Benign false-data injection as a moving-target defense to secure mobile wireless communications

Abstract

The latest advances in the Internet of Vehicle (IoV) networks, and Software Defined Radio (SDR) enabled spectrum sharing between anonymous users. Such enablement raised many concerns in regards to the system’s security and privacy. Therefore, there is a desperate need for transformative solutions to ensure wireless communication security against eavesdropping attacks.

This paper presents a novel Moving-target Defense (MtD) to enhance the channel secrecy capacity in a Decode-and- Forward (DF) dual-phase large network containing K relays and source nodes with multi-antennas operating on different frequencies. Our MtD approach enables multidimensional spatiotemporal diversification for the user’s traffic in cooperative wireless transmission, to obfuscate signal transmission-patterns and data, across the entire spectrum available. In time, we obfuscate the transmitted data by employing real-time shuffling between real and fake data. In space, we enforce real-time hopping between multiple frequencies to evade signal tracing. We examine the ergodic channel secrecy capacity considering two behavioral patterns; cooperative and uncooperative untrustworthy-relays.

Simulation results showed that, for a powerful malicious user with multiple access points, and no pre-knowledge of the diversification patterns used by the system, it is very hard to eavesdrop a meaningful portion of the signal or the data stream.

Introduction

Currently, Wireless communication technologies struggle to cope with the increasing service demands. Several concepts were applied to optimize resource utilization and network performance for different wireless systems operating in urban environments, and smart cities [1], [2].

To this end, cooperative wireless networks were underscored as one of the most promising and efficient solutions to improve channel spectrum utilization in high mobility applications like Vehicular Ad hoc Networks (VANETs) [3], [4].

VANETs evolved over the past decade developing what we note as The Internet of Vehicles (IoV).

IoV is a distributed network that supports data generation, processing, and exchange between connected vehicles and the internet. IoV network enables moving vehicles to communicate with other vehicles, roadside infrastructure, and any other fleet management systems in real-time. Unfortunately, IoV reliance on wireless communication as a base for its infrastructure made it vulnerable to devastating eavesdropping attacks. The heterogeneous nature of the communicating parties, communication technologies, and relying patterns, opened the doors for eavesdroppers to play the role of Man in the Middle and intercept or manipulate the exchanged messages. Researchers investigated wireless communication security issues in cooperative relying environments for years.

Wyner’s [5], [6] introduced the critical notion of the secrecy capacity as a way to evaluate security levels in such networks. Wyner defined secrecy capacity as the difference between the main and wiretap channels.The main channel refers to the legitimate users’ channel while the wiretap channel refers to the eavesdropper’s channel. For years, researches investigated innovative mechanisms aiming to improve the channel secrecy capacity of wireless communication networks [7].

They noted the potential of exploiting the diversified nature of such networks and its constructing components, to enhance the attained level of security. Within this context, approaches like Moving-target defense (MtD) was introduced as a game-changer. Researchers noted that successful realization of MtD could be very effective in mitigating sophisticated eavesdropping attacks [1], [2], [8], [9], [10].

MtD acts on manipulating the network characteristics across multiple dimensions to asymmetrically increase attacker uncertainty of the network behavioral patterns. The induced manipulations reduce the attacker’s window of opportunities and increase the costs of probing and attacks [11].

There is no standardized description or a specific metrics that can help evaluating the effectiveness of MtD in securing wireless networks. In this paper, we demonstrated that channel secrecy capacity can be used as an effective metric to evaluate the efficiency of MtD mechanism securing IoV networks.

In [12], authors assumes the network encompasses both trustworthy and untrustworthy relays. During the analysis, we consider two classes of untrustworthy relays. The first is uncooperative malicious relays, where every relay tries to intercept the message independently. While the second is maliciously cooperative relays, where relays cooperate towards aggregating the received messages to reconstruct as much as possible from the original data.

This paper presents a runtime diversification mechanism as an MtD approach for two-hop DF-cooperative relays based on the DBF technique. Our approach efficiently ensures user’s security by dynamically changing the runtime behavior of the system. The proposed mechanism relies on benign employment of false-data injection to confuse the attacker. Such a technique disorients attackers from compromising user’s information.

The presented approach senses the whole available antennas/frequencies according to the running channel characteristics. Therefore, Software Defined Radio (SDR) technology is presented as a paradigm solution to efficiently re-program/reconfigure the attached SDR-antenna.

The Proposed spatiotemporal multidimensional manipulations randomly re-assign legitimate user’s data to the available frequency bands in a dual-hop cooperative network through a dynamic real-time transmission algorithm. In time, we confuse the attacker by utilizing a real-time shuffler to alternate between real and fake data. In space, we diffuse the transmitted data across all the available frequencies.

The main contribution of this paper can be summarized as follow:

• Presenting a novel spatiotemporal Moving-target Defense (MtD) mechanism against eavesdropping in wireless dual-hop cooperative networks based on the DBF mechanism.

• Presenting security evaluation mechanism based on the induced level of confusion, followed by a closed-form derivation with respect to the channel secrecy capacity.

• Calculating the channel secrecy capacity and intercept probability in case of location-based multi-casting considering two scenarios:

  • The existence of independent uncooperative untrustworthy relays.

  • The existence of cooperative untrustworthy relays aggressively attacking the message.

The remaining part of the paper is organized as follows: Section 2 describes the proposed cooperative wireless communication system model. Section 3 presents a detailed mathematical security assessment for both uncooperative and cooperative untrustworthy relays. Section 4 illustrates the security evaluation model using our randomization mechanism procedure. Section 5 illustrates the intercept probability mathematical calculation for both types of relays. Section 6 presents numerical results to confirm the analytical derivations and provide insight into the system performance. Finally, Section 7 presents the conclusion of this work.