Researches On VANET Middleware For Service Recommendation Based On Context

The Vehicular Ad-hoc Network(VANET)consists of vehicles and roadside infrastructures,which performs wireless communication and information exchange through vehicle-to-vehicle,vehicle-to-road and vehicle-to-Internet according to agreed communication protocols and data exchange standards.Due to the heterogeneity of embedded devices,networks and application systems between different vehicles,middleware is widely used as a link in order to improve service and data management.With the development of service number and service type in VANET,the information explosion in VANET is increasing seriously,so there is an urgent need for VANET to provide services considering users’ individual requirements.Middleware can effectively shield the heterogeneity of the embedded devices,and provide a unified and common interface for applications.This paper proposes a context based middleware for VANET service recommendation called VSROSGi which takes the advantage of OSGi service oriented architecture,vehicles can not only dynamically obtain the service resource information in roadside units,but also use context information and historical service data to process service data.In this middleware,we design two service recommendation mechanisms to improve the customer satisfaction.Secondly,a context-based service recommendation algorithm is proposed for the differences of the vehicle’s own situation and user’s service preference,which can generate the user’s service preference model based on the vehicle historical service data.Then it can carry out contextual reasoning on the optional services when it comes to service request stage,otherwise,it can not only calculate the distance between the optional services and the service request vehicle,but also the similarity between the optional services and the user preference model.Only when services meet the vehicle’s context constraints and the user’s preference model that they could be recommended to the user,as result of which,customer satisfaction can be enhanced.Thirdly,the VANET active service recommendation algorithm for POI(point of interest)-oriented is proposed for the application scenario where users have no clear service request targets.The algorithm combines the thought of crowdsourcing and takes the vehicles with the service records in POI as the crowdsourcing participants.By calculating the similarity of the service preference models between vehicles,the vehicle user who has a similar service preference with the service requester is found,and then service selections of the similar user are recommended to the service requester,which aimed at diversified service recommendation that meets the service requester’s preference in POI.Simulation results show that these two algorithms can provide personalized service recommendation that meet the context constrains for the vehicle users,and effectively improve the customer satisfaction.