PROMPT: A cross-layer position-based communication protocol for delay-aware vehicular access networks

Abstract

Vehicular communication systems facilitate communication devices for exchange of information among vehicles and between vehicles and roadside equipment. These systems are used to provide a myriad of services ranging from traffic safety application to convenience applications for drivers and passengers. In this paper, we focus on the design of communication protocols for vehicular access networks where vehicles access a wired backbone network by means of a multi-hop data delivery service. Key challenges in designing protocols for vehicular access networks include quick adaptability to frequent changes in the network topology due to vehicular mobility and delay awareness in data delivery. To address these challenges, we propose a cross-layer position-based delay-aware communication protocol called PROMPT. It adopts a source routing mechanism that relies on positions independent of vehicle movement rather than on specific vehicle addresses. Vehicles monitor information exchange in their reception range to obtain data flow statistics, which are then used in estimating the delay and selecting best available paths. Through a detailed simulation study using ns-2, we empirically show that PROMPT outperforms existing routing protocols proposed for vehicular networks in terms of end-to-end packet delay, packet loss rate, and fairness of service.

Introduction

In recent years, the awareness of transportation problems has increased due to rising fuel costs, air pollution, and increased number of accidents. In urban environments, where the population density is high, these problems are more pronounced. To alleviate these problems, Intelligent Transportation Systems (ITS) are proposed to utilize information about vehicle traffic through a communication structure. Such systems are useful in many applications including emergency notification systems, vehicle traffic management, and traveler information/support.

An important component of ITS is the vehicular communication network called VANET that enables information exchange among vehicles. VANET is a self-configuring mobile ad hoc network of vehicles interconnected via wireless link. VANETs’ services can be classified as Vehicle-to-Infrastructure (V2I) and Vehicle-to-Vehicle (V2V). V2I communication provides connectivity between roadside equipments and vehicles. V2V provides connectivity directly between vehicles without relying on an existing infrastructure. There are increasing number of efforts such as DSRC¹ and 802.11p² to standardize communication protocols for VANETs.

VANETs are fundamentally different from other ad hoc networks in several ways. They have geographically constrained but highly dynamic topologies, predictable mobility, and fast varying connectivity. These characteristics make it difficult to apply traditional solutions developed for MANETs directly to VANETs. A major focus of recent research on VANETs is to minimize the end-to-end communication delay, which is important both in providing emergency information and also in delay sensitive applications. For example, consider user communication services, which require information exchange among passengers from different cars, and info-tainment applications involving real-time multimedia. They try to improve the convenience and comfort level for travelers and often requires vehicular networks with small end-to-end delay. VANETs may also be used to develop marketing tools for businesses. For instance, restaurants or hotels can broadcast their information over VANETs and interested drivers or passengers can send a query for more information or even make a reservation. VANETs may also be used to develop e-commerce applications wherein travelers can make business transactions. Such applications require networks which are delay-efficient and reliable.

As part of our previous research, we developed a cross-layer protocol CVIA for highways [1] which aimed at providing location-independent throughput and fairness to vehicles. In this work, we focus on urban environments laced with intersections. Such road networks pose several new challenges such as multiple paths, overhead due to routing decisions at intersections, packet congestion, and highly mobile relay nodes. We propose a new protocol called PROMPT,³ which is a cross-layer position-based delay-aware protocol that addresses these challenges and delivers packets over minimum delay paths.

PROMPT adopts a position-based routing approach to eliminate negative effects of high node mobility rates. Paths are determined at source nodes based on network traffic statistics collected during propagation of service advertisements of base stations. Taking advantage of the fact that data traffic along roads vary much slower than vehicles’ positions, statistics collected in individual nodes are mapped to locations. We introduce analytical methods to map such statistics to delay estimations at source nodes, which allows us to distribute the traffic over all available paths.

Through an extensive simulation study, we show that PROMPT is able to use local information effectively to predict the end-to-end delay. We also show that PROMPT outperforms DSR [2], CAR [3], GPSRJ+ [4], and VADD [5] protocols in terms of end-to-end delays, success rates, and fairness.