Francois Baccelli - Prof at Univ. Texas at AUSTIN Communication systems
Date: July 23rd 2019 Location: Eurecom - Eurecom
Title: On the Performance Evaluation of Vehicular Gateways based Wireless Networks. Abstract: This presentation discusses the capacity gains obtained when using vehicles as mobile gateways to enable large-scale connectivity for the Internet of Things (IoT). Two types of delay tolerant architectures are analyzed using techniques of stochastic geometry, queuing theory, and wireless networking. In the single-hop architecture, at any given time, a vehicle is assumed to communicate with a device selected at random in a domain centered at the vehicle, which is referred to as the coverage domain. We first consider the case where vehicles move on roads modeled by a Poisson line process and where vehicles are Poisson point processes on these roads, respectively, and then in the case where vehicles are UAVs moving on planned trajectories. The multi-hop architecture, which is referred to as mesh+vehicular, is based on mesh communications between IoT devices and short range communication between repositories located along the roads, again represented as Poisson lines, and vehicular gateways passing by, also represented as Poisson point processes on thes lines. For both architectures, we propose system level models and evaluate their performance gains w.r.t. the conventional cellular architecture of the Euclidean plane. We analyze the network performance by explicitly deriving the distribution of the signal-to-noise and interference ratio experienced by typical network elements. For the mesh+vehicle multi-hop architecture, we also show that if the density of IoT devices is large and the density of repositories scales with that of IoT devices, then the proposed architecture has a capacity scaling that outperforms that of the conventional mesh+cellular architecture and that the performance gain is polynomial in the density. In both cases, the capacity gain is obtained to the expense of delays linked to the motion of vehicles. Biography: F. Baccelli shares his time between UT Austin and INRIA. His research directions are at the interface between Applied Mathematics and Communications. He is co-author of research monographs on point processes and queues, max plus algebras and network dynamics, stationary queuing networks, stochastic geometry and wireless networks. He received the France Télécom Prize of the French Academy of Sciences in 2002 and the ACM Sigmetrics Achievement Award in 2014. He also received the 2014 Stephen O. Rice Prize and the 2014 Leonard G. Abraham Prize Awards of the IEEE Communications Theory Society for his work on wireless network stochastic geometry. He is a member of the French Academy of Sciences. He just started an ERC project on the mathematics of communication networks.