Algoritmo de gestão de mobilidade para disseminação de vídeos em redes veiculares ultradensas

Abstract

The next generation networks aim to pave the for video distribution via vehicular ad-hoc networks (VANET), which will be composed of heterogeneous ultra-dense networks. These networks will take into account the existing communication infrastructure, in order to obtain greater spectral efficiency, such as the reuse of spectrum fees. However, the increase in the number of cells ends up making the mobility management process (ie, handover) a challenging task in ultra-dense VANETs. This is because vehicles often move between different networks, leading to an increase in unnecessary handover and greater network overhead, called the ping-pong effect. In this sense, an inefficient handover algorithm ends up providing videos with low Quality of Experience (QoE), which is caused by ping-pong handover decisions, which increases the packet and video frames losses. This thesis presents a handover algorithm for video distribution over ultra-dense VANET, called of Skip-HoVe. The proposed algorithm has the following characteristics: a handover mechanism aware of Mobility Prediction and QoE, as well as quality of service, resulting in handover with greater reliability and less frequency. The simulation results show the efficiency of the Skip-HoVe algorithm to deliver videos with 30% higher performance compared to existing handover algorithms, maintaining a ping-pong rate of around 2%.