Trailblazing Safe Routes Leveraging Verkle and Berkle Trees for Enhancing Routing Security and Efficiency in Opportunistic Mobile Social Networks

Abstract

A significant majority of individuals utilize Mobile Social Networks (MSNs), as common platform for interpersonal communication and social interaction. An overview of recent developments in MSNs, including their types, characteristics, applications, and usage, is given in this thesis. This thesis also explores different aspects including, architecture, privacy and security issues, as well as challenges and their existing solutions. Also, recommendations on security and privacy measures, along with few future research directions in mobile social networking are provided in this thesis. MSNs have a high impact on society and are responsible for the emergence of Opportunistic Mobile Social Networks (OMSN). OMSN eliminates the need for the internet for communication and data exchange, prompting a detailed investigation on its architecture, and potential future research directions. One important point of focus is to compare the design and problems of OMSN with those of regular MSNs, highlighting the main contribution of addressing privacy and security issues in OMSN. Routing in OMSN poses significant challenges due to opportunistic interactions between unknown and unreliable nodes. This thesis employs machine learning techniques to introduce the Socialized Proficient Routing (SPR) model. The SPR model consists of three distinct phases. In the feature selection phase, notable features are identified from the training dataset through the implementation of the Boruta wrapper algorithm. The training phase employs various machine learning classifiers, namely Naïve-Bayes, Decision-Tree, Neural-Networks, Support-Vector-Machine, and Random-Forest. In the testing phase, the model accurately identifies trustworthy neighbour (friendship) nodes for routing purposes. By utilizing social features, the SPR model improves trust and dependability among involved nodes and performs better, especially when using the Random Forest classifier. OMSN faces threats from data integrity attacks, leading to an evaluation of tree-based data...