Search

This paper presents a design of a single stage 5W X-Band GaN HEMT Power Amplifier using a die Cree model CGHV1J006D. The amplifier is optimized for the frequency range of 11.5GHz to 11.9GHz. Load pull analysis is performed on the active device to identify the appropriate load impedance and source impedance terminations. With this proposed design approach, it was possible to suppress the power leakage into higher order harmonics and achieve 15.26dB linear small signal gain with more than +37dBm output power at 3dB gain compression with a corresponding maximum power-added efficiency of 46.60% and drain efficiency of greater than 49%.

This paper presents system concept and payload design of very high throughput satellite system based on beam hopping that is time domain switching in multicolour reuse concept. High throughput satellites (HTS), with large aggregate bandwidth, act as a backbone to support terrestrial communication in remote and unconnected areas. The Very high throughput satellites (VHTS), aiming at several hundred Gbps of capacity, employing hundreds of spot beams over coverage area, reduces the cost per bit of space infrastructure. The use of time switched beams (beam hopping, BH) in a VHTS system offers significant advantage by providing system flexibility with reduced space resources. This paper presents the system concept and payload design of a time switched VHTS system for Indian region; its comparison with the conventional frequency re-use (FDMA) system; and associated technological challenges. The time switched VHTS system shows reduction in payload mass and power requirements, with same spectrum and aggregate system throughput.

Wireless sensor networks are infrastructure-less network with numerous sensors to detect environmental conditions. It passes the data to the base station or sinks through relay nodes. Wireless data communication and cryptographic techniques. It also deals with the classification of cryptographic techniques like symmetric key cryptography and asymmetric key cryptography. It explains the efficient Selective Encryption algorithm to minimize the processing time and improve the security against different attack. Energy optimization technique to reduce transmission power. Using the Modified Cuckoo Search Resource allocation (MCS-RA) approach, energy-saving gains of up to 10dB are obtained to the EEP method of network resource allocation and 1.5 dB more than the traditional resource allocation method UEP (GA).

Web-based talking alludes to the way toward sending and getting messages utilizing the web. There are different talking applications accessible on the lookout. In the principal quarter of 2017, the absolute number of clients utilizing visit applications are more than 5.03 Billion, and broadly utilized applications are[1], Facebook Messenger, QQ Mobile[2], WhatsApp[3], and so forth The proposed network-based android talk application utilized for visiting reason with distant customers or clients associated with the web, and it won't let the client send unseemly messages WeChat[4]. We will in general make BONJOUR. For the extent of the venture, the function will be tried as the program is being created. An information base for the clients enrolled will be created and tried, a menu will be created and tried, a customer/worker interface will be created and tried, and GUI's will be created and tried, for the clients' advantages. At the point when the talk application is close to finishing, all the more testing will be done so as to make it less of a hassle or easier to understand.

Social media is one of the most popular form of platform which is most common with people of our age. With passage of time, memes have gained a significant popularity and are often shared on social media platforms. Memes usually have hilarious content but can be offensive sometimes, containing some hateful message or character image. Such memes may have detrimental social impact in our society or to any individual. [1] Thus, an automated system for evaluation of offensiveness in the meme content is required. This paper presents an approach to detect offense in memes using Natural Language Processing (NLP) and deep learning. Due to the increasing number of memes over the internet, it is not an easy task to evaluate each meme before it spreads all around. This raises a demand for a system that can automate the process of evaluating memes before they agitate a crowd or spread a humour. This paper presents a model to detect offensive memes, in three steps. First, it will extract the text from the given image, then it will classify the given text as offensive or not offensive. If the text is found to be offensive then in the third step it will further classify offensive text in three categories namely slight offensive, very offensive and hateful offensive. The dataset used for this work consists of 6,992 memes which were labeled as not offensive, slightly offensive, very offensive, and hateful offensive. The model uses very simple architecture with a multi-layer dense network structure involving NLP with RNN and LSTM along with word embeddings such as GloVe and FastText.

Grid integration of PV panels is trending in developing countries in order to have a sustainable alternative for the power generation. An AC module employed to harness the solar energy comprises a PV module with a grid-coupled inverter. Among different arrangements of AC module, transformerless single stage arrangement is highly efficient. ZSI being one of such arrangements have seen a fast evolution since its first release in 2003, the evolution includes its different topologies as well as its control and modulation schemes to enhance its performance in different perspectives. This paper work includes a comparative analysis of different modulation schemes employed for 3-O ZSI in order to emphasize on the basic differences between them and finally conclude with a suitable technique to obtain the high output voltage along with low voltage stress on the switching devices of the inverter.

In order to obtain a sustainable alternative for the power generation in developing countries, grid coupled PV panels are the best solution which is widely preferred in developed countries. A PV module with a grid coupled inverter is embodied in an AC module. Transformer less single stage arrangements are highly efficient and preferred over different arrangements of an AC module. ZS I being one of such arrangements have seen a fast evolution since its first release in 2003, in which its different arrangements along with its control and modulation schemes are included. This paper work emphasizes on the performance of ZS I under three modulation techniques provided with constant input voltage and modulation index, and finally conclude with a suitable scheme to obtain high output voltage.

For any classification or prediction problem, machine learning algorithm plays an important role in analysis of data. It enhances the performance capability of classifiers and ease the difficulties. In this research analysis, a prediction model is developed for diabetes disease using K-nearest neighbors classification algorithm. The model is implemented in Python using PIMA indian diabetes dataset.The objective of research work is to enhance the prediction capability of KNN classifier with the help of feature selection and normalization of data. A study is also performed to find the optimal number of neighbors on which KNN returns its best result. The F1 score is considered as the main performance metrics for this work.

The relation between solar activity indices and earth atmosphere is studied in this current investigation using correlation as well as periodic analysis. The solar flare index and solar irradiance signal of India over different latitude values are considered in this study as solar activity indices and Earth atmospheric parameter respectively. To evaluate their relationship, last 33 years (1986 – 2018) of data is considered which covers solar cycle 22, 23 and almost the complete solar cycle 24. The solar irradiance of India over different considered geographical location has a positive correlation during solar cycle 23 and a negative correlation during solar cycle 24 with solar flare index. It has also been observed that the maximum correlation is exist between 5° - 10° latitude and 30° - 35° latitude data and the correlation has an increasing trend with higher latitude. The periodicity analysis of solar irradiance signal exhibits a similar type of inherent period with solar flare index as well as other activity indices. It can be suggested that the effect of the solar flare index is exists on the solar irradiance signal of India over different considered geographical location according to the current statistical investigation during solar cycle 22 – 24.

The data of soft x-ray solar flare activity for the interval 2008 to 2019 which covers the ascending, maximum, descending & part of minimum period of solar cycle 24 has been used for investigating the behavior of North-South (N-S) and East-West (E-W) asymmetries. Our statistical analysis revealed that the N-S asymmetry is a real phenomenon and that the solar cycle 24 shows southern predominance which is similar to the past solar cycle 21, 22 and 23. Hence a long-term periodic behavior could be inferred. The E-W asymmetry for the considered is found to be slight western dominated but the associated probability values indicates a statistically insignificant result.

Indian Railways being one of the largest functioning units of India employs a huge amount of manual unorganized labor. Maintenance of railway coaches and bogies is provided from part of this labor. It has always been a great challenge to find an optimal operating scenario for maximized productivity. In this work, we aimed to model and maximize maintenance productivity of Indian Railway coach and bogie . Taking into consideration the volatility of the circumstances and uncorrelated behavior of different human factors affecting productivity, to obtain the best scenario five of the most critical human factors were identified that are expected to affect the maintenance of the coaches and bogies most. Cobb-Douglas production function was used to model maintenance productivity with the chosen human factors. The main goal after designing the model for productivity is to maximize it. To get the optimal productivity Differential Evolution algorithm is used. Finally, the Cuckoo search algorithm is applied to validate the results.

Machine learning algorithm plays an important role for the data generation and analysis to ease the difficulties of life. Whereas the disease classification or prediction can be performed using machine learning algorithms. These learning algorithms are applied to enhance the capability of classifiers. In this research experiments, KNN and machine learning methods are used in the prediction model to classify whether the patient is diabetic or non-diabetic. The PIMA diabetes dataset is used for research purpose in the python implemented model. Research study has been performed to improve the performance of KNN classifier by using feature selection method, normalization and considering different number of neighbors. The performance of classifier is measured based on different metrics such as accuracy, precision, sensitivity, specificity, F1 score and error rate.

Luminescent Solar Concentrators (LSCs) offer an alternative way to generate electricity from solar radiation in urban environment. One of the main advantages of an LSC is its ability to utilize diffuse solar radiation. However, low efficiency of LSCs have inhibited their widespread commercial application. Plasmonic Luminescent concentrator (PLSC) was recently introduced to decrease the escape cone losses in an LSC. In this work, an optimized PLSC device of size 4.5 × 4.5 × 0.3 cm 3 is tested outdoors along with an LSC device of same dimensions. It is observed that the PLSC device performs 2.8 times better than a blank host matrix.

Single stage high boost converters offers an economical solution for highly intermittent and low power generating source. This paper investigates Quasi-Impedance Source Inverter (qZSI) for off-grid photovoltaic application, for an LCL connected qZSI the described control methodology uses current fed control at ac side, where outer loop is used to generate the reference, inner current and voltage loops are used to get the desired modulation index. Paper presents the detailed small signal model for transient response prediction, elaborates passive component size selection and controller parameter for an effective closed loop control. The described methodology is validated for normal and reduced voltage, controllers effectively tracked the reference values and able to reject the disturbance effectively.

In the current era of increasing applications of solar energy, it is most crucial to emphasize on true performance.Also world community is focusing on use of artificial intelligent methods for forecasting the performance of systems. In this paper an attempt is made to emphasize on application of soft computing for performance measurement and also true potential determination. For this, experiment is conducted in laboratory and in outdoor conditions in GLA University, Mathura, India. In outdoor conditions results are enthusiastic because Q indoor value of MSE, RMSE, MBE and SDE is 84%, 9.6%, 6.3 % and 6.65% while the Q outdoor value of MSE, RMSE, and MBE is and SDE is 29%, 5.38%, 5.25 % and 1.19%. Here Q is error between simulated and experimental energy.

Coherent detection schemes require full Channel State Information (CSI) at the receiver for Single Input Single Output (SIMO) system. It is difficult to have full CSI at the receiver end in time varying channel due to large transmit overheads and the need for continuous channel estimation at the receiver for each symbol and channel. In this paper, receiver assumes the perfect channel knowledge of only a preable of two symbols is available. The subsequent received symbols of a frame of size $N$ are detected using the known preamble. The available CSI for two symbols is called as Partial CSI. A simple channel prediction technique to detect data in a time varying channel given partial CSI is proposed for 1xL SIMO system. Use of partial CSI will reduce the complexity at receiver end. A flat fading channel is modelled by a second order autoregressive (AR2) process which has higher correlation to Jakes model. Exact closed form generalized expression of probability of error for M-PSK modulation under partial CSI for 1 × $L$ SIMO system using maximum ratio combining (MRC) scheme is derived. Generalized expression of lower bound of probability of error for $M$ -PSK modulation under partial CSI for 1 × $L$ SIMO system using selection combining (SC) scheme is also derived. Further, these expressions are reduced for some special cases and compared them with prevailing results in the literature. Simulation results are presented for different modulation schemes and time-varying channel conditions for MRC and SC using proposed scheme which are closely matching with derived analytical expressions. Simulation results shows negligible degradation in performance under medium scale fast fading channel and approx. 2 dB degradation at 10–3 order BER under large scale fast fading channel using partial CSI.

To cater to the exponential growth of terrestrial mobile services, several frequency bands have been included for use in International Mobile Telecommunications (IMT). Consequently, the 2500-2690 MHz frequency band is now being shared between Band 41 of 4G-LTE (Long Term Evolution) and regional Geo-Mobile Satellite Services (MSS) [1]. Within a country the inter-system interference is minimized through its national-level coordination, which ensures that the frequencies used for both systems are mutually exclusive. However, other countries (including neighbouring) can independently choose to use this band for either LTE or MSS, resulting in interference to systems of both countries. This paper proposes a novel method to estimate aggregate interference from LTE terrestrial services on S band Geo-MSS return (user to satellite) and forward (satellite to user) links. The results of interference analysis are presented and it is shown that MSS forward and return links of a particular country are severely affected by co-frequency LTE services of other countries.

Pulse Repetition Interval (PRI) timing and its modulation types are significant features of any Radar to comprehend its capabilities and functionalities. In this paper, an algorithm has been proposed to classify PRI modulation by estimating features from time of arrival (ToA) and its difference (dToA) of a pulse radar. Performance of the proposed algorithm is also verified for different PRI modulations with different percentage of missing pulses. PRI modulation type classified covers Dwell and Switch (DS) of 2 to 64 level, Staggered PRI of 2 to 64 levels, Jittered PRI with 20% Gaussian deviation and Sliding PRI. Estimation of features is improved by selection of an optimal window of histogram. Level of staggered/ DS PRI has also been estimated using Kernel Density Estimator (KDE) with an epanechnikov window with length $N^{0.8}$ where, N is total count of dToA. Upon classification, PRI estimation for each modulation type is also estimated.