and solved in two stages, which are cluster formation stage and scheduling stage. In the first stage, we let sensor nodes join to cluster head nodes, which can determine the association of sensor nodes; in the second stage, the total utility function is maximized by allocating time slots for tradeoff between fairness and throughput. Finally, simulation results demonstrate that our algorithm can achieve better performance than compared algorithms.

Orthogonal frequency division multiplexing (OFDM) is a promising technique for dynamic spectrum access due to flexibility in spectrum shaping. However, out-of-band (OOB) power radiation of OFDM introduces significant interference to adjacent users. This is a serious problem in cognitive radio networks, which affects the secondary system to access the instantaneous spectrum hole. Existing methods either do not effectively reduce OOB power leakage or introduce significant bit-error-rate deterioration at the receiver. In this paper, a joint spectral precoding scheme is developed for OOB power reduction by using the matrix operations of orthogonal projection and singular value decomposition. We also propose an algorithm that designs the precoding matrix under receive performance constraint, which is then converted to matrix condition number constraint in practice. This method achieves the desirable spectrum envelope and receive performance by selecting zero-forcing frequencies. Simulation results show that OOB power decreases significantly by the proposed scheme under condition number constraint.

Along with the developments of deep learning, many recent architectures have been proposed for face recognition and even get close to human performance. However, accurately recognizing an identity from seriously noisy face images still remains a challenge. In this paper, we propose a carefully designed deep neural network coined noise-resistant network (NR-Network) for face recognition under noise. We present a multi-input structure in the final fully connected layer of the proposed NR-Network to extract a multi-scale and more discriminative feature from the input image. Experimental results such as the receiver-operating characteristic (ROC) curves on the AR database injected with different noise types show that the NR-Network is visibly superior to some state-of-the-art feature extraction algorithms and also achieves better performance than two deep benchmark networks for face recognition under noise.

Proper migration of neuronal somas and axonal growth cones to designated locations in the developing brain is essential for the assembly of functional neuronal circuits. Rapid progress in research of axon guidance and neuronal migration has been made in the last twenty years. Chinese researchers began their exploration in this field ten years ago and have made significant contributions in clarifying the signal transduction of axon guidance and neuronal migration. Several unique experimental approaches, including the migration assay of single isolated neurons in response to locally delivered guidance cues, have been developed by Chinese neuroscientists to investigate the molecular machinery underlying these guidance events.

This paper focuses on wide-sense stationary signal processing within a compressive sensing framework, proposing a new method of compressive sampling fast Fourier transform (FFT) accumulation method (CS-FAM). Depending on how it is applied, CS-FAM has one or two steps, allowing for versatility in multiband signal detection and parameter extraction. In the first step, the active sub-bands are detected using multiple measurement vectors (MMVs) and multiuser detection is achieved using a bandwidth constraint. In applications where it is required, such as in estimations of carrier frequency, symbol rate, or modulation format identification, the second step can be used to reconstruct the cyclic spectrums of each user individually. Based on the results of first step, parameter extraction is performed by searching for peaks in the cyclic spectrum rather than by the usual method of setting a threshold. Compared to other cyclic feature detection methods based on sub-Nyquist sampling, CS-FAM is low in complexity, allowing for practical implementation. Based on the results of the first step, parameter extraction from the cyclic spectrum is performed by searching for peaks rather than by setting a threshold. Although CS-FAM can only be employed for multiband signal detection, compared to other cyclic feature detection methods based on sub-Nyquist sampling, it is low in complexity, which makes practical implementation possible. Numerical simulations are presented to demonstrate the robustness of CS-FAM's multiband signal detection and the effectiveness of its cyclic spectrum estimation against both sampling rate reduction and noise uncertainty.