Nearest-Vehicle Communication in Regular Street Systems

摘要

We consider a vehicular network with an underlying regular street system formed by horizontally and vertically oriented streets. We model vehicle locations on each street using Poisson point processes, and each vehicle transmits with some probability following slotted ALOHA. Each receiving vehicle connects to its partner-transmitting vehicle based on the following schemes: 1) Nearest-transmitter reception (NTR), 2) Nearest-transmitter reception with selective thinning (NTR-II), and 3) Nearest-receiver transmission (NRT). In NTR and NTR-II, each receiver receives from its nearest transmitter, whereas in NRT, each transmitter transmits to its nearest receiver. NTR-II is a modified version of NTR, where the transmitters not closest to any of the receivers remain idle. Under each scheme, we calculate the probabilities of successful reception for the general and intersection users, i.e., vehicles not at intersections and at intersections. We study how the nearest-transmitter/receiver distance properties in each scheme differ for general and intersection users, and in turn, how those properties affect their probabilities of successful reception. Also, we show that we can obtain higher data rates in NTR and NTR-II compared to NRT in the high-reliability regime.