International Journal of Computer Science and Telecommunications [Volume 6, Issue 11, December 2015] 13 Journal Homepage: www.ijcst.org Sara Esmaeillou 1 , Ramin Karimi 2 and Seyed Mahdi Jameii 3 1,2 Department of Computer Engineering, Malard Branch, Islamic Azad University, Tehran, Iran 3 Department of Computer Engineering, Shahr-e-Qods Branch, Tehran, Iran Abstract– Nowadays, VANETs are one of the most important networks in the mobile network research domain; these networks have highly mobile nodes that are constantly exchanging information. High mobility, large number of vehicles, and lack of concentrated infrastructures has caused many problems regarding routing, communications security, Packet lost, repeated communications, inappropriate information transfer, etc. Moreover, due to the high mobility, the topology of these networks is constantly changing and optimal information propagation is one of the most important challenges in these networks. In this paper, the VANET`s nodes, i.e., vehicles, are grouped with an algorithm and information is exchanged in the form of small groups. This grouping reduces the constant topology changes. Moreover, it is predicted that grouping mitigates lost data and information transmission in different routes. This paper investigates the formation, selection, and identification of each group. Moreover, the grouping algorithm is optimized using particle swarm optimization to optimize the formation and head selection of each group. Index Terms– VANETs, Grouping and Particle Swarm Optimization I. INTRODUCTION ANETs are a type of ad hoc networks that allows communication between adjacent vehicles and between vehicles and fixed equipment usually installed at the side of the roads. The main goal of VANETs is creating more comfort and security for the passengers [2].Accordingly, an electronic device is installed in each vehicle that allows connecting to other vehicles (OBU-On board unit) [14]. Therefore, each vehicle equipped with VANET. Operates as a node in the ad hoc network and is able to send or receive others messages through the network. These messages are mostly used for security and traffic control purposes [6], for instance, accident alert, entertainment, traffic observation, road sign announcement, paying parking expenses, etc. These messages are useful tools for the driver to select an appropriate route. In addition, multimedia and internet facilities are embedded for the passengers. Paying tolls and parking expenses are other services of these networks. Today, large vehicle manufacturing factories have initiated different projects regarding these types of networks and equipping their vehicles with VANET capabilities. Vehicular ad hoc networks (VANETs) are a leading challenging group of mobile ad hoc networks (MANETs) [8], which have currently attracted considerable research interest regarding wireless networking and vehicle industries. Due to their relatively high flexibility, communication in VANETs poses stronger challenges in public communication. Environments without an underlying structure and a highly dynamic network topology constantly change these networks [10], [11]. VANET can be considered as a component of intelligent transportation systems (ITSs) [4]. As mentioned, highly dynamic nodes and servers without a central station cause collisions in wireless VANET communications and packages are mostly lost or delayed. In these scenarios, simultaneous communication easily fails. VANETs employ different wireless technologies, including DSRC, which is a type of Wi-Fi, as well as WIMAX and cellular technology. Other short band wireless protocols, e.g., IEEE802.11, Bluetooth, and CALM are also usable in these networks [12]. This paper proposes VANET grouping that allows using particle swarm optimization to optimize communications in concentrated traffic conditions and the formation these groups. The remainder of the paper is organized as follows: section II presents the related works; section III, particle swarm optimization; section IV, a brief summary of grouping; in section IV-A, group identification; in section IV-B, weight calculation of each node using PSO; in section IV-C, inviting to group formation using PSO; in section IV-D, forming optimal groups using PSO; in section IV-E, group selection; and in section V, conclusions are presented. II. RELATED WORKS There are many studies regarding grouping in VANETs. GAP protocol based grouping and authentication is proposed in [9] to reduce the delay and lost messages. The node clustering methods are introduced in [14], in which the node at the center of the group is considered as the leader node. These two methods do not specify the parctical processes that nodes perform to manage the group and the collected information is not sufficiently reliable. A grouping algorithm is used in [3] to reduce the number of communications and lost packages and prevent information propagation in different V Spatial-based Grouping of VANETs Using Particle Swarm Optimization Algorithm ISSN 2047-3338