Badugu Samatha
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Badugu Samatha has a total of 25 co-authors. Here are the top co-authors by number of shared publications:
- Nagarjuna Karyemsetty
Koneru Lakshmaiah Education Foundation, Vaddeswaram, India
3 shared publications - Thalakola Syamsundararao
Kallam Haranadhareddy Institute of Technology(A),Dept. of CSE,Chowdavaram,India
3 shared publications - Krishna Rupendra Singh
Department of Computer Science and Engineering, Vignan’s Institute of Engineering for Women, Visakhapatnam, Andhra Pradesh, India
2 shared publications - Mahanty Rashmita
Department of Basic Sciences and Humanities, Vignan’s Institute of Engineering for Women, Visakhapatnam, Andhra Pradesh, India
2 shared publications - Abhinav Raj1 shared publication
- B. Raja Koti
Department of CSE, GITAM School of technology, GITAM University, Visakhapatnam, Andhra Pradesh, India
1 shared publication
Latest Publications
- 29 December 2025
- Published by Edtech Publishers (OPC) Private Limited in International Journal of Scientific Research in Engineering and Management
- Vol. 09 (12) , 1-9
Cyber Security Awareness Platform
While significant investments have been made to support new cyber security technologies, there is still a reliance on humans to withstand the actions of attackers or other malicious actors. Cybersecurity is fundamentally about an organization’s ability to protect its assets from those who wish to do harm; therefore, cyber security has changed from simply being about technology to having a much broader mission — that of developing human resources and enabling employees to be more effective protectors of the organization's assets. This paper presents the Cognitive Resilience Framework for Cyber Security (CRFCS) as a new model of an organization’s ability to protect against cyberattack, focusing on how to create psychological capital for the personnel within the organization to ensure their psychosocial resilience to cyberattack. Human security capabilities should be viewed as renewable resource based on a combination of cognitive psychology, organizational behavior theory and security science. Thus, it is imperative to reframe the concept of organization security from a compliance-driven training approach to one that facilitates building human security abilities as part of the overall organizational capabilities of building human assets. This work provides the methodology to develop, measure and leverage human security as strategic organizations capability. Keywords continue to include Psychological capital, cyber security from a human-centric approach, security self-efficacy, threat vigilance, cognitive security architecture. Keywords—component, formatting, style, styling, insert
- 29 December 2025
- Published by Edtech Publishers (OPC) Private Limited in International Journal of Scientific Research in Engineering and Management
- Vol. 09 (12) , 1-9
Cyber Security Awareness Platform
While significant investments have been made to support new cyber security technologies, there is still a reliance on humans to withstand the actions of attackers or other malicious actors. Cybersecurity is fundamentally about an organization’s ability to protect its assets from those who wish to do harm; therefore, cyber security has changed from simply being about technology to having a much broader mission — that of developing human resources and enabling employees to be more effective protectors of the organization's assets. This paper presents the Cognitive Resilience Framework for Cyber Security (CRFCS) as a new model of an organization’s ability to protect against cyberattack, focusing on how to create psychological capital for the personnel within the organization to ensure their psychosocial resilience to cyberattack. Human security capabilities should be viewed as renewable resource based on a combination of cognitive psychology, organizational behavior theory and security science. Thus, it is imperative to reframe the concept of organization security from a compliance-driven training approach to one that facilitates building human security abilities as part of the overall organizational capabilities of building human assets. This work provides the methodology to develop, measure and leverage human security as strategic organizations capability. Keywords continue to include Psychological capital, cyber security from a human-centric approach, security self-efficacy, threat vigilance, cognitive security architecture. Keywords—component, formatting, style, styling, insert
- Open Access
- 30 December 2024
- Published by International Academic Publishing House in INTERNATIONAL JOURNAL OF EXPERIMENTAL RESEARCH AND REVIEW
- Vol. 46, 326-341
Proximal Policy Optimization for Efficient Channel Allocation with Quality of Service (QoS) in Cognitive Radio Networks
A multi-variable relationship exists in Cognitive Radio Networks (CRNs) where factors such as Energy efficiency, Throughput, Delay and Signal Noise Ratio (SINR) are related. The SINR shows the quality of the signal and is defined as the total power of a specific signal over the total power of an inter signal plus noise. This work proposes an effective energy and delay-efficient channel allocation strategy for CRNs (Cognitive Radio Networks) using Q-Learning and actor-criticism algorithms that maximize rewards. We also propose a Proximal Policy Optimization (PPO) algorithm that uses clipping of surrogate objectives to prevent large policy changes and ensure that the other parameters remain stable over time. We study the tradeoff between rewards, energy efficiency and other parameters and compare the algorithms with respect to the same. Results show that the proposed PPO method, while using optimally increased energy consumption, significantly reduces the delay, improves the thought and reduces the packet loss ratio for efficient channel allocation. This is positive with our findings shown in the results section and by comparing the proposed method with other algorithms to identify improved throughput and channel utilization. As the simulation results indicate that the PPO algorithm has very high throughput and significantly minimizes the delay and packet loss, it is suitable for application in all sorts of services such as video, imaging or M2M. The results are also compared with two of the existing channel allocation schemes and they confirm that the proposed algorithm performs better in terms of throughput discussed in one scheme and channel efficiency in the other.
- 30 December 2024
- Published by International Academic Publishing House in INTERNATIONAL JOURNAL OF EXPERIMENTAL RESEARCH AND REVIEW
- Vol. 46, 326-341
Open AccessProximal Policy Optimization for Efficient Channel Allocation with Quality of Service (QoS) in Cognitive Radio Networks
A multi-variable relationship exists in Cognitive Radio Networks (CRNs) where factors such as Energy efficiency, Throughput, Delay and Signal Noise Ratio (SINR) are related. The SINR shows the quality of the signal and is defined as the total power of a specific signal over the total power of an inter signal plus noise. This work proposes an effective energy and delay-efficient channel allocation strategy for CRNs (Cognitive Radio Networks) using Q-Learning and actor-criticism algorithms that maximize rewards. We also propose a Proximal Policy Optimization (PPO) algorithm that uses clipping of surrogate objectives to prevent large policy changes and ensure that the other parameters remain stable over time. We study the tradeoff between rewards, energy efficiency and other parameters and compare the algorithms with respect to the same. Results show that the proposed PPO method, while using optimally increased energy consumption, significantly reduces the delay, improves the thought and reduces the packet loss ratio for efficient channel allocation. This is positive with our findings shown in the results section and by comparing the proposed method with other algorithms to identify improved throughput and channel utilization. As the simulation results indicate that the PPO algorithm has very high throughput and significantly minimizes the delay and packet loss, it is suitable for application in all sorts of services such as video, imaging or M2M. The results are also compared with two of the existing channel allocation schemes and they confirm that the proposed algorithm performs better in terms of throughput discussed in one scheme and channel efficiency in the other.
- Open Access
- 30 December 2024
- Published by International Academic Publishing House in INTERNATIONAL JOURNAL OF EXPERIMENTAL RESEARCH AND REVIEW
- Vol. 46, 297-304
Image Captioning with Convolutional Neural Networks and Autoencoder-Transformer Model
This study deals with emerging machine learning technologies, deep learning, and Transformers with autoencode-decode mechanisms for image captioning. This study is important to provide in-depth and detailed information about methodologies, algorithms and procedures involved in the task of captioning images. In this study, exploration and implementation of the most efficient technologies to produce relevant captions is done. This research aims to achieve a detailed understanding of image captioning using Transformers and convolutional neural networks, which can be achieved using various available algorithms. Methods and utilities used in this study are some of the predefined CNN models, COCO dataset, Transformers (enc-BERT,dec-GPT) and machine learning algorithms which are used for visualization and analysis in the area of model’s performance which would help to contribute to advancements in accuracy and effectiveness of image captioning models and technologies. The evaluation and comparison of metrics that are applied to the generated captions state the model's performance.
- 30 December 2024
- Published by International Academic Publishing House in INTERNATIONAL JOURNAL OF EXPERIMENTAL RESEARCH AND REVIEW
- Vol. 46, 297-304
Open AccessImage Captioning with Convolutional Neural Networks and Autoencoder-Transformer Model
This study deals with emerging machine learning technologies, deep learning, and Transformers with autoencode-decode mechanisms for image captioning. This study is important to provide in-depth and detailed information about methodologies, algorithms and procedures involved in the task of captioning images. In this study, exploration and implementation of the most efficient technologies to produce relevant captions is done. This research aims to achieve a detailed understanding of image captioning using Transformers and convolutional neural networks, which can be achieved using various available algorithms. Methods and utilities used in this study are some of the predefined CNN models, COCO dataset, Transformers (enc-BERT,dec-GPT) and machine learning algorithms which are used for visualization and analysis in the area of model’s performance which would help to contribute to advancements in accuracy and effectiveness of image captioning models and technologies. The evaluation and comparison of metrics that are applied to the generated captions state the model's performance.
- Open Access
- 30 November 2024
- Published by International Academic Publishing House in INTERNATIONAL JOURNAL OF EXPERIMENTAL RESEARCH AND REVIEW
- Vol. 45 (Spl Vol) , 173-185
Advanced Dermatology Platform: Deep Learning with VGG19 and DenseNet201, Integrated Chatbot and Community Forum
The present online application employs a contemporary artificial intelligence (AI)-driven solution to transform the process of diagnosing skin disorders. This research uses DenseNet201 and VGG19, two of the most advanced DNN architectures, to build a Convolutional Neural Network (CNN). The enhanced predictive models, built with a dataset of 930 photos divided into ten groups and strengthened by data augmentation, produce remarkably accurate predictions for a range of skin conditions. The website's intelligent chatbot is a standout feature; it was built to answer questions about skin diagnoses, treatment options, and more. This chatbot is designed to help users understand their diagnostic results and find their way on the health journey. In addition, it keeps track of users' prediction histories, so they may learn a lot about their skin's health over time and make educated choices about their medical treatments. In addition, by giving people a place to talk about their struggles and get advice from others, the website fosters a supportive community. The emphasis here is on real human connections, which are great for learning from one another and helping one another out. Firebase facilitates efficient data administration for monitoring forecasts and engaging with the community, while Replit and Voice flow support the CNN model, chatbot, and forum, guaranteeing optimal performance. By integrating cutting-edge AI with a user-centric approach, this web application empowers users with the tools, insights, and support necessary for proactive skin health management.
- 30 November 2024
- Published by International Academic Publishing House in INTERNATIONAL JOURNAL OF EXPERIMENTAL RESEARCH AND REVIEW
- Vol. 45 (Spl Vol) , 173-185
Open AccessAdvanced Dermatology Platform: Deep Learning with VGG19 and DenseNet201, Integrated Chatbot and Community Forum
The present online application employs a contemporary artificial intelligence (AI)-driven solution to transform the process of diagnosing skin disorders. This research uses DenseNet201 and VGG19, two of the most advanced DNN architectures, to build a Convolutional Neural Network (CNN). The enhanced predictive models, built with a dataset of 930 photos divided into ten groups and strengthened by data augmentation, produce remarkably accurate predictions for a range of skin conditions. The website's intelligent chatbot is a standout feature; it was built to answer questions about skin diagnoses, treatment options, and more. This chatbot is designed to help users understand their diagnostic results and find their way on the health journey. In addition, it keeps track of users' prediction histories, so they may learn a lot about their skin's health over time and make educated choices about their medical treatments. In addition, by giving people a place to talk about their struggles and get advice from others, the website fosters a supportive community. The emphasis here is on real human connections, which are great for learning from one another and helping one another out. Firebase facilitates efficient data administration for monitoring forecasts and engaging with the community, while Replit and Voice flow support the CNN model, chatbot, and forum, guaranteeing optimal performance. By integrating cutting-edge AI with a user-centric approach, this web application empowers users with the tools, insights, and support necessary for proactive skin health management.
- 13 October 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 542, 102-107
Integrated Cloud Security for Data Storage and Access
During the COVID-19 pandemic, cloud computing & storage emerged as a viable alternative to traditional storage systems. Almost all firms are transitioning from traditional modes of operation to online modes of operation. The vast majority of businesses have the long-term goal of moving all or a significant part of their operations to an online (Work-from-Home) format. There are various advantages to cloud computing and storage, including cost savings, mobility, platform independence, accessibility, and elasticity. However, security is the primary impediment to adopting cloud storage services, particularly cloud retrieving. To address the issue of cloud storage security, this research study has proposed a method for cloud storage data security. The suggested strategy makes use of integrated cryptographic techniques to monitor, secure user as well as data that is kept in the cloud storage systems. The method that has been suggested constantly checks user data by performing the sequence of securities. The performance of proposed security is measured by downloading and uploading data iteratively and analyzed by computing time taken for cloud operations. Thus, the proposed strategy not only secures data but also increases user confidence in cloud-based storage services. The proposed approach can be further researched for dynamic data and variable data sizes.
- 13 October 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- Vol. 542, 102-107
Integrated Cloud Security for Data Storage and Access
During the COVID-19 pandemic, cloud computing & storage emerged as a viable alternative to traditional storage systems. Almost all firms are transitioning from traditional modes of operation to online modes of operation. The vast majority of businesses have the long-term goal of moving all or a significant part of their operations to an online (Work-from-Home) format. There are various advantages to cloud computing and storage, including cost savings, mobility, platform independence, accessibility, and elasticity. However, security is the primary impediment to adopting cloud storage services, particularly cloud retrieving. To address the issue of cloud storage security, this research study has proposed a method for cloud storage data security. The suggested strategy makes use of integrated cryptographic techniques to monitor, secure user as well as data that is kept in the cloud storage systems. The method that has been suggested constantly checks user data by performing the sequence of securities. The performance of proposed security is measured by downloading and uploading data iteratively and analyzed by computing time taken for cloud operations. Thus, the proposed strategy not only secures data but also increases user confidence in cloud-based storage services. The proposed approach can be further researched for dynamic data and variable data sizes.
- 27 August 2022
- Published by Conscientia Beam in Review of Computer Engineering Research
- Vol. 9 (3) , 200-208
A Model for the Safety Risk Evaluation of Connected Car Network
The automobile industry is preparing for what is now anticipated to be a big transition toward linked and autonomous vehicles. This introduces significant dangers both to individuals' privacy and to the intricate workings of electrical and electronic systems. Currently, the automotive industry has well-established and government-mandated safety risk management methods. They present an improved risk assessment methodology and demonstrate its validity with a real-world use case to facilitate the derivation of safety procedures and safety solutions for automotive integrated devices. This should make it simpler to ensure the safety of integrated vehicles. Some of the significant factors are the accessibility of functional safety to individuals who are not trained in safety and the appropriateness of existing techniques for said functional safety. The methodology that is used to assess the risks posed by properties and capacities includes a risk analysis as an essential component. The next step in the process of managing risks involves determining both the degree of risk and the magnitude of the potential impact. In this step, various different rules and requirements that are already in place are utilized to make any necessary adjustments. After much deliberation, a level of security has been settled on in order to facilitate the formulation of high-level security objectives. Due to the fact that this design has a good convergence with contemporary principles and criteria, it ought to be capable of being adequately suited to the prerequisites of the automotive sector.
- 27 August 2022
- Published by Conscientia Beam in Review of Computer Engineering Research
- Vol. 9 (3) , 200-208
A Model for the Safety Risk Evaluation of Connected Car Network
The automobile industry is preparing for what is now anticipated to be a big transition toward linked and autonomous vehicles. This introduces significant dangers both to individuals' privacy and to the intricate workings of electrical and electronic systems. Currently, the automotive industry has well-established and government-mandated safety risk management methods. They present an improved risk assessment methodology and demonstrate its validity with a real-world use case to facilitate the derivation of safety procedures and safety solutions for automotive integrated devices. This should make it simpler to ensure the safety of integrated vehicles. Some of the significant factors are the accessibility of functional safety to individuals who are not trained in safety and the appropriateness of existing techniques for said functional safety. The methodology that is used to assess the risks posed by properties and capacities includes a risk analysis as an essential component. The next step in the process of managing risks involves determining both the degree of risk and the magnitude of the potential impact. In this step, various different rules and requirements that are already in place are utilized to make any necessary adjustments. After much deliberation, a level of security has been settled on in order to facilitate the formulation of high-level security objectives. Due to the fact that this design has a good convergence with contemporary principles and criteria, it ought to be capable of being adequately suited to the prerequisites of the automotive sector.
- 27 August 2022
- Published by Conscientia Beam in Review of Computer Engineering Research
- Vol. 9 (3) , 169-180
Deep Learning Based Intrusion Prevention System in Vehicular Network
In the internet of vehicles, safety-based communication is carried out for prevention, mitigation, and alleviation of accidents through cooperative messages, position sharing, and the exchange of speed data between the vehicle (nodes) and corresponding roadside units. However, such networks are susceptible to false alarms and mispositioning of vehicles. It is therefore imperative to authenticate and identify normal messages from aggressive and incorrect messages. In this context, this paper has emphasized on a deep learning technique utilizing binary classification for segregating normal and malicious packets. The procedure is initiated by the preparation of training datasets from KDD99 and CICIDS 2018-type of open-source datasets having 1,20,223 packets and 41 features. An autoencoder is used in the preprocessing stage for the elimination of undesirable data right from the beginning. The 23 salient features are filtered out of 41. For training of the models, a structural deep neural network is utilized along with a Softmax classifier and rectified linear unit (ReLU) activation functions. The complete intrusion prevention (IP) mechanism is further trained & tested with Google co-labs as the open platform cloud service with open-source tensor flow. Furthermore, simulation data set developed in a network-simulating procedure is used for validation of the model. The results of the experimentation have established an accuracy 99.57% greater than existing recurrent neural network and convolution Neural Network models. For work in future, various datasets can be used for training to improve the accuracy and efficacy.
- 27 August 2022
- Published by Conscientia Beam in Review of Computer Engineering Research
- Vol. 9 (3) , 169-180
Deep Learning Based Intrusion Prevention System in Vehicular Network
In the internet of vehicles, safety-based communication is carried out for prevention, mitigation, and alleviation of accidents through cooperative messages, position sharing, and the exchange of speed data between the vehicle (nodes) and corresponding roadside units. However, such networks are susceptible to false alarms and mispositioning of vehicles. It is therefore imperative to authenticate and identify normal messages from aggressive and incorrect messages. In this context, this paper has emphasized on a deep learning technique utilizing binary classification for segregating normal and malicious packets. The procedure is initiated by the preparation of training datasets from KDD99 and CICIDS 2018-type of open-source datasets having 1,20,223 packets and 41 features. An autoencoder is used in the preprocessing stage for the elimination of undesirable data right from the beginning. The 23 salient features are filtered out of 41. For training of the models, a structural deep neural network is utilized along with a Softmax classifier and rectified linear unit (ReLU) activation functions. The complete intrusion prevention (IP) mechanism is further trained & tested with Google co-labs as the open platform cloud service with open-source tensor flow. Furthermore, simulation data set developed in a network-simulating procedure is used for validation of the model. The results of the experimentation have established an accuracy 99.57% greater than existing recurrent neural network and convolution Neural Network models. For work in future, various datasets can be used for training to improve the accuracy and efficacy.