Angularly dispersive links are characterized by frequency-dependent radiation direction. In practice, this property manifests from wide bandwidths, as are expected in the terahertz (THz) regime, and from antenna structures such as the leaky-wave antenna (LWA). To date, angular dispersion has been shown to enable beam steering and path discovery, both are critical for establishing directional THz links. While angular dispersion provides new opportunities for THz communications, it also introduces new security threats. Namely, with angular dispersion, to send a wider band transmission from the transmitter Alice to the receiver Bob necessarily expands the spatial footprint of the transmission, potentially aiding an eavesdropper Eve.

This thesis presents the first security study of THz angularly dispersive links using LWAs via a mix of analytical models and over-the-air experiments. In the first part of the thesis, I consider the threat of a same-distance Eve in the line-of-sight (LoS) scenario and study the unique security properties of angularly dispersive links. I show via both models and experiments that the LWA’s angle-frequency coupling leads to non-uniform secrecy capacity across sub-channels yielding advantages to an eavesdropper at edge frequencies. Yet, because different frequencies emit energy at different angles, the eavesdropper is thwarted from easily intercepting an entire wideband transmission. The experiments diverge from the analytical model in that the model underpredicts the eavesdropper’s advantage at angles smaller than the target user and subsequent asymmetric performance across angles. Nonetheless, both the model and measurements show that increasingly wide bandwidth and correspondingly wide beams have only a modest marginal security penalty.

Next, I study secure coding strategies for angularly dispersive links via two representative secure coding strategies, termed I-SCADL (Independent Secure Coding for Angular Dispersive Links) and J-SCADL (Joint SCADL), with the former must code each frequency channel independently while the latter allows joint coding across frequency channels. I show that, due to angular dispersion, the independently-coded strategy, I-SCADL, results in a notable insecure region expansion both angularly and radially as the transmission band widens, whereas the joint coding strategy, J-SCADL, can effectively alleviate the secrecy degradation with increasing bandwidth as it exploits the a priori known non-uniformity across the frequency channels. The experimental results further demonstrate the advantage of J-SCADL over I-SCADL under beam asymmetry and irregularities as J-SCADL can preserve secrecy when Eve receives strong side lobes only in a subset of frequency channels. Nonetheless, for angularly dispersive links, even with J-SCADL, we find the insecure area expands with bandwidth due to the associated emission angle difference. Yet, we also find that the insecure area growth due to increasing bandwidth is significantly smaller compared to other factors including a wider single-tone beamwidth or a higher secrecy coding rate, suggesting that the security concern for angularly dispersive links under larger bandwidth is minor as long as proper secure coding strategy, such as J-SCADL, is employed.