Fast network selection mechanism for seamless connectivity on vehicular networks

The advanced wireless technologies make vehicle-to-vehicle communication possible. However, vehicular network has a limitation in terms of topology rapidly change. This will cause a discontinuous connection. For a smooth connection, horizontal and vertical handovers are required. A good handover mechanism should be able to select a proper network which meets application QoS parameters. Besides application QoS parameters, these parameters such as data link layer performance metrics, user requirements, should be considered. Thus making a decision for seamless communications falls into multi attribute decision making problems. Several decision making algorithms are proposed. Most algorithms require a real-time computation which causes high processing delay. Thus, the real-time computation is not well suited with delay sensitive applications. Moreover, the hardware box used in vehicles is quite tiny having a limitation in CPU speed, memory space, and etc. This paper proposes a simple space and time tradeoff approach to overcome the highly computational assessment by using a pre-calculated decision tree. The preliminary results show the computation time of the proposed approach is 7 to 10 times faster than the one of multi attribute decision making algorithms. The minimum memory requirement of this approach is only 700 KB.