Mobility Modeling And Data Forwarding Protocols For Vehicular Social Networks

Vehicular Social Networks(VSNs)is an emerging field of communication where relevant concepts are being borrowed from two different disciplines,i.e.,Vehicular Ad-hoc Networks(VANETs)and Mobile Social Networks(MSNs).This emerging paradigm presents new re-search fields for content sharing,data dissemination,and delivery services.Based on Social Network Analysis(SNA)applications and methodologies,interdependencies of network enti-ties can be exploited in VSNs for prospective applications.VSNs involve social interactions of commuters having similar objectives,interests,or mobility patterns in the virtual commu-nity of vehicles,passengers,and drivers on the roads.The mobile crowdsourcing capabilities of commuters in VSNs can provide promising solutions to several challenges in today's smart cities,such as traffic congestion control,traffic management,smart parking,and route recom-mendation.In VSNs,mobile crowdsourcing depends upon the cooperative and reliable data forwarding.Compared to other ad-hoc networks and delay tolerant networks,the mobility of nodes is highly dynamic and restricted to roads in VSNs.The vehicular mobility is profoundly influenced by many factors including traffic regulation,city development,and planning,working hours/days,work and lifestyle of inhabitants,social values,roadblocks.and so on.Validation and evaluation of different proposed protocols for VSNs rely on simulation.The value of the validation and the credibility of results are,therefore,highly dependent on the underlying mo-bility model used for simulation.Unfortunately,the simulation performance evaluation of the context-aware application for VSNs is often biased by the underlying mobility models.Incor-rect representation of vehicular mobility can lead to misleading results and conclusions,even if a flawless network-level simulator is used for performance evaluation.The results obtained will be considered more accurate and credible if and only if the underlying mobility model is analogous to real-world car traffic.This dissertation presents the construction of large-scale urban mobility models and two proposed novel data forwarding protocols based on social features metrics and node selfishness.First,this thesis provides an overview of existing mobility models and socially-aware data for-warding protocols in cooperative and non-cooperative(selfish)VSNs.Second,it demonstrates the characterization of big urban traffic data to characterize the essential features of urban mo-bility and construct large-scale mobility models.Next,a set of socially-aware data forwarding protocols for data forwarding in cooperative and socially-selfish VSNs are presented.Also,it presents both analytical and simulation-based experiments to evaluate generated mobility mod-els and proposed protocols.First,we present the characterization of big urban traffic data to characterize the essential features of urban mobility and construct large-scale urban mobility models.This work exploit-s the road and traffic information to enhance trip generation algorithm and traffic assignment technique based on the weighted segments of roads.Besides,it presents extensive observations and corrections on the OpenStreetMap imported to Simulation of Urban Mobility(SUMO)to make it analogous to real-world road topology.The experimental results and validation pro-cess show that the generated mobility models reveal realistic behavior required for analysis of context-aware applications of VSNs.Second,inspired from the social acquaintance in our daily life,Social Acquaintance based Routing Protocol(SARP)for VSNs is proposed,which collectively considers three social feature metrics to make a forwarding decision.The proposed protocol aims to reduce End-to-End delay and improve the packet delivery ratio in VSNs.Additionally,SARP overcomes the shortcoming of topology-based routing and optimum local situation of geographically based routing protocols by considering the global and local community acquaintance of nodes.Extensive simulations are performed under constant node density with different mobility speed and constant speed with varying node density to study the effect of node mobility speed and density on end-to-end delay and packet delivery ratio.The simulation results show that SARP outperforms GPSR by 22%and 26%in terms of end-to-end delay and packet delivery ratio respectively.Also,SARP outperforms AODV in terms of end-to-end delay.Third,we propose a Cooperative Data Forwarding(CDF)mechanism to stimulate the self-ish nodes to participate in data forwarding.To enhance data forwarding mechanism,CDF is based on a socially-aware routing mechanism and a cooperative algorithm using direct obser-vations and mobile crowdsourcing information to stimulate selfish nodes to participate in data forwarding.Besides,CDF is a multi-hop single copy forwarding mechanism which consid-erably decreases the network overhead.The experimental results and analysis are based on real-world vehicular mobility dataset based on mobile crowdsourcing and mobility models.The results show that CDF encourages more and more nodes to cooperate and improves network per-formance significantly in terms of data delivery ratio,transmission cost,and end-to-end delay which can significantly improve the mobile crowdsourcing applications of VSNs.