Elsevier

Vehicular Communications

Volume 38, December 2022, 100519
Vehicular Communications

Hybrid routing for safety data with intermittent V2I connectivity

https://doi.org/10.1016/j.vehcom.2022.100519Get rights and content

Abstract

V2I communications complement V2V safety applications by dealing with emergency situations such as crash scenarios that the V2V program cannot handle or that could be addressed more effectively. However, depending on the deployed infrastructure, we can distinguish between three situations regarding allowed communications: infrastructure-based, infrastructure-less and intermittent infrastructure. We will focus in vehicular networks with an intermittent connectivity, where the unavailability of direct communication with RSUs poses huge problems, especially when vehicles have to send safety-related messages to the infrastructure. In fact, a scattered vehicular infrastructure leads to non-homogeneous coverage, especially in highway scenarios extending over long distances; it is very expensive to set up an infrastructure across the whole road. Our objective is to address the challenges related to safety information forwarding towards the RSU in such networks by enabling vehicles to make an optimal forwarding decision based on contextual information relative to the infrastructure and the neighbourhood. In this work, we present HyRSIC (Hybrid Routing for Safety data with Intermittent V2I Connectivity), a routing approach, combining geographical and topological routing, proposed to deal with data forwarding in intermittent infrastructure-based networks. We conducted an in-depth evaluation from three different perspectives: scalability, reliability and efficiency. The results show that HyRSIC is able to scale quickly and adapt promptly to density and speed variations. Compared to GPSR and its variant ILTS+GPSR, HyRSIC offers the best packet delivery ratio for sparse networks, 12-27% higher, and the lowest End-to-End delivery delay, 25-50% less.

Introduction

Vehicular networks are promising safer roads and a smooth driving by offering a several number of safety and non-safety applications [15]. Cooperative crash and intersection crash warning are examples of safety applications [8]. This category of application is considered as delay sensitive and requires rapid and reliable forwarding of information to the concerned destination (vehicles or RSUs (Road Side Units)). In traditional VANETs (Vehicular Ad-Hoc Networks), we can distinguish between three major communication levels [27], [12]: V2V (Vehicle to Vehicle), I2V (Infrastructure to Vehicle) and Hybrid Vehicular Communication (HVC) (both V2V and I2V). The effective delivery of VANET applications and services among network nodes is tightly related to the flexibility and performance of the deployed routing protocols, which are held back by the shared medium and the low bandwidth, characterizing the vehicular environment. A routing process is needed in order to select paths, able to relay information between vehicles until reaching the destination. In fact, vehicles must rely on V2V multi-hop communications to reach remote nodes or infrastructure if there is no nearby RSU for direct V2I (Vehicles to Infrastructure) communications. A routing protocol governs the way that entities exchange information; route establishment and packet forwarding are essential components of the routing process, besides required operations for route maintenance and route failure recovery [18], [13], [14]. Geography-based protocols are considered to be best suited for this type of network, as the data dissemination, for many vehicular applications, is closely linked to the geographic position of the involved nodes (emitter, forwarders and receiver) [13], [14]. This family of routing protocols works jointly with Location-based Services (LS) whose mission is to provide and maintain location information. Usually, routing and location service are handled separately, inducing huge control and localization overhead [3]. Merging these two processes reduces overhead and also improves performance [3]. In this context, our objective is to deal with the challenges related to safety information forwarding towards the RSU in intermittent infrastructure-based networks. In fact, scattered vehicular infrastructure leads to non-homogeneous coverage, especially in highway scenarios as they extend over long distances and putting infrastructure throughout is very expensive. With a discontinuous connectivity, vehicles can not have direct communication with the RSUs, which poses enormous challenges, especially when vehicles have safety related messages to send to the infrastructure. In such a scenario, vehicles can rely on V2V communications to forward the packet until reaching the infrastructure [29], [19], [23], [24]. However, the multi-hop V2V and hybrid proposed solutions do not offer any guarantee for the critical data End-to-End delivery delay. In this context, we propose HyRSIC (Hybrid Routing for Safety data with Intermittent V2I Connectivity). It allows vehicles to make the optimal forwarding decision based on contextual information relative to the infrastructure remoteness and the neighbourhood knowledge. It incorporates a decision making algorithm for V2V and V2I communications coupled with an infrastructure location service, for intermittent infrastructure-based networks in highways. We push the merging further by integrating the proposed LS in the vehicular beaconing system, and we make decision based on delay estimation. In fact, based on the paths availability and the estimated delays, the vehicle will be able to take the optimal forwarding solution for critical data.

So, HyRSIC is composed of three main building blocks; a localization service, a decision making algorithm, and a routing algorithm. The main contributions of this work are the following:

  • Localization service: We present an enhanced version, ILTS+, of our data collection and tracking mechanism ILTS (Infrastructure Localization service and Tracking Scheme) [16]. The added value to ILTS is the possibility to compute the distance from the infrastructure not only in straight highways but also in curved ones. ILTS is responsible for determining the location of neighbouring RSUs as well as the paths available towards them. The scheme is integrated in the vehicular beaconing system which significantly reduces network overhead.

  • Decision making algorithm: Two crucial phases are undertaking by the decision making algorithm; the first one is the selection of the targeted RSU based on the paths availability and the delay analysis; four possibilities are presented; The previous RSU via (i) the same or (ii) the opposite direction. The next RSU via (iii) the same or (iv) the opposite direction. And the second one is the selection of the forwarding direction based on the hops number.

  • Forwarding algorithm: We present Best Hop Selection (BHS), a mechanism where the selection of the node is based not only on the distance of the neighbour from the destination, but also the quality of the path proposed by the neighbour, and the quality of the link between the source vehicle and its neighbour. The coverage of the path proposed by the neighbour is used in case no full path is available towards the infrastructure. All these information are proactively gathered thanks to ILTS.

The remainder of this paper is structured as follows. Section 2 presents the related work; Section 3 introduces the notations used in this work. Section 4 details the functional behaviour of our scheme. We present a model verification in Section 5. In Section 6, we evaluate the scalability, reliability and efficiency of HyRSIC; and Section 7 concludes the paper.

Section snippets

Related work

In this section, first, we present localization services propositions. Then, we review the state of the art of previous works that tackled the scattered infrastructure based routing in VANETs.

Model assumptions and notations

In this section, we present the model assumptions and the notations used in this paper. To simplify the discussion, we have made the following assumptions about the general model we are considering:

  • Vehicles are travelling in a multi-lane bidirectional highway without obstacles and side exits.

  • Each vehicle knows its own location and direction, thanks to GPS (Global Positioning System), which feeds vehicles with accurate information about time and position.

  • All vehicles have an omnidirectional

HyRSIC functional behaviour

HyRSIC is a hybrid routing protocol for the forwarding of safety data in networks with intermittent infrastructure, which relies on a decision making algorithm designed to select the most suitable path, based on fresh geographical and topological routing information. Our objective is to enable the vehicles in uncovered areas to reach the infrastructure in a timely manner. First, the forwarding direction is decided according to whether or not full paths are available towards the known RSUs.

Model verification

In the following we will present the best cases scenarios and the worst case scenarios for both Travail Time and Forwarding Time:Delayestimation={TTB<TT<TTWFTB<FT<FTW

  • Travail Time:

    • Worst case scenario: In the worst case the source vehicle has just moved out of the left RSU coverage area and does not have full paths to any of the RSUs. So it will have to buffer the message and cross the whole distance separating both RSUs until it delivers the message to the subsequent infrastructure. So the worst

Performance evaluation

In this section, we assess the performance of HyRSIC while varying the network density, the road length, the maximum speed of vehicles and the number of sources of warning messages. We compare HyRSIC to GPSR (See Section 2.2) and a variant of GPSR integrating ILST [16]. GPSR was used in this evaluation with its basic functionalities, the destination position is provided and the GPSR considers only the distance metric and choose to route the generated alert message towards the closest RSU. In

Conclusion

Delivering warning messages in a vehicular environment is a crucial application and tremendous work were carried out in order to guarantee high reliability and low delay for such a task. Previous approaches tackled mainly infrastructure and infrastructure-less based networks rather than dispersed infrastructure-based ones. In this paper, we proposed the HyRSIC scheme, a routing approach based on neighbourhood and infrastructure awareness. Based on such knowledge, a decision making algorithm

Declaration of Competing Interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

References (33)

  • H. Woo et al.

    A hierarchical location service architecture for vanet with aggregated location update

    Comput. Commun.

    (2018)
  • K. Abrougui et al.

    Location-aided gateway advertisement and discovery protocol for vanets

    IEEE Trans. Veh. Technol.

    (2010)
  • S. Arianmehr et al.

    Hybtgr: a hybrid routing protocol based on topological and geographical information in vehicular ad hoc networks

    J. Ambient Intell. Humaniz. Comput.

    (2020)
  • M. Ayaida et al.

    Hhls: a hybrid routing technique for vanets

  • N.M. Balouchzahi et al.

    An efficient infrastructure based service discovery in vehicular networks using p2p structures

    J. Supercomput.

    (2016)
  • I. Basaran et al.

    Fireworks: an intelligent location discovery algorithm for vehicular ad hoc networks

    Wirel. Netw.

    (2018)
  • A. Boukerche et al.

    An efficient hybrid adaptive location-aided gateway advertisement and discovery protocol for heterogeneous wireless and mobile networks

  • S. Boumerdassi et al.

    A flooding-based solution to improve location services in vanets

  • H. Hartenstein et al.

    VANET: Vehicular Applications and Inter-Networking Technologies, Vol. 1

    (2009)
  • B. Karp et al.

    Gpsr: Greedy perimeter stateless routing for wireless networks

  • K. Katsaros et al.

    Effective implementation of location services for vanets in hybrid network infrastructures

  • K. Katsaros et al.

    End-to-end delay bound analysis for location-based routing in hybrid vehicular networks

    IEEE Trans. Veh. Technol.

    (2015)
  • C. Ksouri et al.

    Data gathering for internet of vehicles safety

  • C. Ksouri et al.

    Vanets routing protocols survey: classifications, optimization methods and new trends

  • C. Ksouri et al.

    Towards general internet of vehicles networking: routing protocols survey

  • C. Ksouri et al.

    A unified smart mobility system integrating terrestrial, aerial and marine intelligent vehicles

  • View full text