I/O Optimized Recovery Algorithm in Vehicular Network Using PM-RBT Codes

Qiyuan He, Yaning Xu, Yuan Luo

Abstract: Network coding has been widely used in vehicular ad hoc networks (VANETs). However, the challenges for network coding come from resource constrains like I/O and memory. Existing network coding does not address the increasingly important problem of I/O overhead. In this paper, optimizing I/O consumption during the whole regeneration process is our main focus. Since vehicles may drive off the ad hoc network at halfway, it is necessary to regenerate the lost data in order to maintain reliability for further data reconstruction. We have come up with one recovery algorithm which is based on PM-RBT codes to combat against the high packet loss probability in VANETs. It significantly reduces the total I/O cost when taking the whole regeneration process into consideration. Simulation demonstrates that the algorithm performs better in case of high packet loss probability and large regeneration times.