Adaptive Anomaly Detection for Internet of Things in Hierarchical Edge Computing: A Contextual-Bandit Approach

The advances in deep neural networks (DNN) have significantly enhanced real-time detection of anomalous data in IoT applications. However, the complexity-accuracy-delay dilemma persists: Complex DNN models offer higher accuracy, but typical IoT devices can barely afford the computation load, and the remedy of offloading the load to the cloud incurs long delay. In this article, we address this challenge by proposing an adaptive anomaly detection scheme with hierarchical edge computing (HEC). Specifically, we first construct multiple anomaly detection DNN models with increasing complexity and associate each of them to a corresponding HEC layer. Then, we design an adaptive model selection scheme that is formulated as a contextual-bandit problem and solved by using a reinforcement learning policy network . We also incorporate a parallelism policy training method to accelerate the training process by taking advantage of distributed models. We build an HEC testbed using real IoT devices and implement and evaluate our contextual-bandit approach with both univariate and multivariate IoT datasets. In comparison with both baseline and state-of-the-art schemes, our adaptive approach strikes the best accuracy-delay tradeoff on the univariate dataset and achieves the best accuracy and F1-score on the multivariate dataset with only negligibly longer delay than the best (but inflexible) scheme.

Experimental analysis of influence of double-layer bump foils on aerodynamic thrust foil bearings performance

Purpose Gas thrust foil bearings (GTFBs) are used to balance the axial load of engines. However, in some working conditions of large axial force, such as the use of single impeller air compressor, the load capacity of GTFBs is still insufficient. To solve this problem, the load capacity can be improved by increasing the stiffness of bump foil. The purpose of this paper is to explore a scheme to effectively improve the performance of thrust foil bearings. In the paper, the stiffness of bump foil is improved by increasing the thickness of bump foil and using double-layer bump foil. Design/methodology/approach The foil deformation of GTFBs supported by three different types of bump foils, the relationship between friction power consumption and external force and the difference of limited load capacity were measured by experimental method. Findings The variation of the foil deformation, bearing stiffness, friction power consumption with the external force at different speeds and limited load capacity are obtained. Based on experimental results, the selection scheme of bump foil thickness is obtained. Originality/value This paper provides a feasible method for the performance optimization of GTFBs.

Computational Prediction of Heteromeric Protein Complex Disassembly Order with Hybrid Monte Carlo/Molecular Dynamics Simulation

The physicochemical entity of biological phenomenon in the cell is a network of biochemical reactions and the activity of such a network is regulated by multimeric protein complexes. Mass spectroscopy (MS) experiments and multimeric protein docking simulations based on structural bioinformatics techniques have revealed the molecular-level stoichiometry and static configuration of subcomplexes in their bound forms, then revealing the subcomplex populations and formation orders. Meanwhile, these methodologies are not designed to straightforwardly examine temporal dynamics of multimeric protein assembly and disassembly, essential physicochemical properties to understand functional expression mechanisms of proteins in the biological environment. To address the problem, we had developed an atomistic simulation in the framework of the hybrid Monte Carlo/Molecular Dynamics (hMC/MD) method and succeeded in observing disassembly of homomeric pentamer of the serum amyloid P component protein in experimentally consistent order. In this study, we improved the hMC/MD method to examine disassembly processes of the tryptophan synthase tetramer, a paradigmatic heteromeric protein complex in MS studies. We employed the likelihood-based selection scheme to determine a dissociation-prone subunit pair at each hMC/MD simulation cycle and achieved highly reliable predictions of the disassembly orders with the success rate over 0.9 without a priori knowledge of the MS experiments and structural bioinformatics simulations. We similarly succeeded in reliable predictions for the other three tetrameric protein complexes. These achievements indicate the potential availability of our hMC/MD approach as the general purpose methodology to obtain microscopic and physicochemical insights into multimeric protein complex formation.

Physiological Characters and Alimentary Needs of Both Divergent Lines of the Barbary Partridge (Alectoris barbara, Bonnaterre, 1792) in Algeria

Background: The Barbary partridge (Alectoris barbara) is a wild endemic species of North-Africa. The lack of information about this species, allows putting research focus in the hunting center of Zeralda, on the selection of a line for an improvement on the zootechnical characteristics of this population. Methods: Over 13 weeks of follow-up, we have performed weekly measurements on 3 successive generations F12-F14 with daily measurements of food intake. We are interested at the fundamental level of the biological characterization of both divergent lines. The analysis of the effects of selection pressures on the growth to estimate the metabolizable energy of each line. Result: During 273 days of study period, we noticed that the metabolizable energy has an average divergence between the fast line and the slow one with 0.328 Kcal/day since birth and an average divergence of 8.899 Kcal/day towards the 13th week in favour of the fast line. Significance noticed between males and females with a favour of the males which are more important. In addition to that, the values of weight-based consumption index imply higher indices for the slow line compared to the values of the fast line. Our results highlight the efficiency of the selection scheme. This progress will allow developing the restoration methods and/or the natural restocking populations on scientific bases.

A Coalition Formation Game for Cooperative Spectrum Sensing in Cognitive Radio Network under the Constraint of Overhead

Cooperative spectrum sensing improves the sensing performance of secondary users by exploiting spatial diversity in cognitive radio networks. However, the cooperation of secondary users introduces some overhead also that may degrade the overall performance of cooperative spectrum sensing. The trade-off between cooperation gain and overhead plays a vital role in modeling cooperative spectrum sensing. This paper considers overhead in terms of reporting energy and reporting time. We propose a cooperative spectrum sensing based coalitional game model where the utility of the game is formulated as a function of throughput gain and overhead. To achieve a rational average throughput of secondary users, the overhead incurred is to be optimized. This work emphasizes on optimization of the overhead incurred. In cooperative spectrum sensing, the large number of cooperating users improve the detection performance, on the contrary, it increases overhead too. So, to limit the maximum coalition size we propose a formulation under the constraint of the probability of false alarm. An efficient fusion center selection scheme and an algorithm to select eligible secondary users for reporting are proposed to reduce the reporting overhead. We also outline a distributed cooperative spectrum sensing algorithm using the properties of the coalition formation game and prove that the utility of the proposed game has non-transferable properties. The simulation results show that the proposed schemes reduce the overhead of reporting without compromising the overall detection performance of cooperative spectrum sensing.

User Quality of Experience (QoE) Satisfaction for Video Content Selection (VCS) Framework in Smartphone Devices

Video streaming is widely available nowadays. Moreover, since the pandemic hit all across the globe, many people stayed home and used streaming services for news, education, and entertainment. However, when streaming in session, user Quality of Experience (QoE) is unsatisfied with the video content selection while streaming on smartphone devices. Users are often irritated by unpredictable video quality format displays on their smartphone devices. In this paper, we proposed a framework video selection scheme that targets to increase QoE user satisfaction. We used a video content selection algorithm to map the video selection that satisfies the user the most regarding streaming quality. Video Content Selection (VCS) are classified into video attributes groups. The level of VCS streaming will gradually decrease to consider the least video selection that users will not accept depending on video quality. To evaluate the satisfaction level, we used the Mean Opinion Score (MOS) to measure the adaptability of user acceptance towards video streaming quality. The final results show that the proposed algorithm shows that the user satisfies the video selection, by altering the video attributes.

Analysis of Selection Methods Used in Genetic Algorithms

This paper offers a comprehensive review of selection methods used in the generational genetic algorithms.Firstly, a brief description of the following selection methods is presented: fitness proportionate selection methods including roulette-wheel selection (RWS) and its modifications, stochastic remainder selection with replacement (SRSWR), remainder stochastic independent selection (RSIS), and stochastic universal selection (SUS); ranking selection methods including linear and nonlinear rankings; tournament selection methods including deterministic and stochastic tournaments as well as tournaments with and without replacement; elitist and truncation selection methods; fitness uniform selection scheme (FUSS).Second, basic theoretical statements on selection method properties are given. Particularly, the selection noise, selection pressure, growth rate, reproduction rate, and computational complexity are considered. To illustrate selection method properties, numerous runs of genetic algorithms using the only selection method and no other genetic operator are conducted, and numerical characteristics of analyzed properties are computed. Specifically, to estimate the selection pressure, the takeover time and selection intensity are computed; to estimate the growth rate, the ratio of best individual copies in two consecutive populations is computed; to estimate the selection noise, the algorithm convergence speed is analyzed based on experiments carried out on a specific fitness function assigning the same fitness value to all individuals.Third, the effect of selection methods on the population fitness distribution is investigated. To do this, there are conducted genetic algorithm runs starting with a binomially distributed initial population. It is shown that most selection methods keep the distribution close to the original one providing an increased mean value of the distribution, while others (such as disruptive RWS, exponential ranking, truncation, and FUSS) change the distribution significantly. The obtained results are illustrated with the help of tables and histograms.