B-FERL: Blockchain based framework for securing smart vehicles

Highlights

• Securing in-vehicle network components enhances the security of smart vehicles and prevents the potential for successful remote exploitation.

• Integrity assessments of in-vehicle components enhances the credibility of data produced by smart vehicles.

• Proposed security framework satisfies critical automotive functions including vehicular forensics and trust management.

• Security framework facilitates the identification of compromised vehicles in the vehicular network.

Abstract

The ubiquity of connecting technologies in smart vehicles and the incremental automation of its functionalities promise significant benefits, including a significant decline in congestion and road fatalities. However, increasing automation and connectedness broadens the attack surface and heightens the likelihood of a malicious entity successfully executing an attack. In this paper, we propose a Blockchain based Framework for sEcuring smaRt vehicLes (B-FERL). B-FERL uses permissioned blockchain technology to tailor information access to restricted entities in the connected vehicle ecosystem. It also uses a challenge–response data exchange between the vehicles and roadside units to monitor the internal state of the vehicle to identify cases of in-vehicle network compromise. In order to enable authentic and valid communication in the vehicular network, only vehicles with a verifiable record in the blockchain can exchange messages. Through qualitative arguments, we show that B-FERL is resilient to identified attacks. Also, quantitative evaluations in an emulated scenario show that B-FERL ensures a suitable response time and required storage size compatible with realistic scenarios. Finally, we demonstrate how B-FERL achieves various important functions relevant to the automotive ecosystem such as trust management, vehicular forensics and secure vehicular networks.

Introduction

With technological advancements in the automotive industry in recent times, modern vehicles are no longer made up of only mechanical devices but are also an assemblage of complex electronic devices called electronic control units (ECUs) which provide advanced vehicle functionality and facilitate independent decision making. ECUs receive input from sensors and runs computations for their required tasks (Swawibe Ul Alam, 2018). These vehicles are also fitted with an increasing number of sensing and communication technologies to facilitate driving decisions and to be self aware (Chattopadhyay & Lam, 2018). However, the proliferation of these technologies has been found to facilitate the remote exploitation of the vehicle (Greenberg, 2015). Malicious entities could inject malware in ECUs to compromise the internal network of the vehicle (Chattopadhyay & Lam, 2018). The internal network of a vehicle refers to the communications between the multiple ECUs in the vehicle over on-board buses such as the controller area network (CAN) (Han, Weimerskirch, & Shin, 2014). The authors in Greenberg (2015) and Woo, Jo, and Lee (2015) demonstrated the possibility of such remote exploitation on a connected and autonomous vehicle (CAV), which allowed the malicious entity to gain full control of the driving system and bring the vehicle to a halt.

To comprehend the extent to which smart vehicles are vulnerable, we conducted a risk analysis for connected vehicles in Oham, Jurdak and Jha (2019) and identified likely threats and their sources. Furthermore, using the Threat Vulnerability Risk Assessment (TVRA) methodology, we classified identified threats based on their impact on the vehicles and found that compromising one or more of the myriad of ECUs installed in the vehicles poses a considerable threat to the security of smart vehicles and the vehicular network. Vehicular network here refers to communication between smart vehicles and roadside units (RSUs) which are installed and managed by the transport authority. These entities exchange routine and safety messages according to the IEEE802.11p standard (Chen, Li, & Panneerselvam, 2017). By compromising ECUs fitted in a vehicle, a malicious entity could for example, broadcast false information in the network to affect the driving decisions of other vehicles. Therefore, this paper presents a countermeasure to mitigate ECU exploitation by monitoring the state of the in-vehicle network to facilitate the detection of an ECU compromise.

Previous efforts that focus on the security of in-vehicle networks have focused on intrusion and anomaly detection which enables the detection of unauthorized access to in-vehicle network (Akosan et al., 2015, Aloqaily et al., 2019, CUBE, 2018, Nilsson and Larson, 2008, Oguma et al., 2008, Salem et al., 2019), and the identification of deviations from acceptable vehicle behavior (Wasicek & Weimerskirch, 2014). Several challenges however persist. First, proposed security solutions are based on a centralized design which relies on a Master ECU that is responsible for ensuring valid communications between in-vehicle ECUs (Nilsson and Larson, 2008, Oguma et al., 2008, Salem et al., 2019). However, these solutions are vulnerable to a single point of failure attack where an attacker’s aim is to compromise the centralized security design. Furthermore, if the Master ECU is either compromised or faulty, the attacker could easily execute actions that undermine the security of the in-vehicle network. Also, actions that constitute intrusion have not been properly defined and intrusion detection is towards the request and provision of requisite services for smart vehicles (Aloqaily et al., 2019) and not towards secure inter-vehicular communications. In addition, efforts that focus on intrusion detection by comparing ECU firmware versions (Akosan et al., 2015, CUBE, 2018, Nilsson and Larson, 2008) are also vulnerable to a single point of exploitation whereby the previous version which is centrally stored could be altered. These works (Akosan et al., 2015, CUBE, 2018) also rely on the vehicle manufacturer to ultimately verify the state of ECUs. However, vehicle manufacturers could be motivated to execute malicious actions for their benefits such as to evade liability (Oham, Jurdak, Kanhere, Dorri and Jha, 2018). Therefore, decentralization of the ECU state verification among entities in the vehicular ecosystem is desirable for the security of smart vehicles. Finally, the solution proposed in Wasicek and Weimerskirch (2014) which focuses on observing deviations from acceptable behavior utilized data generated from a subset of ECUs. However, this present a data reliability challenge when an ECU not included in the ECU subset is compromised.

We argue in this paper that Blockchain (BC) (Nakamoto, 2008) technology has the potential to address the aforementioned challenges including centralization, availability and data reliability.

BC is an immutable and distributed ledger technology that provides verifiable record of transactions in the form of an interconnected series of data blocks. BC was initially introduced as a security solution for the Bitcoin cryptocurrency system (Nakamoto, 2008) but is now widely utilized for non-monetary applications including tackling the dissemination of fake news (Chena, Srivastava, Parizic, Aloqailyd, & Ridhawie, 2020), auditing of public cloud storage (Lia, Wu, Jiang, & Srikanthanb, 2020), and task scheduling optimization (Baniata, Anaqreh, & Kertesza, 2021). Furthermore, in Berdika, Otoum, Schmidta, Portera, and Jararweha (2020), the authors discuss more non-monetary BC applications.

BC can be public or permissioned (Oham, Jurdak, Kanhere et al., 2018) to differentiate user capabilities including who has the right to participate in the BC network. Compared to identified intrusion (CUBE, 2018, Nilsson and Larson, 2008, Salem et al., 2019), and anomaly detection (Swawibe Ul Alam, 2018) solutions, BC replaces centralization with a trustless consensus which when applied to our context can ensure that no single entity can assume full control of verifying the state of ECUs in a smart vehicle and could facilitate the identification of rogue actions executed by vehicle manufacturers (Oham, Jurdak, Kanhere et al., 2018). Furthermore, the decentralized consensus provided by BC is well-suited for securing the internal network of smart vehicles by keeping track of historical operations executed on the vehicle’s ECUs such as firmware updates, thus easily identifying any change to the ECU and entities responsible for that change. Finally, the distributed structure of BC provides robustness to a single point of failure.

Having identified the limitations of existing works, we propose a Blockchain based Framework for sEcuring smaRt vehicLes (B-FERL). B-FERL is an apposite countermeasure for in-vehicle network security that exposes threats in smart vehicles by ascertaining the state of the vehicle’s internal controls. Also, given that data modification depicts a successful attempt to alter the state of an ECU, B-FERL also suffices as a data reliability solution that ensures that a vehicle’s data is trustworthy. We utilize a permissioned BC to allow only trusted entities manage the record of vehicles in the BC network. This means that state changes of an ECU are summarized, stored and managed distributedly in the BC.

The key contributions of this paper are summarized as follows:

(1) We present B-FERL; a decentralized security framework for in-vehicle networks. B-FERL ascertains the integrity of in-vehicle ECUs and highlights the existence of threats in a smart vehicle. To achieve this, we define a two-tier blockchain-based architecture, which introduces an initialization operation used to create record vehicles for authentication purposes and a challenge–response mechanism where the integrity of a vehicle’s internal network is queried when it connects to an RSU to ensure its security.

(2) We conduct a qualitative evaluation of B-FERL to evaluate its resilience to identified attacks. We also conduct a comparative evaluation with existing approaches and highlight the practical benefits of B-FERL. Finally, we characterize the performance of B-FERL via simulations using the CORE simulator against key performance measures such as the time and storage overheads for smart vehicles and RSUs.

(3) Our proposal is tailored to meet the integrity requirement for securing smart vehicles and the availability requirement for securing vehicular networks and we provide succinct discussion on the applicability of our proposal to achieve various critical automotive functions such as vehicular forensics, secure vehicular communication and trust management.

This paper is an extension of our preliminary ideas presented in Oham, Jurdak, Jha et al. (2019). Here, we present a security framework for detecting when an in-vehicle network compromise occurs and provide evidence that reflect actions on ECUs in a vehicle. Also, we present evaluations to demonstrate the efficacy of B-FERL.

The rest of the paper is structured as follows. In Section 2, we discuss related works and present an overview of our proposed framework in Section 3, where we also describe our system, network and threat model. Section 4 describes the details of our proposed framework. In Section 5, we discuss results of the performance evaluation. Section 6 present discussions on the potential use cases of B-FERL, comparative evaluation with closely related works, and we conclude the paper in Section 7.