The MongoDB injection dataset: A comprehensive collection of MongoDB - NoSQL injection attempts and vulnerabilities.

We present the 'NoSQL Injection Dataset for MongoDB, a comprehensive collection of data obtained from diverse projects focusing on NoSQL attacks on MongoDB databases. In the present era, we can classify databases into three main types: structured, semi-structured, and unstructured. While structured databases have played a prominent role in the past, unstructured databases like MongoDB are currently experiencing remarkable growth. Consequently, the vulnerabilities associated with these databases are also increasing. Hence, we have gathered a comprehensive dataset comprising 400 NoSQL injection commands. These commands are segregated into two categories: 221 malicious commands and 179 benign commands. The dataset was meticulously curated by combining both manually authored commands and those acquired through web scraping from reputable sources. The collected dataset serves as a valuable resource for studying and analysing NoSQL injection vulnerabilities, offering insights into potential security threats and aiding in the development of robust protection mechanisms against such attacks. The dataset includes a blend of complex and simple commands that have been enhanced. The dataset is well-suited for machine learning and data analysis, especially for security enthusiasts. The security professionals can use this dataset to train or fine tune the AI-models or LLMs in order to achieve higher attack detection accuracy. The security enthusiasts can also augment this dataset to generate more NoSQL commands and create robust security tools.

Private 5G networks: a survey on enabling technologies, deployment models, use cases and research directions.

Today's modern enterprises are adjusting to new realities of connectivity. As companies become more distributed and autonomous, emerging applications demand more bandwidth, low latency, more spectrum, and higher reliability. 5G technology can aid many industries or enterprises to make quicker and better business decisions. Private 5G networks, also called 5G Non-Public Networks (5G-NPN), is a 3GPP-based standalone 5G network positioned for a particular enterprise or use case that delivers dedicated network access. It sets to transform industry landscapes with networks capable of rapidly deploying modern use cases and the scalability to meet constantly increasing demands of data capacity and speed. They help generate more revenue for operators who can partner with enterprises to build and manage networks on-premise or in the cloud. The objective of this work is to offer a thorough summary of private 5G networks to assist academicians, researchers, and network developers to quickly grasp their functionalities without needing to go through the standards, specifications, or documentation. This paper discusses various key private 5G network design goals and requirements, examines its deployment scenarios, and explores spectrum considerations and security aspects. The paper presents several enterprise use cases to illuminate how the networks can deliver the demands and services expected by the industries. It also provides an overview of some of the open-source projects considered by various organizations for private network deployment. Finally, several research directions are introduced, emphasizing enterprise challenges to deploying 5G networks.

Defending against adversarial attacks on Covid-19 classifier: A denoiser-based approach.

Covid-19 has posed a serious threat to the existence of the human race. Early detection of the virus is vital to effectively containing the virus and treating the patients. Profound testing methods such as the Real-time reverse transcription-polymerase chain reaction (RT-PCR) test and the Rapid Antigen Test (RAT) are being used for detection, but they have their limitations. The need for early detection has led researchers to explore other testing techniques. Deep Neural Network (DNN) models have shown high potential in medical image classification and various models have been built by researchers which exhibit high accuracy for the task of Covid-19 detection using chest X-ray images. However, it is proven that DNNs are inherently susceptible to adversarial inputs, which can compromise the results of the models. In this paper, the adversarial robustness of such Covid-19 classifiers is evaluated by performing common adversarial attacks, which include the Fast Gradient Sign Method (FGSM) and Projected Gradient Descent (PGD). Using these attacks, it is found that the accuracy of the models for Covid-19 samples decreases drastically. In the medical domain, adversarial training is the most widely explored technique to defend against adversarial attacks. However, using this technique requires replacing the original model and retraining it by including adversarial samples. Another defensive technique, High-Level Representation Guided Denoiser (HGD), overcomes this limitation by employing an adversarial filter which is also transferable across models. Moreover, the HGD architecture, being suitable for high-resolution images, makes it a good candidate for medical image applications. In this paper, the HGD architecture has been evaluated as a potential defensive technique for the task of medical image analysis. Experiments carried out show an increased accuracy of up to 82% in the white box setting. However, in the black box setting, the defense completely fails to defend against adversarial samples.