Securing inductively-coupled communication

Abstract

Abstract-Communication over inductively-coupled links is becoming prevalent in service delivery for medical, financial, and physical security applications and so there is a growing need to prevent eavesdropping. This paper presents circuittheoretic and communication-theoretic models of inductivelycoupled communication systems. Due to coupling, the presence of an eavesdropper detunes the transfer function between the legitimate users. It is shown this detuning can be detected to reveal the presence of the eavesdropper. Further, if capacityapproaching codes are employed, neither the eavesdropper nor the legitimate receiver are able to reconstruct the transmitted message with low error probability, effectively destroying the message. Building on this insight, a coding-based secure communication protocol for inductively-coupled communication, inspired by quantum key distribution, is developed. The notion of security is defined operationally in terms of probabilities rather than through traditional notions of equivocation. Separately, there has been interest in understanding the privacy and security aspects of inductively-coupled systems. With the emergence of RFID in customer-facing service delivery rather than simply back-office logistical operations [10], there is a growing business need for securing such communications. This is particularly important when transmitted information involves medical data, financial transactions, or physical access control signals. One particular kind of privacy attack is eavesdropping, where a third party antenna is used to couple into the communication channel and capture some information; the literature in this area (including a recount of effective attacks in practice) is well-summarized in [11]; one recently proposed approach to protect against eavesdroppers is active jamming [12].Contrary to traditional studies of wireless communication that use far-field models of electromagnetics, inductive coupling is a near-field effect, i.e. the distance between the participating antennas is comparable to (or smaller than) the

Near-Field Passive RFID Communication: Channel Model and Code Design

“…where I(X; Y ) denotes the mutual information between X and Y and E[X] the expectation of X with respect to the input probability distribution f X (x). Now, the capacity of H L,T (in bits per symbol) can be obtained by maximizing the fraction in (7) over all input probability distributions f X (x). Note that since the channel is memoryless, it is sufficient to consider only a single use of the channel, i.e., not sequences of length N as in [15,Eq.…”

“…, L − 1. Thus, we may substitute f X (L) = 1 − L−1 x=1 f X (x) into (7) and then compute the partial derivatives with respect to f X (x), x = 1, . .…”

“…The partial derivative of the normalized mutual informationĨ(X; Y ) in (7) with respect to f X (x), x = 1, . .…”

“…For details, we refer the interested reader to [6]. In [7], a coding-based secure communication protocol for inductively coupled communication, inspired by quantum key distribution, was recently proposed.…”

On non‐destructive detection of hidden passive radio‐frequency eavesdroppers

“…The main rationale of these approaches is to trace unique features of RF eavesdroppers in the wireless channel. It showed that the channel state between the legitimate receiver and transmitter changes when an eavesdropper presents due to the near-field inductively coupling [4]. As the near-field coupling is not universal in eavesdropping implementations, a more commonly used feature for detecting eavesdroppers in wireless channels is the local oscillator (LO) leakage level [5][6].…”