TIET Digital Repository

Recent Submissions

• Loading...

  Item type:Item,

  Performance Improvement of Generalized Frequency Division Multiplexing based 5G Communication System

  (2026-05-21) Kaur, Manpreet; Joshi, Hem Dutt

  Show more

  The current fifth generation (5G) technology has brought a paradigm shift in wireless communication systems, offering unprecedented speeds, ultra-low latency, massive device connectivity, and remarkable network capacity. While 5G is still in its early stages of deployment and expansion, researchers and industry experts are already contemplating the requirements, challenges, and opportunities for ”beyond 5G” (B5G) communication. The development of B5G is gaining attention as the global community seeks to redefine wireless communication systems for the future. The diverse requirements of the B5G standards include data rates of 1 Gb/s, device density of up to 107 devices/km2, high mobility of up to 1000 km/h, and latency in the range of 10−100 μs. To cater for the evolving requirements of future communication systems many current and future standards have adopted multicarrier techniques which include both orthogonal (i.e., Orthogonal Frequency Division Multiplexing (OFDM) and Universal Filtered Multicarrier (UFMC) and non-orthogonal (i.e., Generalized Frequency Division Multiplexing (GFDM), Orthogonal Time Frequency Space (OTFS), and Filter Bank Multicarrier (FBMC) techniques. GFDM has gained attention as a non-orthogonal waveform with several advantageous features, making it a viable choice for future generations (like B5G, 6G). These advantages include its flexible structure, resilience against frequency selective fading, minimal OOB emissions, high spectral efficiency attributed to reduced CP overhead, and low PAPR. This dissertation aims to enhance the performance of GFDM-based 5G communication systems. It includes a performance analysis of the GFDM system over mmWave channels, particularly the FTR channel. The impact of channel impairments, such as the unavailability of perfect channel state information (CSI), on the performance of GFDM systems is also presented. The second contribution of this thesis is the introduction of a novel timing synchronization algorithm, which significantly improves the performance of GFDM systems. This improvement leads to increased reliability in data transmission, a crucial aspect of the B5G system. The third contribution involves a novel filter design method based on the discrete biorthogonality condition and Wigner Distribution (WD). The proposed filter enhances the GFDM system’s performance in terms of both Symbol Error Rate (SER) and Power Spectral Density (PSD). The analytical results derived from the proposed expressions are verified through Monte Carlo simulations. In some cases, numerical integration is required over finite limits, which can be easily implemented with negligible error using tools such as MATLAB and Mathematica.

  Show more
• Loading...

  Item type:Item,

  Growth and characterization of Cu2BaSn(S1-xSex)4 thin films for solar cell application

  (2026-05-13) Jyoti; Mohanty, Bhaskar Chandra

  Show more

  In recent years, Cu2ZnSnS4 (CZTS) has been focus of study as a promising absorber layer in thin film solar cells due to its high absorption coefficient, appropriate bandgap, and environment-friendly earth-abundant constituents. Despite global efforts the efficiency of the CZTS based devices is stagnated at 12.6% since 2014. This stagnated performance is attributed to a large Voc deficit caused by recombinations due to charged point defects and defect clusters. The dominant defect in CZTS is CuZn antisite defect, formed due to similar ionic radii of Cu and Zn (0.74 Å) ions. A potential approach to curb the formation of these antisite defects is isoelectronic substitution of one of the cation by cation of larger or smaller ionic radii. Among various substitutions proposed in literature (e.g. Ag for Cu, Cd for Zn, Ge for Sn, etc.), substitution of Zn by Ba is considered to significantly change the cationic disorder in CZTS. In Cu2BaSnS4 (CBTS), Ba has a larger ionic radii (1.56 Å) and the dominant defect in CBTS is VCu which results in p-type conductivity of CBTS, similar to the more matured absorber material CuInGaS2. Owing to the large size difference, significant structural changes in CBTS, and hence, opto-electronic properties compared to CZTS are expected. This work deals with the growth and characterization of CBTS and Se-alloyed CBTS (CBTSSe) thin films. In view of the differences in the optical and electrical properties of CBTS films with those of CZTS, we have numerically investigated and compared the performance of both devices using the SCAPS software. Simulations were carried out by considering the typical solar cell structure of glass/Mo/CZTS/CdS/i-ZnO/ITO. For a more realistic approach, a thin MoS2 layer is considered between Mo and CZTS. Simulations revealed an efficiency of about 17.68%, which is much higher than the experimentally obtained record efficiency of 11%. This suggest that the simulation should include an appropriate amount of defects in the bulk and at interfaces (i.e., back interface MoS2/CZTS and front interface CZTS/CdS). The experimental champion device parameters could be successfully simulated only when bulk defect density of 5.5 × 1015 cm-3, defect density of ~1×1015 cm-2 and ~1 × 1014 cm-2 at back and front interfaces was introduced. A possible route – by inserting a back surface field (BSF) layer - to improve the efficiency of the devices with CBTS films having these amounts of defect density has been demonstrated. It is shown that the CZTS solar cell efficiency can be increased up to 14.7% and 15.7% by optimizing Cu2O and SnS films as BSF layers, respectively. On the other hand, for CBTS films with similar defect density that resulted 11% efficiency for CZTS (experimentally obtained champion cell) simulations yielded an efficiency of only 4.55%. This is because of larger bandgap (2.0 eV for CBTS vs 1.5 eV for CZTS) and different nature of defects. Performance of the CBTS devices could be increased to 6.9% (reported experimental value) only when the defect densities were considerably reduced (interface defect density NMoS2/CBTS ~ 1015 cm-2, NCdS/CBTS ~ 1010 cm-2 and bulk density NCBTS ~ 1014 cm-3). The results suggest that the performance improvement of CBTS solar cells is more challenging than that for CZTS cells and hence, experimental conditions for the fabrication of CBTS films are expected to be more stringent. The CBTS films were synthesized by a solution based approach. A precursor film was prepared by spin coating of a non-toxic 2-methoxy ethanol based molecular precursor solution and was heat treated in presence of sulphur powder to obtain the eventual film. Since the formation of the secondary phases must be suppressed during the growth of the films as they degrade the performance of the solar cells, the process parameters were carefully optimized and the reaction pathway leading to the formation of single phase CBTS was established. We have systematically varied the molar concentration ratio in the solution and the sulfurization parameters (temperature, dwelling time and sulphur amount) and studied the impact thereof on the evolution of single phase CBTS. It was found that ideal molar concentration ratio [Ba]/[Sn] =1.0 always yielded secondary phases in spite of a large variation in the sulfurization parameters. Single phase CBTS thin films are obtained only for [Ba]/[Sn] = 1.4 in the precursor solution and sulfurization at 575 °C for 45 min with 1.0 g of powder S. UV–visible and room temperature PL measurements revealed a band gap of ~2.0 eV for these films. A symmetric PL peaks suggests reduced cationic disorder in the films compared to CZTS. The films showed white light sensitivity (~30%) for illumination of 24 mW/cm2. Detailed electrical and electro-impedance measurements showed p-type conductivity with a carrier concentration of 1.7×1014 cm-3 for the films. The CBTSSe films were obtained by heat treating the as-prepared precursor films in the presence of 1.0 g of sulphur (that yielded CBTS films) and varying amounts of selenium. It was found that the process parameters that produced CBTS films yielded various secondary phases that necessitated further optimization of the parameters including Ba/Sn ratio in the precursor solution, sulpho-selenization temperature and dwelling time, etc. Phase pure CBTSSe thin films were obtained for Ba/Sn=1.7 and annealing at 550 °C for 45 min with 1.0 g of S and 0.1 g of Se. By placing varying amounts of Se in the furnace during the sulpho-selenization process step, the concentration of Se in the films was systematically varied and the impact thereof was investigated. It was observed that by varying the Se amount from 0.1 to 0.4 g during sulpho-selenization, the Se/(Se+S) ratio in the resulted films increased from 0.05 to 0.22. With increase in the Se amount in the films, the bandgap of the films decreased gradually from 1.93 to 1.55 eV. The electro-impedance spectroscopy measurements on the film grown with 0.4 g of Se during sulpho-selenization revealed its p-type conductivity with an acceptor concentration of 1.58 × 1017 cm-3. The results indicate that these films can be potentially used as photocathode for hydrogen evolution.

  Show more
• Loading...

  Item type:Item,

  Human Resource Climate and Faculty Retention: Evidences from Institutions of Higher Education in Northern India

  (2026-05-13) Verma, Sahil; Kaur, Gurvinder

  Show more

  This study investigates the role of Human Resource (HR) Climate in shaping faculty retention within higher educational institutions (Central, State, Private and Deemed to be Universities) across Northern India. Data was collected from 770 faculty members across Punjab, Haryana, Delhi, and Chandigarh, selected from the top 100 universities in these states as per NIRF rankings, with final participation based on institutional and individual consent., the research aims to (1) examine and compare the HR Climate of different types of higher education institutions, (2) assess the influence of HR Climate and Individual Factors on faculty retention, and (3) explore the mediating roles of Organizational Trust, Organizational Commitment, and Job Satisfaction in this relationship. The study adopts a two-phase quantitative methodology. Exploratory Factor Analysis (EFA) on an initial sample of 300 faculty members identified five key sub-dimensions of HR Climate: Senior Support, Peer Support, Research Environment, Rewards, and Task Environment. Subsequently, Partial Least Squares Structural Equation Modeling (PLS-SEM) was employed on a sample of 470 to validate the measurement and structural models and test hypothesized relationships among variables. Additionally, the entire sample was used for comparing HR Climate perceptions across institution types using ANOVA. A self-designed questionnaire was used to measure HR Climate, while standardized and validated instruments from existing literature were adopted to assess Individual Factors, Organizational Trust, Commitment, Job Satisfaction, and Faculty Retention. Findings reveal that faculty in central and state universities perceive a significantly more positive HR Climate than their counterparts in private institutions. However, no significant differences emerged between central, state, and deemed-to-be universities in overall HR Climate perception, fulfilling the first objective. Structural model results confirmed that HR Climate has a strong and positive impact on faculty retention, job satisfaction, organizational commitment, and trust. Interestingly, Individual Factors—such as work life balance and availability of alternate job opportunities—do not directly influence faculty retention, showing a non-significant path coefficient. However, they significantly contribute to job satisfaction, commitment, and trust, which in turn strongly influence retention, thereby fulfilling the second objective. 6 Mediation analysis provided further insights. The relationship between HR Climate and Faculty Retention is partially mediated by job satisfaction, commitment, and trust—indicating complementary partial mediation. In contrast, the relationship between Individual Factors and Faculty Retention is fully mediated through these three variables, highlighting their critical indirect role in retention decisions. These results underscore the importance of positive psychological states in translating individual and organizational dynamics into long-term faculty engagement, achieving the third objective. This research significantly advances the understanding of faculty retention through the lens of Organizational Support Theory (OST) by positioning HR Climate as a central contextual variable that shapes perceptions of organizational support. The study offers practical implications for educational administrators, particularly in private universities, where faculty perceive HR practices as comparatively less favourable. Institutions aiming to enhance retention must focus on building a supportive HR Climate, fostering research culture, developing transparent reward systems, and strengthening faculty experiences through job satisfaction, trust, and commitment. The establishment of dedicated HR departments in private universities is recommended to institutionalize such practices, which are essential for sustaining talent, improving academic outcomes, and achieving long-term institutional success.

  Show more
• Loading...

  Item type:Item,

  Some Best Proximity Point Problems and their Applications

  (2026-05-12) Sharma, Shagun; Chandok, Sumit

  Show more

  Approximation theory is a subject with a long history and a huge importance in classical and contemporary research. Over the years, the theory has become so extensive that it intersects with every other branch of analysis. One of the problems in approximation theory is to detect a point that minimizes the distance between two subsets $\mathcal{E}_{1},\mathcal{E}_{2}$ of a metric space $(\mathcal{W},d)$. Once this is done, it motivates us to study the solution of minimization problem. For example, \begin{align*} \min_{\acute{k} \in \mathcal{E}_{1}}d(\acute{k},\mathcal{B}\acute{k}), ~\min_{\acute{m} \in \mathcal{E}_{2}}d(\acute{m},\mathcal{B}\acute{m}) ~\text{and}~~ \min_{(\acute{k},\acute{m})\in \mathcal{E}_{1}\times \mathcal{E}_{2}}d(\acute{k},\acute{m}), \end{align*} where $ \mathcal{B}$ is a mapping on $\mathcal{E}_{1} \cup \mathcal{E}_{2}$, such that $\mathcal{B}(\mathcal{E}_{1}) \subseteq \mathcal{E}_{1}~ \text{and} ~ \mathcal{B}(\mathcal{E}_{2}) \subseteq \mathcal{E}_{1}$. It is fascinating to arise a question whether is possible to find a pair $(\acute{k},\acute{m}) \in \mathcal{E}_{1} \times \mathcal{E}_{2}$ which is a solution of above problem that is, to find a pair $(\acute{k},\acute{m}) \in \mathcal{E}_{1} \times \mathcal{E}_{2}$ such that $\acute{k}=\mathcal{B}\acute{k}, \acute{m}=\mathcal{B}\acute{m} ~\text{and}~~ d(\acute{k},\acute{m}) = d(\mathcal{E}_{1},\mathcal{E}_{2})$. If such a pair exists, it is called the best proximity pair for a mapping $\mathcal{B}$. If we take $\mathcal{B}$ a non-self mapping we find an approximate solution $\acute{k}$ such that the error $d(\acute{k}, \mathcal{B}\acute{k})$ is minimum. The existence of an approximate solution $\acute{k}$, called best proximity point, that is, to find $\acute{k} \in \mathcal{E}_{1}$ such that \begin{align*} d(\acute{k}, \mathcal{B}\acute{k}) = d(\mathcal{E}_{1},\mathcal{E}_{2}) = \inf\left\lbrace d(\acute{k}, \acute{m}) : \acute{k} \in \mathcal{E}_{1}, \acute{m} \in \mathcal{E}_{2}\right\rbrace. \end{align*} Approximation theory can be used to solve many kinds of problem such as systems of nonlinear matrices, integral and differential equations, fractals, split feasibility problems, and variational inequalities. The study of approximation theory is appropriately inspired by the fact that particular instances of approximation frequently arise from problems connected with science and technology. The first aim of this project is to construct algorithms for the existence and uniqueness of a best proximity point. Another aim of this project is to discuss some applications of best proximity points. In the first chapter, we provide the supplementary material such as some definitions, preliminary results that are useful for upcoming chapters. It also includes the literature survey, thesis goals, as well as a synopsis of the information included in each of the thesis's chapters. In the second chapter, we discuss the existence of best proximity points for non-self mappings satisfying some contractive conditions in the setting of metric spaces, relational metric spaces, quasi partial metric spaces, normed and binormed linear spaces. Furthermore, we discuss the existence of best proximity pair results using noncyclic contraction mapping at the end of this chapter. Third chapter concerns with iterative schemes. In this chapter, we propose some algorithms that converge to a best proximity point and fixed point. Also, we introduce some algorithms which converge to solution of split fixed and split best proximity point problem. The fourth chapter deals with applications of best proximity point problems. In this chapter, we present the solution for variational inequality problems in the framework of Hilbert spaces. We provide a solution for a system of differential equations in the context of metric spaces. We also solve the model that spreads a virus using a non-linear integral equation.

  Show more
• Loading...

  Item type:Item,

  Interpreting decisions of AI models using XAI techniques

  (2026-05-04) Garg, Priya; Sharma, Mahesh Kumar; Kumar, Parteek

  Show more

  This thesis explores the intersection of Artificial Intelligence, interpretability, and application-specific challenges across diverse domains, including hate speech detection, AI text classification, medical image analysis, and feature interpretability. The research underscores the critical need for robust, transparent, and high-performing AI models to address pressing challenges in these fields. In hate speech detection, this study evaluates various word embedding techniques—CountVectorizer, GloVe, and Bidirectional Encoder Representations from Transformers (BERT)—combined with machine learning and deep learning classifiers. The BERT-BiGRU model achieves a notable accuracy of 92%, with interpretability enhanced through eXplainable AI (XAI) techniques such as Local Interpretable Model-Agnostic Explanations (LIME) and SHapley Additive exPlainations (SHAP), providing actionable insights into model predictions. Similarly, the growing prevalence of AI-generated content motivated an investigation into distinguishing human-authored text from AI-modified text. A comparative analysis highlights the effectiveness of word embeddings and text-level feature extraction techniques when integrated with machine learning classifiers, achieving high accuracy and F1-scores exceeding 91%. Transparency and interpretability were further improved using SHAP and LIME explanations. In medical image analysis, the integration of XAI techniques such as LIME and Gradient-weighted Class Activation Mapping (Grad-CAM) with deep learning models enhances model transparency while maintaining clinical reliability. Performance metrics across tasks emphasize the importance of explainable models in healthcare, strengthening trust and applicability. To address the limitations of traditional interpretability methods, this thesis introduces two novel enhancements to LIME. First, Radial Basis Function Based LIME (RBF-LIME) incorporates RBF interpolation to redefine local boundary assumptions as nonlinear relationships. Applied to stroke prediction using healthcare datasets, RBF-LIME demonstrates superior interpretability and effectiveness, outperforming conventional methods. Second, l_2 + l_1$-LIME integrates Ridge and Lasso regression to improve feature selection, enhancing LIME’s stability and interpretability. This method reduces sensitivity to perturbations, particularly in categorical datasets, making it more reliable for high-stakes applications such as healthcare. Evaluations across multiple datasets confirm improved consistency in feature attributions, contributing to more transparent AI decision-making. By integrating XAI techniques with cutting-edge AI architectures, this thesis advances the development of transparent, reliable, and high-performing AI systems across domains. The proposed innovations pave the way for future research into generalizability, computational efficiency, and broader applicability in real-world scenarios.

  Show more