Convolutional Neural Networks (CNNs) are a class of deep neural networks that have proven highly effective in areas such as image and video recognition. CNNs typically include several types of layers, such as convolutional layers, activation layers, pooling layers, and fully connected layers, all of which contribute to the network’s ability to recognize patterns and features. The pooling layer, which often follows the convolutional layer, is crucial for reducing computational complexity by performing down-sampling while maintaining essential features. This layer’s role in balancing the symmetry of information across the network is vital for optimal performance. However, the choice of pooling method is often based on intuition, which can lead to less accurate or efficient results. This research compares various standard pooling methods (MAX and AVERAGE pooling) on standard datasets (MNIST, CIFAR-10, and CIFAR-100) to determine the most effective approach in preserving detail, performance, and overall computational efficiency while maintaining the symmetry necessary for robust CNN performance.

Cloud computing is an online data storage system that can be easily accessed anywhere. Various algorithms and techniques have been developed to secure cloud data, which, if used correctly, can detect cloud attacks. These algorithms can only detect cloud data attacks and cannot prevent such attacks. If an attacker gains access to the data, he or she can tamper with the data in any way that could lead to misuse of the data. This paper will design a secure architecture for securing data with various algorithm implementations on the architecture so that cloud data can be saved from replay attacks. First, a text file will be taken, the text file will be encrypted using the encryption mechanism, and this file will be transmitted on the SaaS cloud server platform. Whenever a file is uploaded to a cloud server, it will be broadcast across the network, after which any user can access it, whether it is valid or invalid. Whenever a file is downloaded from a cloud server, it will be necessary to decrypt it to access it. If the file is successfully decrypted, then the file will be verified. A verification mechanism will be used to verify the data, which will be used to check whether a replay attack has been made on the data or not. If the data authentication is successful, this data will be called authentic data. The novelty of this paper is that an algorithm has been developed in this paper to prevent replay attacks so that if the data is retransmitted with slight changes, then the authenticity and inauthenticity of the data will be identified. Recent papers will be compared with the proposed paper in the conclusion, and it will be pointed out how much better the current technique is than the previous one.

Road traffic accidents become more of a problem as there are more automobiles on the road. Accidents are impossible to completely prevent, but there are techniques to lessen their impact and manage their aftereffects to limit harm. One of the biggest issues is exchanging car accident alerts with other vehicles on the road. A number of technologies are in use to prevent traffic collisions. It is planned to develop a system for detecting collisions between vehicles based on the vehicular ad-hoc network (VANET) concept. Additionally, it immediately alerts the neighborhood hospital and social media about the incident. The proposed technique uses a warning system supported by collision potential and is implemented on OpenStreetMaps and at locations of interest. We assess our system’s performance using the simulation of urban mobility (SUMO) software tool, which employs OpenStreetMap data to simulate vehicle-to-vehicle communication and accident detection. SUMO generates traffic on the road for our evaluation. To gauge the effectiveness of our system, we measure the size of messages exchanged within the VANET and demonstrate its feasibility. Furthermore, we compare our results with previously published works in the same field of study for reference.