Single-image super-resolution (SISR) has experienced vigorous growth with the rapid development of deep learning. However, handling arbitrary scales ( e.g., integers, non-integers, or asymmetric) using a single model remains a challenging task. Existing super-resolution (SR) networks commonly employ static convolutions during feature extraction, which cannot effectively perceive changes in scales. Moreover, these continuous-scale upsampling modules only utilize the scale factors, without considering the diversity of local features. To activate more information for better reconstruction, two plug-in and compatible modules for fixed-scale networks are designed to perform arbitrary-scale SR tasks. Firstly, we design a Scale-aware Local Feature Adaptation Module (SLFAM), which adaptively adjusts the attention weights of dynamic filters based on the local features and scales. It enables the network to possess stronger representation capabilities. Then we propose a Local Feature Adaptation Upsampling Module (LFAUM), which combines scales and local features to perform arbitrary-scale reconstruction. It allows the upsampling to adapt to local structures. Besides, deformable convolution is utilized letting more information to be activated in the reconstruction, enabling the network to better adapt to the texture features. Extensive experiments on various benchmark datasets demonstrate that integrating the proposed modules into a fixed-scale SR network enables it to achieve satisfactory results with non-integer or asymmetric scales while maintaining advanced performance with integer scales.

The effects of the non-parametric multitaper method (MTM) and symmetrical higher-order space-time block-code (STBC) with full diversity techniques on the performance of spectrum sensing (SS) in cognitive radio (CR) systems are studied. The overall wireless link performance can be improved by using the STBC techniques to combat the multipath fading effects in wireless channels. The MTM is integrated with the STBC for spectrum estimation (SE), which will be termed multitaper spectrum estimation (MTSE)-STBC, and the given analysis of which is developed using the quadrature form approximation. Also, classical SE algorithms commonly focus on a fixed performance assessment method based on predefined false alarms or detection probabilities. Licensed users need to be protected against interference that opportunistic users might cause. An appropriate thresholding policy that varies concerning the designated values of false alarm and detection rates can achieve such protection. Utilization factors are instrumental in providing further protection for licensed users or occupied spectrum holes, and they need to be determined in advance and adopted in the detection policy. If utilization factors are not desirable, then a fine selection of appropriate threshold values will need to be decided as they have a major impact on the overall error probability performance. Using analytical and simulation methods, the assessment results revealed improved performance of the MTSE-STBC compared to other classical methods such as the Periodogram especially in the aspects of efficient transmission and less error probability. The proposed adaptive thresholding technique also proved useful in coping with different SE settings.

Whether conventional machine learning-based or current deep neural networks-based single image super-resolution (SISR) methods, they are generally trained and validated on synthetic datasets, in which low-resolution (LR) inputs are artificially produced by degrading high-resolution (HR) images based on a hand-crafted degradation model (e.g., bicubic downsampling). One of the main reasons for this is that it is challenging to build a realistic dataset composed of real-world LR–HR image pairs. However, a domain gap exists between synthetic and real-world data because the degradations in real scenarios are more complicated, limiting the performance in practical applications of SISR models trained with synthetic data. To address these problems, we propose a Self-supervised Cycle-consistent Learning-based Scale-Arbitrary Super-Resolution framework (SCL-SASR) for real-world images. Inspired by the Maximum a Posteriori estimation, our SCL-SASR consists of a Scale-Arbitrary Super-Resolution Network (SASRN) and an inverse Scale-Arbitrary Resolution-Degradation Network (SARDN). SARDN and SASRN restrain each other with the bidirectional cycle consistency constraints as well as image priors, making SASRN adapt to the image-specific degradation well. Meanwhile, considering the lack of targeted training images and the complexity of realistic degradations, SCL-SASR is designed to be online optimized solely with the LR input prior to the SR reconstruction. Benefitting from the flexible architecture and the self-supervised learning manner, SCL-SASR can easily super-resolve new images with arbitrary integer or non-integer scaling factors. Experiments on real-world images demonstrate the high flexibility and good applicability of SCL-SASR, which achieves better reconstruction performance than state-of-the-art self-supervised learning-based SISR methods as well as several external dataset-trained SISR models.

Multiple people can have similar appearances and portions of images can be occluded or have viewpoint changes in real scenarios, causing the increased difficulty of person re-identification (Re-ID). To address these problems, we propose a dual-channel person Re-ID algorithm that integrates person Re-ID image–text pairs into the classification network for end-to-end learning. We construct an image channel and a text channel, and subsequently extract visual information and text information using a convolutional neural network (CNN) and a simple recurrent units (SRUs) network, respectively. The text information is used to assist in the learning of visual information, consequently improving the robustness of the visual information. In addition, the visual features are divided into two branches to calculate the global and local features. Global features focus on the overall appearance of a person, whereas local features provide more fine-grained detail. Text information is more accurate and reliable, and it is thus more robust to occlusion and viewpoint changes. Visual information complemented by text information can describe a person more accurately and reliably. Extensive experiments demonstrate our method achieves state-of-the-art performance.

Even though Single Degradation Image Restoration (SDIR) has made significant progress and achieved remarkable performance, Multiple Degradation Image Restoration (MDIR) remains a long-term and arduous challenge to achieve the similar levels of success. To further improve the performance and efficiency of MDIR, we propose a novel MDIR method named RestorNet, which comprises an unsupervised degradation encoder for the learning of multi-scale degradation representations and a Multi-scale Degradation-assisted Restoration Module (MDRM) for image reconstruction. Our RestorNet aims to remove noise, rain, and haze in a unified network from the following three aspects. Firstly, to better distinguish among different degradations and learn the corruption information more accurately, we introduce a degradation-specific contrastive loss based on contrastive learning. Next, we develop a multi-scale degradation representation learning method to improve preservation of the spatial structure and distribution of inputs, and to extract multi-scale information to satisfy the diverse requirements of restoring different degraded images. Finally, to make a more reasonable use of degradation representation, we present a novel semi-guided strategy for effective feature transformation, where the multi-scale degradation representations are only incorporated into the MDRM encoder. For image denoising, deraining, and dehazing, by integrating the approaches above, RestorNet not only outperforms the recent state-of-the-art MDIR algorithms with lower computational complexity, but also achieves impressive performance in SDIR. Extensive experiments demonstrate the effectiveness and superiority of our proposed method.

Single image super-resolution (SISR), which aims to reconstruct a high-resolution (HR) image from a low-resolution (LR) observation, has been an active research topic in the area of image processing in recent decades. Particularly, deep learning-based super-resolution (SR) approaches have drawn much attention and have greatly improved the reconstruction performance on synthetic data. Recent studies show that simulation results on synthetic data usually overestimate the capacity to super-resolve real-world images. In this context, more and more researchers devote themselves to develop SR approaches for realistic images. This article aims to make a comprehensive review on real-world single image super-resolution (RSISR). More specifically, this review covers the critical publically available datasets and assessment metrics for RSISR, and four major categories of RSISR methods, namely the degradation modeling-based RSISR, image pairs-based RSISR, domain translation-based RSISR, and self-learning-based RSISR. Comparisons are also made among representative RSISR methods on benchmark datasets, in terms of both reconstruction quality and computational efficiency. Besides, we discuss challenges and promising research topics on RSISR.

With the increasing demand for maximizing efficiency in food production, technology can offer and perform farm operations for yield optimization. Modern agriculture can automate the entire process of cultivation from land preparation and preplanting to postharvest processes through data collection and processing. It implies that the analysis of big agricultural/farm data can be performed with the use of new information and communication technologies. This chapter applies precision agricultural technology with an unmanned aerial vehicle, embedded with optic and radiometric sensors, to obtain high spectral resolution images of a plantation’s status during a normal production/growth cycle. Then, the convolution neural network deep learning technique is employed to train images to develop an end-to-end multiclass classification system to determine the plant’s overall health status. Applying the pretrained model to the new images showed that the model was accurately able to predict any plant condition with an average of 99% accuracy.

Video coding effectively reduces the amount of video data while unavoidably producing compression noise. Compression noise can cause significant artifacts in compressed video, such as blocking, ringing, and blurring, which seriously affects the visual quality of videos and the value of videos for content analysis. In compressed video quality enhancement, few methods based on deep learning fully consider the relationship between video content and compression noise or the possibility of uniting the encoder or the decoder to enhance the quality of compressed video. In an approach different from existing methods, we propose a video quality enhancement framework based on the distribution characteristics of compression noise. The proposed framework consists of two parts: at the encoder, we propose a convolutional neural network (CNN)-based in-loop filtering network combined with noise distribution (IFN-ND) characteristics for the I frame instead of high efficiency video coding (HEVC) standard in-loop filters; at the decoder, we propose a CNN-based quality enhancement network combined with the noise distribution characteristics (PQEN-ND) for the P frames. The noise characteristics are extracted from the code stream to further improve the performance of the proposed networks. The experiments show that the proposed method can significantly improve the quality of HEVC compressed video, achieving an average 12.84% reduction in the BD rate and up to a 1.0476 dB increase in the peak signal-to-noise ratio (PSNR).

Fairness in bandwidth resource allocation is highly significance to the advancement of the future generation mobile and wireless technologies. It is likely that restriction of bandwidth due to the employment of some scheduling scheme would not be an appropriate option for the future development of communication systems. However, there is need to consider an implementation that would lead to good network performance and avoid unguaranteed bandwidth delivery. This paper focusses on evaluating the performance of Bandwidth Allocation using Dynamic Label Switching Paths (LSPs) Tunnelling and Label Distribution Protocol (LDP) signalling in Multi-Protocol Label Switching (MPLS) network. This will make provision for bandwidth allocation and reservation possible. An appropriate bandwidth allocation would have a positive impact on throughput as well as the delay. The results of an IP (Internet Protocol) Network without MPLS enabled is compared with MPLS model network. Furthermore, implementation of dynamic and static LSPs models are presented with about 75% decrease in packet delay variation for dynamic LSP when compared from static LSP. In addition, the models of bandwidth estimation, bandwidth allocation, delay and jitter are provided. Performance metrics used in this respect for multimedia services (Voice and Video conferencing) confirm that the modified models are improved in comparison with the baseline, having highest throughput of about 51% increment, and packet delay variation decreases drastically.

The current inefficient utilization of frequency spectrum has alerted regulatory bodies to streamline improvements. Cognitive radio (CR) has recently received considerable attention and is widely perceived as a promising improvement tool in estimating, or equivalently sensing, the frequency spectrum for wireless communication systems. The cognitive cycle in CR systems is capable of recognizing and processing better spectrum estimation (SE) and hence promotes the efficiency of spectrum utilization. Among different SE methods, the multi-taper method (MTM) shows encouraging results. Further performance improvement in the SE for CR can be achieved by applying multiple antennas and combining techniques. This paper proposes a constructive development of SE using MTM, abbreviated as MTSE, and by employing multiple-input multiple-output (MIMO), parsed into separate parallel channels using singular value decomposition (SVD), and maximum ratio combining (MRC) configurations. Deviating from these improvements, however, multicarrier systems such as orthogonal frequency division multiplexing (OFDM) show inferior sensing performances due to the noise multiplicity generated and combined from all subcarrier channels. By means of the quadrature matrix form, the probabilities for such integrated settings of SE have been derived to reach at their approximate asymptotes. Numerical simulations revealed specific better performances stemmed from coupling the fashionable MTSE and MIMO technologies.

This paper delivers an accurate approximation for adaptive threshold and optimal frame detection algorithms based on the robust multi-taper method aiming at an efficient spectrum sensing in cognitive radio systems. An appropriate adaptive thresholding allows for seamless vacation of unlicensed secondary users from certain bands upon primary users’ requests, while arbitrary optimal frame detection contributes to the computational and throughput demands. Simulation exercises corroborate the given analysis over Rayleigh channel and multiple-input multiple-output configuration and emphasize the critical role of adopting applicable adaptive threshold and optimal frame detection policies.

A rigorous model for automatic modulation classification (AMC) in cognitive radio systems is proposed in this paper. This is achieved by exploiting the Kalman filter (KF) integrated with an adaptive interacting multiple model (IMM) for resilient estimation of the channel state information. A novel approach is proposed, in adding up the square root singular values of the decomposed channel using the singular value decompositions algorithm. This new scheme, termed Frobenius eigenmode transmission, is chiefly intended to maintain the total power of all individual effective eigenmodes, as opposed to keeping only the dominant one. The analysis is applied over multiple-input multiple-output (MIMO) antennas in combination with a Rayleigh fading channel using a quasi-likelihood ratio test algorithm for AMC. The expectation-maximization is employed for recursive computation of the underlying estimation and classification algorithms. Novel simulations demonstrate the advantages of the combined IMM-KF structure when compared to the perfectly known channel and maximum likelihood estimate, in terms of achieving the targeted optimal performance with the desirable benefit of less computational complexity loads.

This Letter addresses the necessity for adaptive threshold (AT) and optimum frame duration (OFD) in spectrum sensing (SS). An adaptation paradigm of Kalman filter incorporated with interacting multiple model to interrogate the channel state information and adjusting the AT and OFD algorithms is proposed. The robust multitaper method is embraced for spectrum leakage control. The analysis of augmenting these algorithms is given and corroborated with simulations, which show their perks for efficient SS applications.

An effective approach for the design of spectrum estimation (SE) in cognitive radio systems using multi-taper method (MTM) and spatiotemporal features is presented in this study, whereas the MTM balances the bias-variance dilemma, the multiple-input–multiple-output (MIMO) and space–time block code (STBC) are customarily aimed to defeat the adverse channel effects and enhance the system capacity and performance. The singular value decomposition is exploited to determine the dominant eigenchannels in MIMO and STBC setups. The maximum-ratio combining, on the other hand, is adopted to produce higher signal-to-noise ratios usually intended for high data rates and reliability levels. The statistical analysis and modelling of the performance metrics associated with the SE based on MTM-STBC and MIMO are approached using the quadratic form approximation. Simulation exercises are employed to compare this different SE, which will be called multitaper spectrum estimation (MTSE), against other typical methods such as the periodogram without tapering options. The results exhibit performance gains due to the merger of MTSE and STBC technologies over MIMO and periodogram SE (PSE) algorithms. Further computational analysis shows that the MTSE–STBC has no extra burdens compared with its PSE counterpart.

The recent surge in the development of new technologies, most especially in the field of mobile and wireless communications, requires the adequate maintenance and overall procurement of network infrastructures. This is due to a great deal of accelerating demand from Mobile users having access to real-time information such as data, voice and video services. Therefore, the operators and service providers require seamless integration of network protocols with an improved quality of service (QoS). This paper addresses the performance of multimedia services in Multiprotocol Label Switching (MPLS) nodes and network models design using a simulation approach. MPLS ensures the reliability of the communication minimizing the delays and enhancing the speed of packet transfer. It is valuable in its capability of providing Traffic Engineering (TE) for minimizing the congestion by efficient throughput. The verification of the MPLS model will be the focus of the performance evaluation. An elaborate description of MPLS and its principle of operation will be required. It will eventually address the challenges of packet loss, high latency, high operational cost, more bandwidth utilization, and poor QoS.

Empirical modeling of wireless fading channels using common schemes such as autoregression and the finite state Markov chain (FSMC) is investigated. The conceptual background of both channel structures and the establishment of their mutual dependence in a confined manner are presented. The novel contribution lies in the proposal of a new approach for deriving the state transition probabilities borrowed from economic disciplines, which has not been studied so far with respect to the modeling of FSMC wireless fading channels. The proposed approach is based on equal portioning of the received signal-to-noise ratio, realized by using an alternative probability construction that was initially highlighted by Tauchen. The associated statistical procedure shows that a first-order FSMC with a limited number of channel states can satisfactorily approximate fading. The computational overheads of the proposed technique are analyzed and proven to beless demanding compared to the conventional FSMC approach based on the level crossing rate. Simulations confirm the analytical results and promising performance of the new channel model based on the Tauchen approach without extra complexity costs.

Agriculture is essential to the continued existence of human life as they directly depend on it for the production of food. The exponential rise in population calls for a rapid increase in food with the application of technology to reduce the laborious work and maximize production. Precision Agriculture is thought to be solution required to achieve the production rate required. There has been a significant improvement in the area of image processing and data processing which has being a major challenge previously in the practice of precision Agriculture. A database of images is collected through remote sensing, analyzed and a model is developed to determine the right treatment plans for different crop types and different regions. Features of images from vegetation's need to be extracted, classified, segmented and finally fed into the model. Different techniques have been applied to the processes from the use of the neural network, support vector machine, fuzzy logic approach and recently, the most effective approach generating fast and excellent results using the deep learning approach of convolution neural network for image classifications. Deep Convolution neural network is used in plant images recognition and classification to optimize production on a maize plantation. The experimental results on the developed model yielded results with an average accuracy of 99.58%.

This paper investigates the spectrum estimation (SE) in cognitive radio (CR) using the non-parametric multitaper method (MTM) and higher-order space-time block-code (STBC). Conventional SE methods usually suffer problems from biasing and leakage dilemma and the MTM is considered as an enhancement against such challenging situations. The STBC techniques help alleviate the multipath fading effects in wireless channels and hence improve the overall wireless link performance. The integration of MTM and STBC for SE will be developed using quadrature approximation and assessed using analytical and simulation methods. The simulation results revealed favorable MTM performance gains compared to other classical methods like Periodogram and also the STBC over MIMO for the same test settings.

Initial ranging is the primary and important process in wireless networks for the customer premise equipments (CPEs) to access the network and establish their connections with the base station. Contention may occur during the initial ranging process. To avoid contention, the mandatory solution defined in the standards is based on a truncated binary exponential random backoff (TBERB) algorithm with a fixed initial contention window size. However, the TBERB algorithm does not take into account the possibility that the number of contended CPEs may change dynamically over time, leading to a dynamically changing collision probability. To the best of our knowledge, this is the first attempt to address this issue. There are three major contributions presented in this paper. First, a comprehensive analysis of initial ranging mechanisms in wireless networks is provided and initial ranging request success probability is derived based on number of contending CPEs and the initial contention window size. Second, the average ranging success delay is derived for the maximum backoff stages. It is found that the collision probability is highly dependent on the size of the initial contention window and the number of contending CPEs. To achieve the higher success probability or to reduce the collision probability among CPEs, the OS needs to adjust the initial contention window size. To keep the collision probability at a specific value for the particular number of contending CPEs, it is necessary for the BS to schedule the required size of the initial contention window to facilitate the maximum number of CPEs to establish their connections with reasonable delay. In our third contribution, the initial window size is optimized to provide the least upper bound that meets the collision probability constraint for a particular number of contending CPEs. The numerical results validate our analysis. (C) 2016 Elsevier Inc. All rights reserved.

Frequency spectrum sensing is a key and challenging task of modern wireless communication systems. Cognitive radio (CR) has recently been envisioned as an intelligent platform capable of achieving better spectrum sensing and, hence, promotes efficient spectrum utilization. The multi-taper method (MTM) is a good candidate to achieve such a requirement. This paper focuses on the application of MTM spectrum sensing using multiple antenna arrangement environments. Maximal ratio combining (MRC) is employed as near to optimum candidate for diversity combining technique to process the spectrum estimate at the output of multiple receive antennas. Analysis and results show that such an arrangement provides better performance compared to the single antenna option under additive white Gaussian noise (AWGN) conditions. Moreover, a proposal for a new and simple approach to derive the decision statistical measures reveals that the probability density functions (PDFs) share common factors in both time and frequency domains. Furthermore, transmit antenna diversity also assumes direct positive influence on the spectrum sensing performance.