Search results for: jet noise

Authors: Stefano Zaninotto, Geoffrey Farrugia, Johan Debattista, Jason Gauci

Abstract:

The need to reduce fuel and noise during taxi operations in the airports with a scenario of constantly increasing air traffic has resulted in an effort by the aerospace industry to move towards electric taxiing. In fact, this is one of the problems that is currently being addressed by SESAR JU and two main solutions are being proposed. With the first solution, electric motors are installed in the main (or nose) landing gear of the aircraft. With the second solution, manned or unmanned electric tow trucks are used to tow aircraft from the gate to the runway (or vice-versa). The presence of the tow trucks results in an increase in vehicle traffic inside the airport. Therefore, it is important to design the system in a way that the workload of Air Traffic Control (ATC) is not increased and the system assists ATC in managing all ground operations. The aim of this work is to develop an electric taxiing system, based on the use of autonomous tow trucks, which optimizes aircraft ground operations while keeping ATC in the loop. This system will consist of two components: an optimization tool and a Graphical User Interface (GUI). The optimization tool will be responsible for determining the optimal path for arriving and departing aircraft; allocating a tow truck to each taxiing aircraft; detecting conflicts between aircraft and/or tow trucks; and proposing solutions to resolve any conflicts. There are two main optimization strategies proposed in the literature. With centralized optimization, a central authority coordinates and makes the decision for all ground movements, in order to find a global optimum. With the second strategy, called decentralized optimization or multi-agent system, the decision authority is distributed among several agents. These agents could be the aircraft, the tow trucks, and taxiway or runway intersections. This approach finds local optima; however, it scales better with the number of ground movements and is more robust to external disturbances (such as taxi delays or unscheduled events). The strategy proposed in this work is a hybrid system combining aspects of these two approaches. The GUI will provide information on the movement and status of each aircraft and tow truck, and alert ATC about any impending conflicts. It will also enable ATC to give taxi clearances and to modify the routes proposed by the system. The complete system will be tested via computer simulation of various taxi scenarios at multiple airports, including Malta International Airport, a major international airport, and a fictitious airport. These tests will involve actual Air Traffic Controllers in order to evaluate the GUI and assess the impact of the system on ATC workload and situation awareness. It is expected that the proposed system will increase the efficiency of taxi operations while reducing their environmental impact. Furthermore, it is envisaged that the system will facilitate various controller tasks and improve ATC situation awareness.

Keywords: air traffic control, electric taxiing, autonomous tow trucks, graphical user interface, ground operations, multi-agent, route optimization

Procedia PDF Downloads 129

Authors: Getaneh Berie Tarekegn, Li-Chia Tai

Abstract:

With the rapid advancement of artificial intelligence, low-power built-in sensors on Internet of Things devices, and communication technologies, location-aware services have become increasingly popular and have permeated every aspect of people’s lives. Global navigation satellite systems (GNSSs) are the default method of providing continuous positioning services for ground and aerial vehicles, as well as consumer devices (smartphones, watches, notepads, etc.). However, the environment affects satellite positioning systems, particularly indoors, in dense urban and suburban cities enclosed by skyscrapers, or when deep shadows obscure satellite signals. This is because (1) indoor environments are more complicated due to the presence of many objects surrounding them; (2) reflection within the building is highly dependent on the surrounding environment, including the positions of objects and human activity; and (3) satellite signals cannot be reached in an indoor environment, and GNSS doesn't have enough power to penetrate building walls. GPS is also highly power-hungry, which poses a severe challenge for battery-powered IoT devices. Due to these challenges, IoT applications are limited. Consequently, precise, seamless, and ubiquitous Positioning, Navigation and Timing (PNT) systems are crucial for many artificial intelligence Internet of Things (AI-IoT) applications in the era of smart cities. Their applications include traffic monitoring, emergency alarming, environmental monitoring, location-based advertising, intelligent transportation, and smart health care. This paper proposes a generative AI-based positioning scheme for large-scale wireless settings using fingerprinting techniques. In this article, we presented a novel semi-supervised deep convolutional generative adversarial network (S-DCGAN)-based radio map construction method for real-time device localization. We also employed a reliable signal fingerprint feature extraction method with t-distributed stochastic neighbor embedding (t-SNE), which extracts dominant features while eliminating noise from hybrid WLAN and long-term evolution (LTE) fingerprints. The proposed scheme reduced the workload of site surveying required to build the fingerprint database by up to 78.5% and significantly improved positioning accuracy. The results show that the average positioning error of GAILoc is less than 0.39 m, and more than 90% of the errors are less than 0.82 m. According to numerical results, SRCLoc improves positioning performance and reduces radio map construction costs significantly compared to traditional methods.

Keywords: location-aware services, feature extraction technique, generative adversarial network, long short-term memory, support vector machine

Procedia PDF Downloads 47

Authors: Rachel M. Edelstein, John D. Van Horn, Karen M. Schmidt, Sydney N. Cushing

Abstract:

As a consequence of sports-related concussions, female athletes have been documented as reporting more symptoms than their male counterparts, in addition to incurring longer periods of recovery. However, the role of sex and its potential influence on symptom reporting and recovery outcomes in concussion management has not been completely explored. The present aims to investigate the relationship between female concussion symptom severity and the presence of assessment bias. The Sport Concussion Assessment Tool 3 (SCAT3), collected by the NCAA and DoD CARE Consortium, was quantified at five different time points post-concussion. N= 1,258 NCAA athletes, n= 473 female (soccer, rugby, lacrosse, ice hockey) and n=785 male athletes (football, rugby, lacrosse, ice hockey). A polytomous Item Response Theory (IRT) Graded Response Model (GRM) was used to assess the relationship between sex and symptom reporting. Differential Item Functioning (DIF) and Differential Group Functioning (DGF) were used to examine potential group-level bias. Interactions for DIF were utilized to explore the impact of sex on symptom reporting among NCAA male and female athletes throughout and after their concussion recovery. DIF was significantly detected after B-H corrections displayed in limited items; however, one symptom, “Pressure in Head” (-0.29, p=0.04 vs -0.20, p =0.04), was statistically significant at both < 6 hours and 24-48 hours. Thus, implies that at < 6 hours, males were 29% less likely to indicate “Pressure in the Head” compared to female athletes and 20% less likely at 24-48 hours. Overall, the DGF suggested significant group differences, suggesting that male athletes might be at a higher risk for returning to play prematurely (logits = -0.38, p < 0.001). However, after analyzing the SCAT 3, a clinically relevant trend was discovered. Twelve out of the twenty-two symptoms suggest higher difficulty in female athletes within three or more of the five-time points. These symptoms include Balance Problems, Blurry Vision, Confusion, Dizziness, Don’t Feel Right, Feel in Fog, Feel Slow Down, Low Energy, Neck Pain, Sensitivity to Light, Sensitivity to Noise, Trouble Falling Asleep. Despite a lack of statistical significance, this tendency is contrary to current literature stating that males may be unclear on symptoms, but females may be more honest in reporting symptoms. Further research, which includes possible modifying socioecological factors, is needed to determine whether females may consistently experience more symptoms and require longer recovery times or if, parsimoniously, males tend to present their symptoms and readiness for play differently than females. Such research will help to improve the validity of current assumptions concerning male as compared to female head injuries and optimize individualized treatments for sports-related head injuries.

Keywords: female athlete, sports-related concussion, item response theory, concussion assessment

Procedia PDF Downloads 77

Authors: Maya de Vries Kedem, Abir Jubran, Diana Baron

Abstract:

Employment in Israel offers Palestinian workers an income double what they can earn in the West Bank. The need to support their families leads many educated Palestinians to forgo finding work in their profession in the Palestinian Authority and instead look for employment in those sectors open to them in Israel, particularly the construction, agriculture, and industry sectors. The International Labor Organization estimated that about 1,200 workers in Israel die every year because of occupational diseases (diseases caused by working conditions). Construction workers in Israel are constantly exposed to dust, noise, chemical materials, and work in awkward postures, which require prolonged bending, repetitive motion, and other risk factors that can lead to illnesses and death. Occupational health is vastly neglected in Israel and construction workers are particularly at risk . As of June 2022, the Israeli quota in the construction sector for Palestinian workers stood at 80,000. Kav LaOved released a new study on the state of occupational health among Palestinian workers employed in construction in Israel. The study Roadmap to Health: Palestinian Workers in Israel's Construction Sector reviews the extent to which the health of Palestinian workers is protected at work in Israel. The report includes analysis of a survey administered to 256 workers as well as interviews with 10 workers and with 5 Israeli occupational health experts. Report highlights: • Among survey respondents, 63.9% stated that safety procedures to protect their health are rarely followed in their workplace (e.g., taking breaks, using protective gear, following restrictions on lifting heavy items, and having inspectors regularly on site to monitor safety). • All 256 Palestinian workers who participated to the survey said that their health has been directly or indirectly harmed by working in Israel and reported suffering from the following problems: orthopedic problems such as joint, hand, leg or knee problems (100%); headaches (75%); back problems (36.3%); eye problems (23.8%); breathing problems (17.6%); chronic pain (14.8%); heart problems (7.8%); and skin problems (3.5%). • Workers who are injured or do not feel well often continue working for fear of losing their payment for that day. About half of the 256 survey respondents reported that they pay brokerage fees to find an employer with a work permit, often paying between 2,000 and 3,000 NIS per month. “I have an obligation—I pay about NIS 120 a day for my permit, [and] I have to pay for it whether I work or not" a worker said. • Most Palestinian construction workers suffer from stress and mental health problems. Workers pointed to several issues that greatly affect their mood and mental state: daily crossings at crowded checkpoints where workers stand for hours; lack of sleep due to leaving home daily at 3:00-3:30 am; commuting two to four hours to work in each direction; and abusive work environments. A worker told KLO that the sight of thousands of workers standing together at the checkpoint causes “high blood pressure and the feeling that you are going to be squeezed.” Another said, “I felt that my bones would break.” In the survey workers reported suffering from insomnia (70.1%), breathing difficulties (35.8%), chest pressure (27.6%), or rapid pulse rate (12.2%).

Keywords: construction sector, palestinian workers, occupational health, Israel, occupation

Procedia PDF Downloads 88

Authors: Katherine Dropiewski, Michael Yakimov, Vadim Tokranov, Allan Minns, Pavel Murat, Serge Oktyabrsky

Abstract:

InAs Quantum Dots (QDs) in a GaAs matrix is a well-documented luminescent material with high light yield, as well as thermal and ionizing radiation tolerance due to quantum confinement. These benefits can be leveraged for high-efficiency, room temperature scintillation detectors. The proposed scintillator is composed of InAs QDs acting as luminescence centers in a GaAs stopping medium, which also acts as a waveguide. This system has appealing potential properties, including high light yield (~240,000 photons/MeV) and fast capture of photoelectrons (2-5ps), orders of magnitude better than currently used inorganic scintillators, such as LYSO or BaF2. The high refractive index of the GaAs matrix (n=3.4) ensures light emitted by the QDs is waveguided, which can be collected by an integrated photodiode (PD). Scintillation structures were grown using Molecular Beam Epitaxy (MBE) and consist of thick GaAs waveguiding layers with embedded sheets of modulation p-type doped InAs QDs. An AlAs sacrificial layer is grown between the waveguide and the GaAs substrate for epitaxial lift-off to separate the scintillator film and transfer it to a low-index substrate for waveguiding measurements. One consideration when using a low-density material like GaAs (~5.32 g/cm³) as a stopping medium is the matrix thickness in the dimension of radiation collection. Therefore, luminescence properties of very thick (4-20 microns) waveguides with up to 100 QD layers were studied. The optimization of the medium included QD shape, density, doping, and AlGaAs barriers at the waveguide surfaces to prevent non-radiative recombination. To characterize the efficiency of QD luminescence, low temperature photoluminescence (PL) (77-450 K) was measured and fitted using a kinetic model. The PL intensity degrades by only 40% at RT, with an activation energy for electron escape from QDs to the barrier of ~60 meV. Attenuation within the waveguide (WG) is a limiting factor for the lateral size of a scintillation detector, so PL spectroscopy in the waveguiding configuration was studied. Spectra were measured while the laser (630 nm) excitation point was scanned away from the collecting fiber coupled to the edge of the WG. The QD ground state PL peak at 1.04 eV (1190 nm) was inhomogeneously broadened with FWHM of 28 meV (33 nm) and showed a distinct red-shift due to self-absorption in the QDs. Attenuation stabilized after traveling over 1 mm through the WG, at about 3 cm⁻¹. Finally, a scintillator sample was used to test detection and evaluate timing characteristics using 5.5 MeV alpha particles. With a 2D waveguide and a small area of integrated PD, the collected charge averaged 8.4 x10⁴ electrons, corresponding to a collection efficiency of about 7%. The scintillation response had 80 ps noise-limited time resolution and a QD decay time of 0.6 ns. The data confirms unique properties of this scintillation detector which can be potentially much faster than any currently used inorganic scintillator.

Keywords: GaAs, InAs, molecular beam epitaxy, quantum dots, III-V semiconductor

Procedia PDF Downloads 256

Authors: Safaa M. Hassan; Safwat S. Gabr, Mohamed F. Sadek

Abstract:

This study is proposed for the lithological and iron oxides detection in the old mine areas of El-Bahariya Depression, Western Desert, using ASTER and Landsat-8 remote sensing data. Four old iron ore occurrences, namely; El-Gedida, El-Haraa, Ghurabi, and Nasir mine areas found in the El-Bahariya area. This study aims to find new high potential areas for iron mineralization around El-Baharyia depression. Image processing methods such as principle component analysis (PCA) and band ratios (b4/b5, b5/b6, b6/b7, and 4/2, 6/7, band 6) images were used for lithological identification/mapping that includes the iron content in the investigated area. ASTER and Landsat-8 visible and short-wave infrared data found to help mapping the ferruginous sandstones, iron oxides as well as the clay minerals in and around the old mines area of El-Bahariya depression. Landsat-8 band ratio and the principle component of this study showed well distribution of the lithological units, especially ferruginous sandstones and iron zones (hematite and limonite) along with detection of probable high potential areas for iron mineralization which can be used in the future and proved the ability of Landsat-8 and ASTER data in mapping these features. Minimum Noise Fraction (MNF), Mixture Tuned Matched Filtering (MTMF), pixel purity index methods as well as Spectral Ange Mapper classifier algorithm have been successfully discriminated the hematite and limonite content within the iron zones in the study area. Various ASTER image spectra and ASD field spectra of hematite and limonite and the surrounding rocks are compared and found to be consistent in terms of the presence of absorption features at range from 1.95 to 2.3 μm for hematite and limonite. Pixel purity index algorithm and two sub-pixel spectral methods, namely Mixture Tuned Matched Filtering (MTMF) and matched filtering (MF) methods, are applied to ASTER bands to delineate iron oxides (hematite and limonite) rich zones within the rock units. The results are validated in the field by comparing image spectra of spectrally anomalous zone with the USGS resampled laboratory spectra of hematite and limonite samples using ASD measurements. A number of iron oxides rich zones in addition to the main surface exposures of the El-Gadidah Mine, are confirmed in the field. The proposed method is a successful application of spectral mapping of iron oxides deposits in the exposed rock units (i.e., ferruginous sandstone) and present approach of both ASTER and ASD hyperspectral data processing can be used to delineate iron-rich zones occurring within similar geological provinces in any parts of the world.

Keywords: Landsat-8, ASTER, lithological mapping, iron exploration, western desert

Procedia PDF Downloads 145

Authors: Sardorbek Numonov, Hyohun Kim, Dongwha Shin, Yeonseok Kim, Ji-Su Ahn, Dongeun Choi, Byung-Woo Hong

Abstract:

The detection of curvilinear structures often plays an important role in the analysis of images. In particular, it is considered as a crucial step for the diagnosis of chronic respiratory diseases to localize the fissures in chest CT imagery where the lung is divided into five lobes by the fissures that are characterized by linear features in appearance. However, the characteristic linear features for the fissures are often shown to be subtle due to the high intensity variability, pathological deformation or image noise involved in the imaging procedure, which leads to the uncertainty in the quantification of anatomical or functional properties of the lung. Thus, it is desired to enhance the linear features present in the chest CT images so that the distinctiveness in the delineation of the lobe is improved. We propose a recursive diffusion process that prefers coherent features based on the analysis of structure tensor in an anisotropic manner. The local image features associated with certain scales and directions can be characterized by the eigenanalysis of the structure tensor that is often regularized via isotropic diffusion filters. However, the isotropic diffusion filters involved in the computation of the structure tensor generally blur geometrically significant structure of the features leading to the degradation of the characteristic power in the feature space. Thus, it is required to take into consideration of local structure of the feature in scale and direction when computing the structure tensor. We apply an anisotropic diffusion in consideration of scale and direction of the features in the computation of the structure tensor that subsequently provides the geometrical structure of the features by its eigenanalysis that determines the shape of the anisotropic diffusion kernel. The recursive application of the anisotropic diffusion with the kernel the shape of which is derived from the structure tensor leading to the anisotropic scale-space where the geometrical features are preserved via the eigenanalysis of the structure tensor computed from the diffused image. The recursive interaction between the anisotropic diffusion based on the geometry-driven kernels and the computation of the structure tensor that determines the shape of the diffusion kernels yields a scale-space where geometrical properties of the image structure are effectively characterized. We apply our recursive anisotropic diffusion algorithm to the detection of curvilinear structure in the chest CT imagery where the fissures present curvilinear features and define the boundary of lobes. It is shown that our algorithm yields precise detection of the fissures while overcoming the subtlety in defining the characteristic linear features. The quantitative evaluation demonstrates the robustness and effectiveness of the proposed algorithm for the detection of fissures in the chest CT in terms of the false positive and the true positive measures. The receiver operating characteristic curves indicate the potential of our algorithm as a segmentation tool in the clinical environment. This work was supported by the MISP(Ministry of Science and ICT), Korea, under the National Program for Excellence in SW (20170001000011001) supervised by the IITP(Institute for Information and Communications Technology Promotion).

Keywords: anisotropic diffusion, chest CT imagery, chronic respiratory disease, curvilinear structure, fissure detection, structure tensor

Procedia PDF Downloads 232

Authors: Farhan A. Alenizi

Abstract:

Digital watermarking has evolved in the past years as an important means for data authentication and ownership protection. The images and video watermarking was well known in the field of multimedia processing; however, 3D objects' watermarking techniques have emerged as an important means for the same purposes, as 3D mesh models are in increasing use in different areas of scientific, industrial, and medical applications. Like the image watermarking techniques, 3D watermarking can take place in either space or transform domains. Unlike images and video watermarking, where the frames have regular structures in both space and temporal domains, 3D objects are represented in different ways as meshes that are basically irregular samplings of surfaces; moreover, meshes can undergo a large variety of alterations which may be hard to tackle. This makes the watermarking process more challenging. While the transform domain watermarking is preferable in images and videos, they are still difficult to implement in 3d meshes due to the huge number of vertices involved and the complicated topology and geometry, and hence the difficulty to perform the spectral decomposition, even though significant work was done in the field. Spatial domain watermarking has attracted significant attention in the past years; they can either act on the topology or on the geometry of the model. Exploiting the statistical characteristics in the 3D mesh models from both geometrical and topological aspects was useful in hiding data. However, doing that with minimal surface distortions to the mesh attracted significant research in the field. A 3D mesh blind watermarking technique is proposed in this research. The watermarking method depends on modifying the vertices' positions with respect to the center of the object. An optimal method will be developed to reduce the errors, minimizing the distortions that the 3d object may experience due to the watermarking process, and reducing the computational complexity due to the iterations and other factors. The technique relies on the displacement process of the vertices' locations depending on the modification of the variances of the vertices’ norms. Statistical analyses were performed to establish the proper distributions that best fit each mesh, and hence establishing the bins sizes. Several optimizing approaches were introduced in the realms of mesh local roughness, the statistical distributions of the norms, and the displacements in the mesh centers. To evaluate the algorithm's robustness against other common geometry and connectivity attacks, the watermarked objects were subjected to uniform noise, Laplacian smoothing, vertices quantization, simplification, and cropping. Experimental results showed that the approach is robust in terms of both perceptual and quantitative qualities. It was also robust against both geometry and connectivity attacks. Moreover, the probability of true positive detection versus the probability of false-positive detection was evaluated. To validate the accuracy of the test cases, the receiver operating characteristics (ROC) curves were drawn, and they’ve shown robustness from this aspect. 3D watermarking is still a new field but still a promising one.

Keywords: watermarking, mesh objects, local roughness, Laplacian Smoothing

Procedia PDF Downloads 160

Authors: Pratap Chandra Pati, Prabina Rajib, Parama Barai

Abstract:

The traditional finance theory neglects the role of sentiment factor in asset pricing. However, the behavioral approach to asset-pricing based on noise trader model and limit to arbitrage includes investor sentiment as a priced risk factor in the assist pricing model. Investor sentiment affects stock more that are vulnerable to speculation, hard to value and risky to arbitrage. It includes small stocks, high volatility stocks, growth stocks, distressed stocks, young stocks and non-dividend-paying stocks. Since the introduction of Chicago Board Options Exchange (CBOE) volatility index (VIX) in 1993, it is used as a measure of future volatility in the stock market and also as a measure of investor sentiment. CBOE VIX index, in particular, is often referred to as the ‘investors’ fear gauge’ by public media and prior literature. The upward spikes in the volatility index are associated with bouts of market turmoil and uncertainty. High levels of the volatility index indicate fear, anxiety and pessimistic expectations of investors about the stock market. On the contrary, low levels of the volatility index reflect confident and optimistic attitude of investors. Based on the above discussions, we investigate whether market-wide fear levels measured volatility index is priced factor in the standard asset pricing model for the Indian stock market. First, we investigate the performance and validity of Fama and French three-factor model and Carhart four-factor model in the Indian stock market. Second, we explore whether India volatility index as a proxy for fearful market-based sentiment indicators affect the cross section of stock returns after controlling for well-established risk factors such as market excess return, size, book-to-market, and momentum. Asset pricing tests are performed using monthly data on CNX 500 index constituent stocks listed on the National stock exchange of India Limited (NSE) over the sample period that extends from January 2008 to March 2017. To examine whether India volatility index, as an indicator of fear sentiment, is a priced risk factor, changes in India VIX is included as an explanatory variable in the Fama-French three-factor model as well as Carhart four-factor model. For the empirical testing, we use three different sets of test portfolios used as the dependent variable in the in asset pricing regressions. The first portfolio set is the 4x4 sorts on the size and B/M ratio. The second portfolio set is the 4x4 sort on the size and sensitivity beta of change in IVIX. The third portfolio set is the 2x3x2 independent triple-sorting on size, B/M and sensitivity beta of change in IVIX. We find evidence that size, value and momentum factors continue to exist in Indian stock market. However, VIX index does not constitute a priced risk factor in the cross-section of returns. The inseparability of volatility and jump risk in the VIX is a possible explanation of the current findings in the study.

Keywords: India VIX, Fama-French model, Carhart four-factor model, asset pricing

Procedia PDF Downloads 252

Authors: Vikrant Gupta, Amrit Goswami

Abstract:

The fixed income market forms the basis of the modern financial market. All other assets in financial markets derive their value from the bond market. Owing to its over-the-counter nature, corporate bonds have relatively less data publicly available and thus is researched upon far less compared to Equities. Bond price prediction is a complex financial time series forecasting problem and is considered very crucial in the domain of finance. The bond prices are highly volatile and full of noise which makes it very difficult for traditional statistical time-series models to capture the complexity in series patterns which leads to inefficient forecasts. To overcome the inefficiencies of statistical models, various machine learning techniques were initially used in the literature for more accurate forecasting of time-series. However, simple machine learning methods such as linear regression, support vectors, random forests fail to provide efficient results when tested on highly complex sequences such as stock prices and bond prices. hence to capture these intricate sequence patterns, various deep learning-based methodologies have been discussed in the literature. In this study, a recurrent neural network-based deep learning model using long short term networks for prediction of corporate bond prices has been discussed. Long Short Term networks (LSTM) have been widely used in the literature for various sequence learning tasks in various domains such as machine translation, speech recognition, etc. In recent years, various studies have discussed the effectiveness of LSTMs in forecasting complex time-series sequences and have shown promising results when compared to other methodologies. LSTMs are a special kind of recurrent neural networks which are capable of learning long term dependencies due to its memory function which traditional neural networks fail to capture. In this study, a simple LSTM, Stacked LSTM and a Masked LSTM based model has been discussed with respect to varying input sequences (three days, seven days and 14 days). In order to facilitate faster learning and to gradually decompose the complexity of bond price sequence, an Empirical Mode Decomposition (EMD) has been used, which has resulted in accuracy improvement of the standalone LSTM model. With a variety of Technical Indicators and EMD decomposed time series, Masked LSTM outperformed the other two counterparts in terms of prediction accuracy. To benchmark the proposed model, the results have been compared with traditional time series models (ARIMA), shallow neural networks and above discussed three different LSTM models. In summary, our results show that the use of LSTM models provide more accurate results and should be explored more within the asset management industry.

Keywords: bond prices, long short-term memory, time series forecasting, empirical mode decomposition

Procedia PDF Downloads 136

Authors: Brayden Mi

Abstract:

Reverse time migration has been widely used in the Petroleum exploration industry to reveal subsurface images and to detect rock and fluid properties since the early 1980s. The seismic technology involves the construction of a velocity model through interpretive model construction, seismic tomography, or full waveform inversion, and the application of the reverse-time propagation of acquired seismic data and the original wavelet used in the acquisition. The methodology has matured from 2D, simple media to present-day to handle full 3D imaging challenges in extremely complex geological conditions. Conventional Ultrasound computed tomography (USCT) utilize travel-time-inversion to reconstruct the velocity structure of an organ. With the velocity structure, USCT data can be migrated with the “bend-ray” method, also known as migration. Its seismic application counterpart is called Kirchhoff depth migration, in which the source of reflective energy is traced by ray-tracing and summed to produce a subsurface image. It is well known that ray-tracing-based migration has severe limitations in strongly heterogeneous media and irregular acquisition geometries. Reverse time migration (RTM), on the other hand, fully accounts for the wave phenomena, including multiple arrives and turning rays due to complex velocity structure. It has the capability to fully reconstruct the image detectable in its acquisition aperture. The RTM algorithms typically require a rather accurate velocity model and demand high computing powers, and may not be applicable to real-time imaging as normally required in day-to-day medical operations. However, with the improvement of computing technology, such a computational bottleneck may not present a challenge in the near future. The present-day (RTM) algorithms are typically implemented from a flat datum for the seismic industry. It can be modified to accommodate any acquisition geometry and aperture, as long as sufficient illumination is provided. Such flexibility of RTM can be conveniently implemented for the application in USCT imaging if the spatial coordinates of the transmitters and receivers are known and enough data is collected to provide full illumination. This paper proposes an implementation of a full 3D RTM algorithm for USCT imaging to produce an accurate 3D acoustic image based on the Phase-shift-plus-interpolation (PSPI) method for wavefield extrapolation. In this method, each acquired data set (shot) is propagated back in time, and a known ultrasound wavelet is propagated forward in time, with PSPI wavefield extrapolation and a piece-wise constant velocity model of the organ (breast). The imaging condition is then applied to produce a partial image. Although each image is subject to the limitation of its own illumination aperture, the stack of multiple partial images will produce a full image of the organ, with a much-reduced noise level if compared with individual partial images.

Keywords: illumination, reverse time migration (RTM), ultrasound computed tomography (USCT), wavefield extrapolation

Procedia PDF Downloads 74

Authors: Stanislavs Gendelis, Maris Sinka, Andris Jakovics

Abstract:

Insulation materials made from natural sources have become more popular for the ecologisation of buildings, meaning wide use of such renewable materials. Such natural materials replace synthetic products which consume a large quantity of energy. The most common and the cheapest natural materials in Latvia are cellulose-based (wood and agricultural plants). The ecological aspects of such materials are well known, but experimental data about physical properties remains lacking. In this study, six different samples of wood wool panels and a mixture of hemp shives and lime (hempcrete) are analysed. Thermal conductivity and heat capacity measurements were carried out for wood wool and cement panels using the calibrated hot plate device. Water vapor permeability was tested for hempcrete material by using the gravimetric dry cup method. Studied wood wool panels are eco-friendly and harmless material, which is widely used in the interior design of public and residential buildings, where noise absorption and sound insulation is of importance. They are also suitable for high humidity facilities (e.g., swimming pools). The difference in panels was the width of used wood wool, which is linked to their density. The results of measured thermal conductivity are in a wide range, showing the worsening of properties with the increasing of the wool width (for the least dense 0.066, for the densest 0.091 W/(m·K)). Comparison with mineral insulation materials shows that thermal conductivity for such materials are 2-3 times higher and are comparable to plywood and fibreboard. Measured heat capacity was in a narrower range; here, the dependence on the wool width was not so strong due to the fact that heat capacity value is related to mass, not volume. The resulting heat capacity is a combination of two main components. A comparison of results for different panels allows to select the most suitable sample for a specific application because the dependencies of the thermal insulation and heat capacity properties on the wool width are not the same. Hempcrete is a much denser material compared to conventional thermal insulating materials. Therefore, its use helps to reinforce the structural capacity of the constructional framework, at the same time, it is lightweight. By altering the proportions of the ingredients, hempcrete can be produced as a structural, thermal, or moisture absorbent component. The water absorption and water vapor permeability are the most important properties of these materials. Information about absorption can be found in the literature, but there are no data about water vapor transmission properties. Water vapor permeability was tested for a sample of locally made hempcrete using different air humidity values to evaluate the possible difference. The results show only the slight influence of the air humidity on the water vapor permeability value. The absolute ‘sd value’ measured is similar to mineral wool and wood fiberboard, meaning that due to very low resistance, water vapor passes easily through the material. At the same time, other properties – structural and thermal of the hempcrete is totally different. As a result, an experimentally-based knowledge of thermal and water vapor transmission properties for cellulose-based materials was significantly improved.

Keywords: heat capacity, hemp concrete, thermal conductivity, water vapor transmission, wood wool

Procedia PDF Downloads 221

Authors: Vitor M. Vilas Boas, Cleison D. Silva, Gustavo S. Mafra, Alexandre Trofino Neto

Abstract:

Motor imagery (MI) based brain-computer interfaces (BCI) uses event-related (de)synchronization (ERS/ ERD), typically recorded using electroencephalography (EEG), to translate brain electrical activity into control commands. To mitigate undesirable artifacts and noise measurements on EEG signals, methods based on band-pass filters defined by a specific frequency band (i.e., 8 – 30Hz), such as the Infinity Impulse Response (IIR) filters, are typically used. Spatial techniques, such as Common Spatial Patterns (CSP), are also used to estimate the variations of the filtered signal and extract features that define the imagined motion. The CSP effectiveness depends on the subject's discriminative frequency, and approaches based on the decomposition of the band of interest into sub-bands with smaller frequency ranges (SBCSP) have been suggested to EEG signals classification. However, despite providing good results, the SBCSP approach generally increases the computational cost of the filtering step in IM-based BCI systems. This paper proposes the use of the Fast Fourier Transform (FFT) algorithm in the IM-based BCI filtering stage that implements SBCSP. The goal is to apply the FFT algorithm to reduce the computational cost of the processing step of these systems and to make them more efficient without compromising classification accuracy. The proposal is based on the representation of EEG signals in a matrix of coefficients resulting from the frequency decomposition performed by the FFT, which is then submitted to the SBCSP process. The structure of the SBCSP contemplates dividing the band of interest, initially defined between 0 and 40Hz, into a set of 33 sub-bands spanning specific frequency bands which are processed in parallel each by a CSP filter and an LDA classifier. A Bayesian meta-classifier is then used to represent the LDA outputs of each sub-band as scores and organize them into a single vector, and then used as a training vector of an SVM global classifier. Initially, the public EEG data set IIa of the BCI Competition IV is used to validate the approach. The first contribution of the proposed method is that, in addition to being more compact, because it has a 68% smaller dimension than the original signal, the resulting FFT matrix maintains the signal information relevant to class discrimination. In addition, the results showed an average reduction of 31.6% in the computational cost in relation to the application of filtering methods based on IIR filters, suggesting FFT efficiency when applied in the filtering step. Finally, the frequency decomposition approach improves the overall system classification rate significantly compared to the commonly used filtering, going from 73.7% using IIR to 84.2% using FFT. The accuracy improvement above 10% and the computational cost reduction denote the potential of FFT in EEG signal filtering applied to the context of IM-based BCI implementing SBCSP. Tests with other data sets are currently being performed to reinforce such conclusions.

Keywords: brain-computer interfaces, fast Fourier transform algorithm, motor imagery, sub-band common spatial patterns

Procedia PDF Downloads 128

Authors: Tamar Trop

Abstract:

Bike sharing systems can be important elements of smart cities as they have the potential for impact on multiple levels. These systems can add a significant alternative to other modes of mass transit in cities that are continuously looking for measures to become more livable and maintain their attractiveness for citizens, businesses and tourism. Bike-sharing began in Europe in 1965, and a viable format emerged in the mid-2000s thanks to the introduction of information technology. The rate of growth in bike-sharing schemes and fleets has been very rapid since 2008 and has probably outstripped growth in every other form of urban transport. Today, public bike-sharing systems are available on five continents, including over 700 cities, operating more than 800,000 bicycles at approximately 40,000 docking stations. Since modern bike sharing systems have become prevalent only in the last decade, the existing literature analyzing these systems and their sustainability is relatively new. The purpose of the presented study is to assess the sustainability of these newly emerging transportation systems, by using the Envision™ rating system as a methodological framework and the Israeli 'Tel -O-Fun' – bike sharing project as a case study. The assessment was conducted by project team members. Envision™ is a new guidance and rating system used to assess and improve the sustainability of all types and sizes of infrastructure projects. This tool provides a holistic framework for evaluating and rating the community, environmental, and economic benefits of infrastructure projects over the course of their life cycle. This evaluation method has 60 sustainability criteria divided into five categories: Quality of life, leadership, resource allocation, natural world, and climate and risk. 'Tel -O-Fun' project was launched in Tel Aviv-Yafo on 2011 and today provides about 1,800 bikes for rent, at 180 rental stations across the city. The system is based on a complex computer terminal that is located in the docking stations. The highest-rated sustainable features that the project scored include: (a) Improving quality of life by: offering a low cost and efficient form of public transit, improving community mobility and access, enabling the flexibility of travel within a multimodal transportation system, saving commuters time and money, enhancing public health and reducing air and noise pollution; (b) improving resource allocation by: offering inexpensive and flexible last-mile connectivity, reducing space, materials and energy consumption, reducing wear and tear on public roads, and maximizing the utility of existing infrastructure, and (c) reducing of greenhouse gas emissions from transportation. Overall, 'Tel -O-Fun' project was highly scored as an environmentally sustainable and socially equitable infrastructure. The use of this practical framework for evaluation also yielded various interesting insights on the shortcoming of the system and the characteristics of good solutions. This can contribute to the improvement of the project and may assist planners and operators of bike sharing systems to develop a sustainable, efficient and reliable transportation infrastructure within smart cities.

Keywords: bike sharing, Envision™, sustainability rating system, sustainable infrastructure

Procedia PDF Downloads 340

Authors: Zhuang Ma, Zihan Zhang, Yu Li

Abstract:

Environmental concerns from stakeholders (e.g., governments & customers) have pushed firms to integrate environmental management systems into business processes such as R&D, manufacturing, and marketing. Environmental systems include managing environmental risks and pollution control (e.g., air pollution control, waste-water treatment, noise control, energy recycling & solid waste treatment) through raw material management, the elimination and reduction of contaminants, recycling, and reuse in firms' operational processes. Despite increasing studies on firms' proactive adoption of environmental management, their focus is primarily on large corporations operating in developed economies. Investigations in the environmental management efforts of small and medium-sized enterprises (SMEs) are scarce. This is problematic for SMEs because, unlike large corporations, SMEs have limited awareness, resources, capabilities to adapt their operational routines to address environmental impacts. The purpose of this study is to explore how SMEs develop organizational capabilities through interactions with business partners (e.g., environmental management specialists & customers). Drawing on the resource-based view (RBV) and an organizational capabilities perspective, this study investigates the interactively developed capabilities that allow SMEs to adopt environmental management systems. Using an exploratory approach, the study includes 12 semi-structured interviews with senior managers from four SMEs, two environmental management specialists, and two customers in the pharmaceutical sector in Chongqing, China. Findings of this study include four key organizational capabilities: 1) ‘dynamic marketing’ capability, which allows SMEs to recoup the investments in environmental management systems by developing environmentally friendly products to address customers' ever-changing needs; 2) ‘process improvement’ capability, which allows SMEs to select and adopt the latest technologies from biology, chemistry, new material, and new energy sectors into the production system for improved environmental performance and cost-reductions; and 3) ‘relationship management’ capability which allows SMEs to improve corporate image among the public, social media, government agencies, and customers, who in turn help SMEs to overcome their competitive disadvantages. These interactively developed capabilities help SMEs to address larger competitors' foothold in the local market, reduce market constraints, and exploit competitive advantages in other regions (e.g., Guangdong & Jiangsu) of China. These findings extend the RBV and organizational capabilities perspective; that is, SMEs can develop the essential resources and capabilities required for environmental management through interactions with upstream and downstream business partners. While a limited number of studies did highlight the importance of interactions among SMEs, customers, suppliers, NGOs, industrial associations, and consulting firms, they failed to explore the specific capabilities developed through these interactions. Additionally, the findings can explain how a proactive adoption of environmental management systems could help some SMEs to overcome the institutional and market restraints on their products, thereby springboarding into larger, more environmentally demanding, yet more profitable markets compared with their existing market.

Keywords: capabilities, environmental management systems, interactions, SMEs

Procedia PDF Downloads 180

Authors: Sunil Chandra, Himanshu Rawat, Vikas Gusain, Triparna Barman

Abstract:

Forest fires are a recurrent phenomenon in the Himalayan region owing to the presence of vulnerable forest types, topographical gradients, climatic weather conditions, and anthropogenic pressure. The present study focuses on the identification of forest fire-affected areas in a small part of the West Himalayan region using a differential normalized burnt ratio method and spectral unmixing methods. The study area has a rugged terrain with the presence of sub-tropical pine forest, montane temperate forest, and sub-alpine forest and scrub. The major reason for fires in this region is anthropogenic in nature, with the practice of human-induced fires for getting fresh leaves, scaring wild animals to protect agricultural crops, grazing practices within reserved forests, and igniting fires for cooking and other reasons. The fires caused by the above reasons affect a large area on the ground, necessitating its precise estimation for further management and policy making. In the present study, two approaches have been used for carrying out a burnt area analysis. The first approach followed for burnt area analysis uses a differenced normalized burnt ratio (dNBR) index approach that uses the burnt ratio values generated using the Short-Wave Infrared (SWIR) band and Near Infrared (NIR) bands of the Sentinel-2 image. The results of the dNBR have been compared with the outputs of the spectral mixing methods. It has been found that the dNBR is able to create good results in fire-affected areas having homogenous forest stratum and with slope degree <5 degrees. However, in a rugged terrain where the landscape is largely influenced by the topographical variations, vegetation types, tree density, the results may be largely influenced by the effects of topography, complexity in tree composition, fuel load composition, and soil moisture. Hence, such variations in the factors influencing burnt area assessment may not be effectively carried out using a dNBR approach which is commonly followed for burnt area assessment over a large area. Hence, another approach that has been attempted in the present study utilizes a spectral mixing method where the individual pixel is tested before assigning an information class to it. The method uses a neural network approach utilizing Sentinel-2 bands. The training and testing data are generated from the Sentinel-2 data and the national field inventory, which is further used for generating outputs using ML tools. The analysis of the results indicates that the fire-affected regions and their severity can be better estimated using spectral unmixing methods, which have the capability to resolve the noise in the data and can classify the individual pixel to the precise burnt/unburnt class.

Keywords: categorical data, log linear modeling, neural network, shifting cultivation

Procedia PDF Downloads 54

Authors: Desmond Appiah, Fan Zhang, Shouqi Yuan, Wei Xueyuan, Stephen N. Asomani

Abstract:

The side channel pump has distinctive hydraulic performance characteristics over other vane pumps because of its generation of high pressure heads in only one impeller revolution. Hence, there is soaring utilization and application in the fields of petrochemical, food processing fields, automotive and aerospace fuel pumping where high heads are required at low flows. The side channel pump is characterized by unstable flow because after fluid flows into the impeller passage, it moves into the side channel and comes back to the impeller again and then moves to the next circulation. Consequently, the flow leaves the side channel pump following a helical path. However, the pressure fluctuation exhibited in the flow greatly contributes to the unwanted noise and vibration which is associated with the flow. In this paper, a side channel pump prototype was examined thoroughly through numerical calculations based on SST k-ω turbulence model to ascertain the pressure fluctuation behavior. The pressure fluctuation intensity of the 3D unstable flow dynamics were carefully investigated under different working conditions 0.8QBEP, 1.0 QBEP and 1.2QBEP. The results showed that the pressure fluctuation distribution around the pressure side of the blade is greater than the suction side at the impeller and side channel interface (z=0) for all three operating conditions. Part-load condition 0.8QBEP recorded the highest pressure fluctuation distribution because of the high circulation velocity thus causing an intense exchanged flow between the impeller and side channel. Time and frequency domains spectra of the pressure fluctuation patterns in the impeller and the side channel were also analyzed under the best efficiency point value, QBEP using the solution from the numerical calculations. It was observed from the time-domain analysis that the pressure fluctuation characteristics in the impeller flow passage increased steadily until the flow reached the interrupter which separates low-pressure at the inflow from high pressure at the outflow. The pressure fluctuation amplitudes in the frequency domain spectrum at the different monitoring points depicted a gentle decreasing trend of the pressure amplitudes which was common among the operating conditions. The frequency domain also revealed that the main excitation frequencies occurred at 600Hz, 1200Hz, and 1800Hz and continued in the integers of the rotating shaft frequency. Also, the mass flow exchange plots indicated that the side channel pump is characterized with many vortex flows. Operating conditions 0.8QBEP, 1.0 QBEP depicted less and similar vortex flow while 1.2Q recorded many vortex flows around the inflow, middle and outflow regions. The results of the numerical calculations were finally verified experimentally. The performance characteristics curves from the simulated results showed that 0.8QBEP working condition recorded a head increase of 43.03% and efficiency decrease of 6.73% compared to 1.0QBEP. It can be concluded that for industrial applications where the high heads are mostly required, the side channel pump can be designed to operate at part-load conditions. This paper can serve as a source of information in order to optimize a reliable performance and widen the applications of the side channel pumps.

Keywords: exchanged flow, pressure fluctuation, numerical simulation, side channel pump

Procedia PDF Downloads 136

Authors: Zhao Gao, Eran Edirisinghe

Abstract:

The production of an accurate sketch of a suspect based on a verbal description obtained from a witness is an essential task for most criminal investigations. The criminal investigation system employs specifically trained professional artists to manually draw a facial image of the suspect according to the descriptions of an eyewitness for subsequent identification. Within the advancement of Deep Learning, Recurrent Neural Networks (RNN) have shown great promise in Natural Language Processing (NLP) tasks. Additionally, Generative Adversarial Networks (GAN) have also proven to be very effective in image generation. In this study, a trained GAN conditioned on textual features such as keywords automatically encoded from a verbal description of a human face using an RNN is used to generate photo-realistic facial images for criminal investigations. The intention of the proposed system is to map corresponding features into text generated from verbal descriptions. With this, it becomes possible to generate many reasonably accurate alternatives to which the witness can use to hopefully identify a suspect from. This reduces subjectivity in decision making both by the eyewitness and the artist while giving an opportunity for the witness to evaluate and reconsider decisions. Furthermore, the proposed approach benefits law enforcement agencies by reducing the time taken to physically draw each potential sketch, thus increasing response times and mitigating potentially malicious human intervention. With publically available 'CelebFaces Attributes Dataset' (CelebA) and additionally providing verbal description as training data, the proposed architecture is able to effectively produce facial structures from given text. Word Embeddings are learnt by applying the RNN architecture in order to perform semantic parsing, the output of which is fed into the GAN for synthesizing photo-realistic images. Rather than the grid search method, a metaheuristic search based on genetic algorithms is applied to evolve the network with the intent of achieving optimal hyperparameters in a fraction the time of a typical brute force approach. With the exception of the ‘CelebA’ training database, further novel test cases are supplied to the network for evaluation. Witness reports detailing criminals from Interpol or other law enforcement agencies are sampled on the network. Using the descriptions provided, samples are generated and compared with the ground truth images of a criminal in order to calculate the similarities. Two factors are used for performance evaluation: The Structural Similarity Index (SSIM) and the Peak Signal-to-Noise Ratio (PSNR). A high percentile output from this performance matrix should attribute to demonstrating the accuracy, in hope of proving that the proposed approach can be an effective tool for law enforcement agencies. The proposed approach to criminal facial image generation has potential to increase the ratio of criminal cases that can be ultimately resolved using eyewitness information gathering.

Keywords: RNN, GAN, NLP, facial composition, criminal investigation

Procedia PDF Downloads 161

Authors: Ja-Keoung Koo, Kensuke Nakamura, Hyohun Kim, Dongwha Shin, Yeonseok Kim, Ji-Su Ahn, Byung-Woo Hong

Abstract:

The machine learning techniques based on a convolutional neural network (CNN) have been actively developed and successfully applied to a variety of image analysis tasks including reconstruction, noise reduction, resolution enhancement, segmentation, motion estimation, object recognition. The classical visual information processing that ranges from low level tasks to high level ones has been widely developed in the deep learning framework. It is generally considered as a challenging problem to derive visual interpretation from high dimensional imagery data. A CNN is a class of feed-forward artificial neural network that usually consists of deep layers the connections of which are established by a series of non-linear operations. The CNN architecture is known to be shift invariant due to its shared weights and translation invariance characteristics. However, it is often computationally intractable to optimize the network in particular with a large number of convolution layers due to a large number of unknowns to be optimized with respect to the training set that is generally required to be large enough to effectively generalize the model under consideration. It is also necessary to limit the size of convolution kernels due to the computational expense despite of the recent development of effective parallel processing machinery, which leads to the use of the constantly small size of the convolution kernels throughout the deep CNN architecture. However, it is often desired to consider different scales in the analysis of visual features at different layers in the network. Thus, we propose a CNN model where different sizes of the convolution kernels are applied at each layer based on the random projection. We apply random filters with varying sizes and associate the filter responses with scalar weights that correspond to the standard deviation of the random filters. We are allowed to use large number of random filters with the cost of one scalar unknown for each filter. The computational cost in the back-propagation procedure does not increase with the larger size of the filters even though the additional computational cost is required in the computation of convolution in the feed-forward procedure. The use of random kernels with varying sizes allows to effectively analyze image features at multiple scales leading to a better generalization. The robustness and effectiveness of the proposed CNN based on random kernels are demonstrated by numerical experiments where the quantitative comparison of the well-known CNN architectures and our models that simply replace the convolution kernels with the random filters is performed. The experimental results indicate that our model achieves better performance with less number of unknown weights. The proposed algorithm has a high potential in the application of a variety of visual tasks based on the CNN framework. Acknowledgement—This work was supported by the MISP (Ministry of Science and ICT), Korea, under the National Program for Excellence in SW (20170001000011001) supervised by IITP, and NRF-2014R1A2A1A11051941, NRF2017R1A2B4006023.

Keywords: deep learning, convolutional neural network, random kernel, random projection, dimensionality reduction, object recognition

Procedia PDF Downloads 289

Authors: Matej Komanec, Tomas Nemecek, Dmytro Suslov, Petr Chvojka, Stanislav Zvanovec

Abstract:

Nonlinear phenomena in the near- and mid-infrared region are attracting scientific attention mainly due to the supercontinuum generation possibilities and subsequent utilizations for ultra-wideband applications like e.g. absorption spectroscopy or optical coherence tomography. Thulium-based fiber lasers provide access to high-power ultrashort pump pulses in the vicinity of 2000 nm, which can be easily exploited for various nonlinear applications. The paper presents a simulation and experimental study of a pulsed thulium laser based for near-infrared (NIR) and mid-infrared (MIR) nonlinear applications in specialty optical fibers. In the first part of the paper the thulium laser is discussed. The thulium laser is based on a gain-switched seed-laser and a series of amplification stages for obtaining output peak powers in the order of kilowatts for pulses shorter than 200 ps in full-width at half-maximum. The pulsed thulium laser is first studied in a simulation software, focusing on seed-laser properties. Afterward, a pre-amplification thulium-based stage is discussed, with the focus of low-noise signal amplification, high signal gain and eliminating pulse distortions during pulse propagation in the gain medium. Following the pre-amplification stage a second gain stage is evaluated with incorporating a thulium-fiber of shorter length with increased rare-earth dopant ratio. Last a power-booster stage is analyzed, where the peak power of kilowatts should be achieved. Examples of analytical study are further validated by the experimental campaign. The simulation model is further corrected based on real components – parameters such as real insertion-losses, cross-talks, polarization dependencies, etc. are included. The second part of the paper evaluates the utilization of nonlinear phenomena, their specific features at the vicinity of 2000 nm, compared to e.g. 1550 nm, and presents supercontinuum modelling, based on the thulium laser pulsed output. Supercontinuum generation simulation is performed and provides reasonably accurate results, once fiber dispersion profile is precisely defined and fiber nonlinearity is known, furthermore input pulse shape and peak power must be known, which is assured thanks to the experimental measurement of the studied thulium pulsed laser. The supercontinuum simulation model is put in relation to designed and characterized specialty optical fibers, which are discussed in the third part of the paper. The focus is placed on silica and mainly on non-silica fibers (fluoride, chalcogenide, lead-silicate) in their conventional, microstructured or tapered variants. Parameters such as dispersion profile and nonlinearity of exploited fibers were characterized either with an accurate model, developed in COMSOL software or by direct experimental measurement to achieve even higher precision. The paper then combines all three studied topics and presents a possible application of such a thulium pulsed laser system working with specialty optical fibers.

Keywords: nonlinear phenomena, specialty optical fibers, supercontinuum generation, thulium laser

Procedia PDF Downloads 321

Authors: Shilpi Bora, Abhirup Chatterjee, Debkumar Chakrabarti

Abstract:

Organizations of all the sectors around the world are increasingly revisiting their workplace strategies with due concern for women working therein. Limited office space and rigid work arrangements contribute to lesser job satisfaction and greater work impoundments for any organization. Flexible workspace strategies are indispensable to accommodate the progressive rise of modular workstations and involvement of women. Today’s generation of employees deserves malleable office environments with employee-friendly job conditions and strategies. The workplace nowadays stands on rapid organizational changes in progressive and flexible work culture. Occupational well-being practices need to keep pace with the rapid changes in office-based work. Working at the office (workspace) with awkward postures or for long periods can cause pain, discomfort, and injury. The world is stirring towards the era of globalization and progress. The 4000 women police personnel constitute less than one per cent of the total police strength of India. Lots of innovative fields are growing fast, and it is important that we should accommodate women in those arenas. The timeworn trends should be set apart to set out for fresh opportunities and possibilities of development and success through more involvement of women in the workplace. The notion of women policing is gaining position throughout the world, and various countries are putting solemn efforts to mainstream women in policing. As the role of women policing in a society is budding, and thus it is also notable that the accessibility of women at general police stations should be considered. Accordingly, the impact of workspace at police station on the employee productivity has been widely deliberated as a crucial contributor to employee satisfaction leading to better functional motivation. Thus the present research aimed to look into the office workstation design of police station with reference to womanhood specific issues to uplift occupational wellbeing of the policewomen. Personal interview and individual responses collected through administering to a subjective assessment questionnaire on thirty women police as well as to have their views on these issues by purposive non-probability sampling of women police personnel of different ranks posted in Guwahati, Assam, India. Scrutiny of the collected data revealed that office design has a substantial impact on the policewomen job satisfaction in the police station. In this study, the workspace was designed in such a way that the set of factors would impact on the individual to ensure increased productivity. Office design such as furniture, noise, temperature, lighting and spatial arrangement were considered. The primary feature which affected the productivity of policewomen was the furniture used in the workspace, which was found to disturb the everyday and overall productivity of policewomen. Therefore, it was recommended to have proper and adequate ergonomics design intervention to improve the office design for better performance. This type of study is today’s need-of-the-hour to empower women and facilitate their inner talent to come up in service of the nation. The office workspace design also finds critical importance at several other occupations also – where office workstation needs further improvement.

Keywords: office workspace design, policewomen, womanhood concerns at workspace, occupational wellbeing

Procedia PDF Downloads 225

Authors: Ammar Alali, Mahmoud Abughaban

Abstract:

Cost reduction and drilling optimization is the goal of many drilling operators. Historically, stuck pipe incidents were a major segment of non-productive time (NPT) associated costs. Traditionally, stuck pipe problems are part of the operations and solved post-sticking. However, the real key to savings and success is in predicting the stuck pipe incidents and avoiding the conditions leading to its occurrences. Previous attempts in stuck-pipe predictions have neglected the local geology of the problem. The proposed predictive tool utilizes geophysical data processing techniques and Machine Learning (ML) algorithms to predict drilling activities events in real-time using surface drilling data with minimum computational power. The method combines two types of analysis: (1) real-time prediction, and (2) cause analysis. Real-time prediction aggregates the input data, including historical drilling surface data, geological formation tops, and petrophysical data, from wells within the same field. The input data are then flattened per the geological formation and stacked per stuck-pipe incidents. The algorithm uses two physical methods (stacking and flattening) to filter any noise in the signature and create a robust pre-determined pilot that adheres to the local geology. Once the drilling operation starts, the Wellsite Information Transfer Standard Markup Language (WITSML) live surface data are fed into a matrix and aggregated in a similar frequency as the pre-determined signature. Then, the matrix is correlated with the pre-determined stuck-pipe signature for this field, in real-time. The correlation used is a machine learning Correlation-based Feature Selection (CFS) algorithm, which selects relevant features from the class and identifying redundant features. The correlation output is interpreted as a probability curve of stuck pipe incidents prediction in real-time. Once this probability passes a fixed-threshold defined by the user, the other component, cause analysis, alerts the user of the expected incident based on set pre-determined signatures. A set of recommendations will be provided to reduce the associated risk. The validation process involved feeding of historical drilling data as live-stream, mimicking actual drilling conditions, of an onshore oil field. Pre-determined signatures were created for three problematic geological formations in this field prior. Three wells were processed as case studies, and the stuck-pipe incidents were predicted successfully, with an accuracy of 76%. This accuracy of detection could have resulted in around 50% reduction in NPT, equivalent to 9% cost saving in comparison with offset wells. The prediction of stuck pipe problem requires a method to capture geological, geophysical and drilling data, and recognize the indicators of this issue at a field and geological formation level. This paper illustrates the efficiency and the robustness of the proposed cross-disciplinary approach in its ability to produce such signatures and predicting this NPT event.

Keywords: drilling optimization, hazard prediction, machine learning, stuck pipe

Procedia PDF Downloads 227

Authors: Sanguk Nam, Van Ha Nguyen, Hanjung Song

Abstract:

For low-power applications such as adapters for portable devices and USB chargers, the primary side regulation (PSR) fly-back converter is widely used in lieu of the conventional fly-back converter using opto-coupler because of its simpler structure and lower cost. In the literature, there has been studies focusing on the design of PSR circuit; however, the conventional sensing method in PSR circuit using RC delay has a lower accuracy as compared to the conventional fly-back converter using opto-coupler. In this paper, we propose a novel PWM/PFM controller using new sensing technique for the PSR fly-back converter which can control an accurate output voltage. The conventional PSR circuit can sense the output voltage information from the auxiliary winding to regulate the duty cycle of the clock that control the output voltage. In the sensing signal waveform, there has two transient points at time the voltage equals to Vout+VD and Vout, respectively. In other to sense the output voltage, the PSR circuit must detect the time at which the current of the diode at the output equals to zero. In the conventional PSR flyback-converter, the sensing signal at this time has a non-sharp-negative slope that might cause a difficulty in detecting the output voltage information since a delay of sensing signal or switching clock may exist which brings out an unstable operation of PSR fly-back converter. In this paper instead of detecting output voltage at a non-sharp-negative slope, a sharp-positive slope is used to sense the proper information of the output voltage. The proposed PRS circuit consists of a saw-tooth generator, a summing circuit, a sample and hold circuit and a peak detector. Besides, there is also the start-up circuit which protects the chip from high surge current when the converter is turned on. Additionally, to reduce the standby power loss, a second mode which operates in a low frequency is designed beside the main mode at high frequency. In general, the operation of the proposed PSR circuit can be summarized as following: At the time the output information is sensed from the auxiliary winding, a saw-tooth signal from the saw-tooth generator is generated. Then, both of these signals are summed using a summing circuit. After this process, the slope of the peak of the sensing signal at the time diode current is zero becomes positive and sharp that make the peak easy to detect. The output of the summing circuit then is fed into a peak detector and the sample and hold circuit; hence, the output voltage can be properly sensed. By this way, we can sense more accurate output voltage information and extend margin even circuit is delayed or even there is the existence of noise by using only a simple circuit structure as compared with conventional circuits while the performance can be sufficiently enhanced. Circuit verification was carried out using 0.35μm 700V Magnachip process. The simulation result of sensing signal shows a maximum error of 5mV under various load and line conditions which means the operation of the converter is stable. As compared to the conventional circuit, we achieved very small error only used analog circuits compare with conventional circuits. In this paper, a PWM/PFM controller using a simple and effective sensing method for PSR fly-back converter has been presented in this paper. The circuit structure is simple as compared with the conventional designs. The gained results from simulation confirmed the idea of the design

Keywords: primary side regulation, PSR, sensing technique, peak detector, PWM/PFM control, fly-back converter

Procedia PDF Downloads 338

Authors: Prajakta Patil, Yash Huzurbazar, Abhijeet Shinde

Abstract:

Aim: To develop a treatment protocol for the management of tinnitus through multi-modality brain stimulation. Methodology: Present study included 33 adults with unilateral (31 subjects) and bilateral (2 subjects) chronic tinnitus with and/or without hearing loss independent of their etiology. The Treatment protocol included 5 consecutive sessions with follow-up of 6 months. Each session was divided into 3 parts: • Pre-treatment: a) Informed consent b) Pitch and loudness matching. • Treatment: Bimanual paper pen task with tinnitus masking for 30 minutes. • Post-treatment: a) Pitch and loudness matching b) Directive counseling and obtaining feedback. Paper-pen task is to be performed bimanually that included carrying out two different writing activities in different context. The level of difficulty of the activities was increased in successive sessions. Narrowband noise of a frequency same as that of tinnitus was presented at 10 dBSL of tinnitus for 30 minutes simultaneously in the ear with tinnitus. Result: The perception of tinnitus was no longer present in 4 subjects while in remaining subjects it reduced to an intensity that its perception no longer troubled them without causing residual facilitation. In all subjects, the intensity of tinnitus decreased by an extent of 45 dB at an average. However, in few subjects, the intensity of tinnitus also decreased by more than 45 dB. The approach resulted in statistically significant reductions in Tinnitus Functional Index and Tinnitus Handicap Inventory scores. The results correlate with pre and post treatment score of Tinnitus Handicap Inventory that dropped from 90% to 0%. Discussion: Brain mapping(qEEG) Studies report that there is multiple parallel overlapping of neural subnetworks in the non-auditory areas of the brain which exhibits abnormal, constant and spontaneous neural activity involved in the perception of tinnitus with each subnetwork and area reflecting a specific aspect of tinnitus percept. The paper pen task and directive counseling are designed and delivered respectively in a way that is assumed to induce normal, rhythmically constant and premeditated neural activity and mask the abnormal, constant and spontaneous neural activity in the above-mentioned subnetworks and the specific non-auditory area. Counseling was focused on breaking the vicious cycle causing and maintaining the presence of tinnitus. Diverting auditory attention alone is insufficient to reduce the perception of tinnitus. Conscious awareness of tinnitus can be suppressed when individuals engage in cognitively demanding tasks of non-auditory nature as the paper pen task used in the present study. To carry out this task selective, divided, sustained, simultaneous and split attention act cumulatively. Bimanual paper pen task represents a top-down activity which underlies brain’s ability to selectively attend to the bimanual written activity as a relevant stimulus and to ignore tinnitus that is the irrelevant stimuli in the present study. Conclusion: The study suggests that this novel treatment approach is cost effective, time saving and efficient to vanish the tinnitus or to reduce the intensity of tinnitus to a negligible level and thereby eliminating the negative reactions towards tinnitus.

Keywords: multi-modality brain stimulation, neural subnetworks, non-auditory areas, paper-pen task, top-down activity

Procedia PDF Downloads 147

Authors: Rodrigo Arnaldo Scarpel

Abstract:

Trend following is a widely accepted investment strategy that adopts a rule-based trading mechanism that rather than striving to predict market direction or on information gathering to decide when to buy and when to sell a stock. Thus, in trend following one must respond to market’s movements that has recently happen and what is currently happening, rather than on what will happen. Optimally, in trend following strategy, is to catch a bull market at its early stage, ride the trend, and liquidate the position at the first evidence of the subsequent bear market. For applying the trend following strategy one needs to find the trend and identify trade signals. In order to avoid false signals, i.e., identify fluctuations of short, mid and long terms and to separate noise from real changes in the trend, most academic works rely on moving averages and other technical analysis indicators, such as the moving average convergence divergence (MACD) and the relative strength index (RSI) to uncover intelligible stock trading rules following trend following strategy philosophy. Recently, some works has applied machine learning techniques for trade rules discovery. In those works, the process of rule construction is based on evolutionary learning which aims to adapt the rules to the current environment and searches for the global optimum rules in the search space. In this work, instead of focusing on the usage of machine learning techniques for creating trading rules, a time series trend classification employing a semi-supervised approach was used to early identify both the beginning and the end of upward and downward trends. Such classification model can be employed to identify trade signals and the decision-making procedure is that if an up-trend (down-trend) is identified, a buy (sell) signal is generated. Semi-supervised learning is used for model training when only part of the data is labeled and Semi-supervised classification aims to train a classifier from both the labeled and unlabeled data, such that it is better than the supervised classifier trained only on the labeled data. For illustrating the proposed approach, it was employed daily trade information, including the open, high, low and closing values and volume from January 1, 2000 to December 31, 2022, of the São Paulo Exchange Composite index (IBOVESPA). Through this time period it was visually identified consistent changes in price, upwards or downwards, for assigning labels and leaving the rest of the days (when there is not a consistent change in price) unlabeled. For training the classification model, a pseudo-label semi-supervised learning strategy was used employing different technical analysis indicators. In this learning strategy, the core is to use unlabeled data to generate a pseudo-label for supervised training. For evaluating the achieved results, it was considered the annualized return and excess return, the Sortino and the Sharpe indicators. Through the evaluated time period, the obtained results were very consistent and can be considered promising for generating the intended trading signals.

Keywords: evolutionary learning, semi-supervised classification, time series data, trading signals generation

Procedia PDF Downloads 89

Authors: Viktor D. Dubrovin, Masoud Mollaee, Jie Zong, Xiushan Zhu, Nasser Peyghambarian

Abstract:

Glasses doped with high rare-earth (RE) elements concentration attracted considerable attention since the middle of the 20th century due to their particular magneto-optical properties. Such glasses exhibit the Faraday effect in which the polarization plane of a linearly polarized light beam is rotated by the interaction between the incident light and the magneto-optical material. That effect found application in optical isolators that are useful for laser systems, which can prevent back reflection of light into lasers or optical amplifiers and reduce signal instability and noise. Glasses are of particular interest since they are cost-effective and can be formed into fibers, thus breaking the limits of traditional bulk optics requiring optical coupling for use with fiber-optic systems. The advent of high-power fiber lasers operating near 2µm revealed a necessity in the development of all fiber isolators for this region. Ce³⁺, Pr³⁺, Dy³⁺, and Tb³⁺ ions provide the biggest contribution to the Verdet constant value of optical materials among the RE. It is known that Pr³⁺ and Tb³⁺ ions have strong absorption bands near 2 µm, thus making Dy³⁺ and Ce³⁺ the only prospective candidates for fiber isolator operating in that region. Due to the high tendency of Ce³⁺ ions pass to Ce⁴⁺ during the synthesis, glasses with high cerium content usually suffers from Ce⁴⁺ ions absorption extending from visible to IR. Additionally, Dy³⁺ (₆H¹⁵/²) same as Ho³⁺ (⁵I₈) ions, have the largest effective magnetic moment (µeff = 10.6 µB) among the RE ions that starts to play the key role if the operating region is far from 4fⁿ→ 4fⁿ⁻¹5 d¹ electric-dipole transition relevant to the Faraday Effect. Considering the high effective magnetic moment value of Er³⁺ ions (µeff = 9.6 µB) that is 3rd after Dy³⁺/ Ho³⁺ and Tb³⁺, it is possible to assume that Er³⁺ doped glasses should exhibit Verdet constant value near 2µm that is comparable with one of Dy doped glasses. Thus, partial replacement of Dy³⁺ on Er³⁺ ions has been performed, keeping the overall concentration of Re₂O₃ equal to 70 wt.% (30.6 mol.%). Al₂O₃-B₂O₃-SiO₂-30.6RE₂O₃ (RE= Er, Dy) glasses had been synthesized, and their thermal, spectral, optical, structural, and magneto-optical properties had been studied. Glasses synthesis had been conducted in Pt crucibles for 3h at 1500 °C. The obtained melt was poured into preheated up to 400 °C mold and annealed from 800 oC to room temperature for 12h with 1h dwell. The mass of obtained glass samples was about 200g. Shown that the difference between crystallization and glass transition temperature is about 150 oC, even taking into account the fact that high content of RE₂O₃ leads to glass network depolymerization. Verdet constant of Al₂O₃-B₂O₃-SiO₂-30.6RE₂O₃ glasses for wavelength 1950 nm can reach more than 5.9 rad/(T*m), which is among the highest number reported for a paramagnetic glass at this wavelength. The refractive index value was found to be equal to 1.7545 at 633 nm. Our experimental results show that Al₂O₃-B₂O₃-SiO₂-30.6RE₂O₃ glasses with high Dy₂O₃ content are expected to be promising material for use as highly effective Faraday isolators and modulators of electromagnetic radiation in the 2μm region.

Keywords: oxide glass, magneto-optical, dysprosium, erbium, Faraday rotator, boron-aluminosilicate system

Procedia PDF Downloads 114

Authors: Daniel Rowe, Christopher R. Vogel, Richard H. J. Willden

Abstract:

The University of Oxford’s recirculating water flume is a combined wave and current test tank with a 1 m depth, 1.1 m width, and 10 m long working section, and is capable of flow speeds up to 1 ms−1 . This study documents the hydrodynamic characteristics of the facility in preparation for experimental testing of horizontal axis tidal stream turbine models. The turbine to be tested has a rotor diameter of 0.6 m and is a modified version of one of two model-scale turbines tested in previous experimental campaigns. An Acoustic Doppler Velocimeter (ADV) was used to measure the flow at high temporal resolution at various locations throughout the flume, enabling the spatial uniformity and turbulence flow parameters to be investigated. The mean velocity profiles exhibited high levels of spatial uniformity at the design speed of the flume, 0.6 ms−1 , with variations in the three-dimensional velocity components on the order of ±1% at the 95% confidence level, along with a modest streamwise acceleration through the measurement domain, a target 5 m working section of the flume. A high degree of uniformity was also apparent for the turbulence intensity, with values ranging between 1-2% across the intended swept area of the turbine rotor. The integral scales of turbulence exhibited a far higher degree of variation throughout the water column, particularly in the streamwise and vertical scales. This behaviour is believed to be due to the high signal noise content leading to decorrelation in the sampling records. To achieve more realistic levels of vertical velocity shear in the flume, a simple procedure to practically generate target vertical shear profiles in open-channel flows is described. Here, the authors arranged a series of non-uniformly spaced parallel bars placed across the width of the flume and normal to the onset flow. By adjusting the resistance grading across the height of the working section, the downstream profiles could be modified accordingly, characterised by changes in the velocity profile power law exponent, 1/n. Considering the significant temporal variation in a tidal channel, the choice of the exponent denominator, n = 6 and n = 9, effectively provides an achievable range around the much-cited value of n = 7 observed at many tidal sites. The resulting flow profiles, which we intend to use in future turbine tests, have been characterised in detail. The results indicate non-uniform vertical shear across the survey area and reveal substantial corner flows, arising from the differential shear between the target vertical and cross-stream shear profiles throughout the measurement domain. In vertically sheared flow, the rotor-equivalent turbulence intensity ranges between 3.0-3.8% throughout the measurement domain for both bar arrangements, while the streamwise integral length scale grows from a characteristic dimension on the order of the bar width, similar to the flow downstream of a turbulence-generating grid. The experimental tests are well-defined and repeatable and serve as a reference for other researchers who wish to undertake similar investigations.

Keywords: acoustic doppler Velocimeter, experimental hydrodynamics, open-channel flow, shear profiles, tidal stream turbines

Procedia PDF Downloads 86

Authors: Amine Ganibardi, Cherif Arab Ali

Abstract:

This paper addresses the data cleaning issue as a part of web usage data preprocessing within the scope of Web Usage Mining. Weblog data recorded by web servers within log files reflect usage activity, i.e., End-users’ clicks and underlying user-agents’ hits. As Web Usage Mining is interested in End-users’ behavior, user-agents’ hits are referred to as noise to be cleaned-off before mining. Filtering hits from clicks is not trivial for two reasons, i.e., a server records requests interlaced in sequential order regardless of their source or type, website resources may be set up as requestable interchangeably by end-users and user-agents. The current methods are content-centric based on filtering heuristics of relevant/irrelevant items in terms of some cleaning attributes, i.e., website’s resources filetype extensions, website’s resources pointed by hyperlinks/URIs, http methods, user-agents, etc. These methods need exhaustive extra-weblog data and prior knowledge on the relevant and/or irrelevant items to be assumed as clicks or hits within the filtering heuristics. Such methods are not appropriate for dynamic/responsive Web for three reasons, i.e., resources may be set up to as clickable by end-users regardless of their type, website’s resources are indexed by frame names without filetype extensions, web contents are generated and cancelled differently from an end-user to another. In order to overcome these constraints, a clustering-based cleaning method centered on the logging structure is proposed. This method focuses on the statistical properties of the logging structure at the requested and referring resources attributes levels. It is insensitive to logging content and does not need extra-weblog data. The used statistical property takes on the structure of the generated logging feature by webpage requests in terms of clicks and hits. Since a webpage consists of its single URI and several components, these feature results in a single click to multiple hits ratio in terms of the requested and referring resources. Thus, the clustering-based method is meant to identify two clusters based on the application of the appropriate distance to the frequency matrix of the requested and referring resources levels. As the ratio clicks to hits is single to multiple, the clicks’ cluster is the smallest one in requests number. Hierarchical Agglomerative Clustering based on a pairwise distance (Gower) and average linkage has been applied to four logfiles of dynamic/responsive websites whose click to hits ratio range from 1/2 to 1/15. The optimal clustering set on the basis of average linkage and maximum inter-cluster inertia results always in two clusters. The evaluation of the smallest cluster referred to as clicks cluster under the terms of confusion matrix indicators results in 97% of true positive rate. The content-centric cleaning methods, i.e., conventional and advanced cleaning, resulted in a lower rate 91%. Thus, the proposed clustering-based cleaning outperforms the content-centric methods within dynamic and responsive web design without the need of any extra-weblog. Such an improvement in cleaning quality is likely to refine dependent analysis.

Keywords: clustering approach, data cleaning, data preprocessing, weblog data, web usage data

Procedia PDF Downloads 170

Authors: Natalie Riley, Anita Quigley, Robert M. I. Kapsa, George W. Greene

Abstract:

As the fields of medicine and bionics grow rapidly in technological advancement, the future and success of it depends on the ability to effectively interface between the artificial and the biological worlds. The biggest obstacle when it comes to implantable, electronic medical devices, is maintaining a ‘clean’, low noise electrical connection that allows for efficient sharing of electrical information between the artificial and biological systems. Implant fouling occurs with the adhesion and accumulation of proteins and various cell types as a result of the immune response to protect itself from the foreign object, essentially forming an electrical insulation barrier that often leads to implant failure over time. Lubricin (LUB) functions as a major boundary lubricant in articular joints, a unique glycoprotein with impressive anti-adhesive properties that self-assembles to virtually any substrate to form a highly ordered, ‘telechelic’ polymer brush. LUB does not passivate electroactive surfaces which makes it ideal, along with its innate biocompatibility, as a coating for implantable bionic electrodes. It is the aim of the study to investigate LUB’s anti-fouling properties and its potential as a safe, bioinspired material for coating applications to enhance the performance and longevity of implantable medical devices as well as reducing the frequency of implant replacement surgeries. Native, bovine-derived LUB (N-LUB) and recombinant LUB (R-LUB) were applied to gold-coated mylar surfaces. Fibroblast, chondrocyte and neural cell types were cultured and grown on the coatings under both passive and electrically stimulated conditions to test the stability and anti-adhesive property of the LUB coating in the presence of an electric field. Lactate dehydrogenase (LDH) assays were conducted as a directly proportional cell population count on each surface along with immunofluorescent microscopy to visualize cells. One-way analysis of variance (ANOVA) with post-hoc Tukey’s test was used to test for statistical significance. Under both passive and electrically stimulated conditions, LUB significantly reduced cell attachment compared to bare gold. Comparing the two coating types, R-LUB reduced cell attachment significantly compared to its native counterpart. Immunofluorescent micrographs visually confirmed LUB’s antiadhesive property, R-LUB consistently demonstrating significantly less attached cells for both fibroblasts and chondrocytes. Preliminary results investigating neural cells have so far demonstrated that R-LUB has little effect on reducing neural cell attachment; the study is ongoing. Recombinant LUB coatings demonstrated impressive anti-adhesive properties, reducing cell attachment in fibroblasts and chondrocytes. These findings and the availability of recombinant LUB brings into question the results of previous experiments conducted using native-derived LUB, its potential not adequately represented nor realized due to unknown factors and impurities that warrant further study. R-LUB is stable and maintains its anti-fouling property under electrical stimulation, making it suitable for electroactive surfaces.

Keywords: anti-fouling, bioinspired, cell attachment, lubricin

Procedia PDF Downloads 124

Authors: Shayan Mohajer Hamidi

Abstract:

Federated Learning (FL) has emerged as a promising paradigm for training machine learning models on distributed data without the need for centralized data aggregation. In the realm of wireless communications, FL has the potential to leverage the vast amounts of data generated by wireless devices to improve model performance and enable intelligent applications. However, the fairness aspect of FL in wireless communications remains largely unexplored. This abstract presents an idea for fair federated learning in wireless communications, addressing the challenges of imbalanced data distribution, privacy preservation, and resource allocation. Firstly, the proposed approach aims to tackle the issue of imbalanced data distribution in wireless networks. In typical FL scenarios, the distribution of data across wireless devices can be highly skewed, resulting in unfair model updates. To address this, we propose a weighted aggregation strategy that assigns higher importance to devices with fewer samples during the aggregation process. By incorporating fairness-aware weighting mechanisms, the proposed approach ensures that each participating device's contribution is proportional to its data distribution, thereby mitigating the impact of data imbalance on model performance. Secondly, privacy preservation is a critical concern in federated learning, especially in wireless communications where sensitive user data is involved. The proposed approach incorporates privacy-enhancing techniques, such as differential privacy, to protect user privacy during the model training process. By adding carefully calibrated noise to the gradient updates, the proposed approach ensures that the privacy of individual devices is preserved without compromising the overall model accuracy. Moreover, the approach considers the heterogeneity of devices in terms of computational capabilities and energy constraints, allowing devices to adaptively adjust the level of privacy preservation to strike a balance between privacy and utility. Thirdly, efficient resource allocation is crucial for federated learning in wireless communications, as devices operate under limited bandwidth, energy, and computational resources. The proposed approach leverages optimization techniques to allocate resources effectively among the participating devices, considering factors such as data quality, network conditions, and device capabilities. By intelligently distributing the computational load, communication bandwidth, and energy consumption, the proposed approach minimizes resource wastage and ensures a fair and efficient FL process in wireless networks. To evaluate the performance of the proposed fair federated learning approach, extensive simulations and experiments will be conducted. The experiments will involve a diverse set of wireless devices, ranging from smartphones to Internet of Things (IoT) devices, operating in various scenarios with different data distributions and network conditions. The evaluation metrics will include model accuracy, fairness measures, privacy preservation, and resource utilization. The expected outcomes of this research include improved model performance, fair allocation of resources, enhanced privacy preservation, and a better understanding of the challenges and solutions for fair federated learning in wireless communications. The proposed approach has the potential to revolutionize wireless communication systems by enabling intelligent applications while addressing fairness concerns and preserving user privacy.

Keywords: federated learning, wireless communications, fairness, imbalanced data, privacy preservation, resource allocation, differential privacy, optimization

Procedia PDF Downloads 75