Abstract—Vehicle cloud computing is a newly emerged research filed which focuses on the usage of cloud computing for vehicular networks. Vehicles interact with each other through wireless communication and sense their surrounding via embedded sensors. In addition, vehicles do not suffer from inadequate storage and processor, so they can be considered as ideal candidate nodes in a cloud while moving on the roads and highways or stopping behind traffic light of parking lots. Despite designing appropriate vehicle cloud architecture is a challenging task owing to the various types of applications and requirements in this area, we have witnessed many architectures are proposed by researchers. Therefore, in this paper we discussed the recent works, and proposed vehicle cloud architecture taxonomy based on architecture and services types in order to find out remaining drawbacks need to be addressed. Index Terms—Cloud Computing, Taxonomy, Vehicular Networks, Vehicle Cloud. I. INTRODUCTION Nowadays by evolving the new intelligent technologies as an impressive contributor of transportation systems, the employing of Intelligent Transportation Systems (ITS) concept has greatly reaped attention of governments and academia in this area. Meanwhile, the use of vehicle Ad Hoc Network (VANET) as a form of wireless ad hoc networks which provides connectivity between vehicles and Road Side Units (RSUs) along the road [1]. The main purpose of VANET is to provide ubiquitous connectivity while on the road to mobile users, who are otherwise connected to outside world by other networks at home, and efficient Vehicle-to-Vehicle (V2V) communications that enable the ITS [2-3]. Although VANET is a form of wireless networks, it has its unique characteristics owing to the high nodes mobility and unreliable channel conditions [3]. These characteristics create many challenging research issues. Cloud Computing (CC) is a comparatively new model in the computing world provided to enable ubiquitous, applicable and on-demand network access to the vast shared computing resources consists of all networks, servers, storages, applications and services [4]. According to Horgan et al. [4], cloud model comprises five essential characteristics containing On-demand self-service, Broad network access, Manuscript received October 10, 2014. This research is supported by UM High Impact Research MoE Grant UM.C/625/1/HIR/MOHE/FCSIT/09 from the Ministry of Education Malaysia. Mohammad Reza Jabbarpour, Ali Jalooli, Alireza marefat and Rafidah Md Noor are with the Department of Computer System and technology, Faculty of Computer Science and Information Technology, University of Malaya, Kuala Lumpur, 50603, Malaysia (e-mail: reza.jabbarpour, ashkansp2, alimarefat69{@siswa.um.edu.my}, and fidah@um.edu.my). Resource pooling, Rapid elasticity or expansion and Measured service. In addition its deployment composed of four main models including Private, Public, Community and Hybrid. In a parallel view with Armbrust et al. in [5], the most significant cloud component refers to its service types that categorized as Infrastructure as a Service (IaaS), Platform as a Service (PaaS), and Software as a Service (SaaS) where IaaS is the most basic type. Besides, some other service types are proposed by researches in the recent years such as Information Resource as a Service (IRaaS), SEnsor as a Service (SEaaS), Industrial Machinery as a Service (IMaaS), Network as a Service (NaaS) and STorage as a Service (STaaS). Vehicular Cloud (VC) computing is a new concept which combines Vehicular Ad-hoc Network (VANET) with Cloud Computing (CC) in order to solve the VANET’s shortcomings. This concept has been introduced by Olariu [6-8] in recent years. Since vehicles do not suffer from inadequate storage and processor, and also they can communicate with each other through vehicular networks, they can be considered as ideal candidate nodes in a cloud while moving on the roads and highways or stopping behind traffic light of parking lots. These abundant and underutilized computational resources can be organized in order to provide third-party or community services. Vehicular Cloud or cloud enabled vehicular architecture is a novel and unique idea among recent amazing technologies that expand properties of vehicular networks and applications, and maximizes performance with more simplicity and flexibility. Although compatibility of vehicular networks with future technology such as smart cities is important, but the main concern of vehicular cloud is to enhance vehicular networks, architectures and applications by cloud computing. High computing performance, security, privacy, scalability, reliability, data mining and aggregation, low cost and on-demand services are some of the advantages that cloud brings for vehicular networks. This paper divides the Vehicle Cloud Architectures (VCAs) into two main categories based on their functionalities: 1.General architectures: in this category, researchers bring the VANETs into Cloud by creating or designing a new VC architecture based on different applications and requirements. 2. Specific architectures: in this category, researchers attempt to use VC architecture in order to improve some issues such as vehicles traffic, routing, disaster management and vehicle or truck monitoring. Further, we proposed taxonomy based on their services and architecture types. The rest of the paper is organized as follows: in section 2, A Taxonomy-based comparison of Vehicle Cloud Architectures Mohammad Reza Jabbarpour, Ali Jalooli, Alireza Marefat and Rafidah Md Noor