For multiple unmanned-aerial-vehicles (UAVs)-assisted mobile-edge computing (MEC) networks, we study the problem of combined computation and communication for user equipments deployed with multitype tasks. Specifically, we consider that the MEC network encompasses both communication and computation uncertainties, where the partial channel state information and the inaccurate estimation of task complexity are only available. We introduce a robust design accounting for these uncertainties and minimize the total weighted energy consumption by jointly optimizing UAV trajectory, task partition, as well as the computation and communication resource allocation in the multi-UAV scenario. The formulated problem is challenging to solve with the coupled optimization variables and the high uncertainties. To overcome this issue, we reformulate a multiagent Markov decision process and propose a multiagent proximal policy optimization with Beta distribution framework to achieve a flexible learning policy. Numerical results demonstrate the effectiveness and robustness of the proposed algorithm for the multi-UAV-assisted MEC network, which outperforms the representative benchmarks of the deep reinforcement learning and heuristic algorithms.

By fully exploiting the cooperative communication capacities among mobile terminals (MTs), the MTs can adapt the offloading designs well to the practical network with dynamic features. In this paper, joint multi-user cooperative partial offloading, transmission scheduling and computation allocating is discussed for device-to-device (D2D) underlay mobile edge computing (MEC). By considering stochastic application requests, unpredictable MTs states, time-varying channel states and computation resources, a customized application offloading model, which aims to minimize the network-wide response latency and energy consumption simultaneously, is formulated. In order to solve this non-convex and non-smooth optimization problem, an online resource coordinating and allocating scheme (ORCAS) is proposed by exploiting Lyapunov optimization theory, variable substitution technique and resource provisioning priority mechanism. Both theoretical analyses and simulation results demonstrate that the proposed ORCAS can 1) drive the application response cost converge to the minimum; 2) achieve superior performance (e.g., the average network-wide response cost under ORCAS is approximately 19.14% lower than that under partial offloading directly); 3) adapt to dynamic situations in terms of stochastic user demands and channel states.

Time series analysis techniques, Lomb–Scargle periodogram and weighted wavelet Z-transform, have been employed to search for year-time-scale periodicities from the radio light curves of a typical blazar NRAO 530. The periodicity analysis using the two methods show consistent results: two strong and persistent periods of ∼10 and ∼6 yr are prominent at all three wavelengths; three other weaker periodicities with shorter time-scales (∼5, ∼3.5 and ∼3 yr) were also seen; in addition, the 14.5-GHz WWZ power spectrum revealed a ∼2-yr periodic component during the time interval 1983–1990. Statistical significance analysis suggests that these periodicities might result from physically periodic process. The characteristic frequencies of these periodicities seem to have a harmonic relationship, and the fundamental frequency, i.e. the first-order harmonic, is about 0.05 yr⁻¹ (P₀ = 20 yr). But the fundamental frequency was not pronounced. The multiplicity and apparent harmonic relationship of the periodicities could be interpreted by the global p-mode oscillation of the accretion disc, if the jet and disc are coupled. In this scenario, the periodic fluctuations in the accretion rate consequently lead to the observed periodic variabilities of the radio luminosity.

Machine learning is increasingly becoming a powerful tool to make decisions in a wide variety of applications, such as medical diagnosis and autonomous driving. Privacy concerns related to the training data and unfair behaviors of some decisions with regard to certain attributes (e.g., sex, race) are becoming more critical. Thus, constructing a fair machine learning model while simultaneously providing privacy protection becomes a challenging problem. In this paper, we focus on the design of classification model with fairness and differential privacy guarantees by jointly combining functional mechanism and decision boundary fairness. In order to enforce ϵ-differential privacy and fairness, we leverage the functional mechanism to add different amounts of Laplace noise regarding different attributes to the polynomial coefficients of the objective function in consideration of fairness constraint. We further propose an utility-enhancement scheme, called relaxed functional mechanism by adding Gaussian noise instead of Laplace noise, hence achieving (ϵ, δ)-differential privacy. Based on the relaxed functional mechanism, we can design (ϵ, δ)-differentially private and fair classification model. Moreover, our theoretical analysis and empirical results demonstrate that our two approaches achieve both fairness and differential privacy while preserving good utility and outperform the state-of-the-art algorithms.

3D object classification is an important task in computer vision. In order to explore the high-order and multi-modal correlations among 3D data, we propose an adaptive multi-hypergraph convolutional networks (AMHCN) framework to enhance 3D object classification performance. The proposed network improves the current hypergraph neural networks in two aspects. Firstly, existing networks rely on hyperedge constrained neighborhoods for feature aggregation, which may introduce noise or ignore positive information outside the hyperedges. To this end, we develop the partially absorbing random walks (PARW) to hypergraph for capturing optimal vertex neighborhoods from hypergraph globally. Then, based on the PARW on hypergraph, we design a new hypergraph convolution operator to learn deep embeddings from the optimized high-order correlation, which enables effective information propagation among the most relevant vertices. Secondly, concerning the multi-modal representations in practice, the current multi-modal hypergraph learning models either treat all modalities equally or introduce abundant parameters to learn weights of different modalities. To overcome these shortcomings, we propose a simple but effective dynamic weighting strategy for combining multi-modal representations, in which the importance of each modality can be adjusted adaptively by the loss function. We apply the proposed model to 3D object classification, and the experimental results on two 3D benchmark datasets demonstrate that our method outperforms the state-of-the-art methods, testifying to the effectiveness of both our convolution method and multi-modality fusion strategy.

Context. NGC 5408 X-1 is one of the few ultraluminous X-ray sources (ULXs) with an extensive monitoring program in X-rays (a temporal baseline of 4.2 yr), making it one of the most suitable candidates to study the long-timescale quasi-periodic oscillations (QPOs). Aims. Previous timing analysis of the Swift data of NGC 5408 X-1 led to detection of multiple periodicities ranging from 2.6 d to 230 d. In this paper, we focus on the statistical significance and the temporal evolution of these periodicities. Methods. A time-series analysis technique in the time-frequency domain, the weighted wavelet Z-transform (WWZ), was employed to identify the periodicities and trace their variations with time. Results. Three periodic components were detected from the WWZ periodogram, corresponding to periods of 2.65 ± 0.01 d, 115.4 ± 14.4 d and 189.1 ± 15.2 d. All three have statistical significance higher than 99.74% (≳3σ). The 2.65-d periodicity is quite stable in the majority of the light curve, whereas it changes to a shorter 2.50 d period in the last third of the time interval covered by the light curve. The 115-d periodicity is the most prominent but appears variable. It starts at an initial value of ~117.8 d and reaches 115.4 d around MJD 54 800; then it shifts into a lower-frequency branch (136 d period) between MJD 55 200−55 750; after that, this periodicity returns to 115 d until the end of the monitoring period. The 189-d periodicity is persistent across the whole time coverage. It shows a steadily decreasing trend from the beginning (193 d period) to the end (181 d period). Conclusions. The long-timescale periodicities in NGC 5408 X-1 are most likely of super-orbital origin, and are probably associated with the precession of a warped accretion disc. The disc may have been broken into two distinct planes with different precessing periods, i.e. the 189-d and 115-d periodicities corresponding to the outer and inner disc, respectively.