http://hdl.handle.net/123456789/102 Wed, 05 Aug 2026 17:10:07 GMT 2026-08-05T17:10:07Z http://hdl.handle.net/123456789/21522 INTERFERENCE MANAGED HYBRID CLUSTERING TECHNIQUES TO IMPROVE 5G RADIO RESOURCE MANAGEMENT Hasan, Naureen Enroll # 02-284162-001 To meet the projected growing demand of future networks, 5G Heterogeneous Network (HetNet) is expected to provide seamless connectivity among the network nodes. In addition, it enable the base-stations and mobile users to maintain consistent data transmission rates in indoor and outdoor locations. To attain such internetworking environment, 5G networks have to guarantee high data throughput, enhanced network capacity, and scalability with reduced network interference and latency compared to previous cellular technologies. However, due to the scarce radio resource availability and conventional network deployment strategies, attainment of these parameters is difficult. Therefore, to achieve an efficient radio resource management mechanism in heterogeneous networks, a novel clustering architecture is proposed in this thesis. The clustering technique can be performed in various ways such as centralized, distributed, and hybrid clustering methods. Nevertheless, the issue of compromised throughput, limited capacity, and high interference persists. This research work contributes a two-fold strategy. Firstly, an efficient hybrid clustering algorithm is proposed, named as Interference-Managed Hybrid Clustering (IMHC) mechanism. The IMHC mechanism categorizes the small cell base stations (SBSs) as High-power SBSs (HSBSs) and Low-power SBSs (LSBSs) based on their transmitting power by introducing small-cell power control (SPC) algorithm. By analyzing the simulation results, it can be concluded that the IMHC mechanism under the orthogonal frequency division multiple access (OFDMA) for SBSs in a three-tier heterogeneous network addresses the issue of co-tier and cross-tier interference. Consequently, by reducing the interference it improves the network throughput and the Signal to Interference Ratio (SIR) at different tiers of HetNet. By managing the small cell nodes of similar power levels at the individual tiers i.e. Pico cells at tier -2, and Femto cells at tier-3, the co-tier and the cross-tier interference are reduced. However, even by implementing the IMHC mechanism still the outlier nodes experiences the issue of low signal reception. Therefore, to improve the signal reception at all users within the network the Power Domain - Non Orthogonal Multiple Access (PD-NOMA) scheme is considered Therefore, to further improve the Radio Resource Management (RRM) and achieve a robust user association emphasizing the edge users within the given mechanism an adaptive scheme is proposed. In this step, clustering with cooperative (PD - NOMA) is performed which resulted in an improved performance of a clustered heterogeneous network. Additionally, to improve user association and network clustered cooperative PD-NOMA performance the proposed technique employs algorithm that limits the number of users associated per cluster. The research has improved the user association, resulting in an increase in system capacity, system throughput, and sum-rate in an ultra-dense heterogeneous network with decreased interference and latency. Statistically, improved results are achieved with the proposed scheme in terms of throughput by 65% when compared with the clustered OFDMA scheme and 23% when compared with the unified PD-NOMA scheme. With a varying number of randomly placed users range from 2 — 80 users and the randomly deployed femto base stations up to 1000 and the number of pico base stations are considered to be 100. For the assumed scenario the system capacity has increased by 8% when compared with unified PD-NOMA and approximately 80% as compared to the clustered OFDMA scheme as previously studied. Thus, with this research an efficient radio resource management clustered topology is achieved along with an interference abating power controlling algorithm for the future networks Supervised by Dr. Safdar Ali Rizvi Mon, 01 Jan 2024 00:00:00 GMT http://hdl.handle.net/123456789/21522 2024-01-01T00:00:00Z http://hdl.handle.net/123456789/21523 A Hybrid Recommendation System Considering Visual-Textual Information Using Machine Learning Techniques Ikram, Fasiha Enroll # 02-284171-001 The increased variety of content in e-commerce and entertainment services creates a new research gap in the field of recommendation systems. Traditional recommendation systems are not capable of dealing with a variety of data because the nature of the data is unstructured and heterogeneous. In the past decade, due to the exponential growth of multimedia content in the video game industry, researchers investigated the importance of personalized video game recommendation techniques. It has been noticed that the previous methods have not investigated the importance of visual and textual semantic content for video game recommendations. The researchers have contributed to the field of multimedia item recommendations using low-level and aesthetic visual features. However, they suffer from less feature utilization and significant performance decay because they usually only consider traditional features such as product information and user information. One potential approach to alleviating such issues is to utilize item semantic information along with hybrid filtering techniques of recommendation systems. This study proposed a novel method named deep hybrid semantic multimedia recommendation systems (MMRSs) to deal with textual and visual semantic features for multimedia recommendation systems. The primary contribution of the presented work is to show improvement in the accuracy of the multimedia recommendations system by using both textual and visual semantic content. In the presented thesis, a semantic-based video game recommendation system utilizing deep learning methods a for visual and textual content learning and user profile learning has been proposed. The proposed approach employs hybrid techniques to expand users’ profiles. The user profile expansion is based on the semantic content of games’ visual and textual modalities. We have used two datasets of video games which are Goggle-Play and Amazon for evaluation purposes. The evaluation results have shown that the proposed pproach accuracy have been improved by up to 94% for Google-Play dataset and for the Amazon datasets we have received 78% as compared to the other state-of-the-art methods Supervised by Dr. Humera Farooq Mon, 01 Jan 2024 00:00:00 GMT http://hdl.handle.net/123456789/21523 2024-01-01T00:00:00Z http://hdl.handle.net/123456789/21520 Dynamic Sign Language Recognition using Deep Learning Javaid, Sameen Enroll # 02-284172-002 Sign Language-based communication is a major source of communication between people with hearing impairments and the general public. Hand, head, body, and gesture developments are immediate aspects of communicating specific feelings or techniques of correspondence in any language, whether marked or spoken. Hand sign recognition is known as manual sign language recognition in sign language, whereas facial expressions and body gesture understanding are known as non-manual sign language recognition. Several researchers are focusing either on manual gestures or non-manual gestures separately; a rare focus is on manual and non-manual gestures concurrently, making the loss of content or complete meaning of the sentence. Nonetheless, one of the great challenges of Sign Language Recognition is dealing with formal and non-formal parameters simultaneously within a single framework or methodology whereas using dynamic sign language recognition. However, dynamic sign language recognition has several difficulties recognizing complex features, accurate classifications, and extensive video sequence data training and testing in the Spatio-temporal domain. To issues, the current research study presents a Multimodal Dynamic Sign Language Recognition in the Spatio-temporal domain based on vision-based deep learning. This research consists of two-fold contributions. First, for the training and testing, there is a lack of such a dataset, where manual and non-manual modalities combine with affective facial expressions. So first contribution is compiling a Pakistan Sign Language (PSL) dataset with Manual and Non-Manual modalities named PkSLMNM. Further, we proposed a system called Sign Language Action Transformer Network (SLATN) is restricts hand, body, and facial signals in video arrangements. Here we are using a Transformer-style structural design as a "base network" to remove highlights from a spatiotemporal space. The model hastily figures out how to follow individual people and their setting of activity in different edges. Further, a "head network" at the same time classify and region out hand movement and facial expression, which is frequently critical to figuring out communication through signing. It uses its attention mechanism for creating tight bounding boxes around classified gestures. Later, the model's performance is evaluated against state-of-the-art datasets and conventional identification techniques. It not only completes tasks more efficiently but also with good accuracy. Our suggested network achieves 82.66% testing accuracy and 94.13 Giga FLOPs of a notable processing performance Supervised by Dr. Safdar Ali Rizvi Sun, 01 Jan 2023 00:00:00 GMT http://hdl.handle.net/123456789/21520 2023-01-01T00:00:00Z http://hdl.handle.net/123456789/21521 VISION BASED GAIT RECOGNITION ROBUST TO VIEW AND APPEARANCE VARIANCE Masood, Hajra Enroll # 02-284151-002 Vision-based gait recognition has excellent potential for biometric identification due to its non-intrusive, non-invasive and remote access person data collection. Gait recognition has diverse applications for visual surveillance due to its adaptability for person identification and making predictions about age, gender, and ethnic background. The vision-based gait recognition is adaptable on low-resolution video as it only lequires the visibility of the human body for feature extraction. The vision-based gait recognition-based person identification is highly affected by the factors altering the perceivable shape of the human body, including variance in the subject’s appearance and viewing angle. These factors reduce the adaptability variance of conventional gait feature extraction techniques, including Gait Energy Image and Gait Silhouette. The problem of gait recognition robust to appearance variance is twofold complex as it introduces higher intra class and lower inter-class variance. These two problems require developing gait features that strongly correlate within the same class and discriminant enough for multi-class classification. This thesis proposes gait feature extraction technique named “Dynamic Gait Feature”, by estimating the relative motion between key poses of the gait cycle and encoding it as feature vectors. The Dynamic Gait Features are evaluated on dual criteria of the problem and are established to be strongly correlated within class and adaptable with Support Vector Machine classifier-based gait recognition. These Dynamic Gait Features are further transformed into Spatio Temporal Power Spectral (STPS) gait features. The robustness of STPS gait features towards different, appearances and views is established by adapting machine learning classifiers. The Dynamic Gait Feature based gait recognition has achieved 97.53% accuracy despite significant appearance variance. The STPS based gait recognition has achieved 99.87% accuracy despite view and appearance variance. This thesis also addressed the effects of south Asian clothing on the subject’s appearance. A local dataset has been developed to address the effects of south Asian clothing on the subject’s appearance. We expect this research to set a new dimension for the adaptation of vision-based gait recognition for automated visual surveillance that is adaptable in a real time environment. Supervised by Dr. Humera Farooq Sun, 01 Jan 2023 00:00:00 GMT http://hdl.handle.net/123456789/21521 2023-01-01T00:00:00Z