Abstract
Efficient chip-scale interconnects are critical for modern microelectronic–photonic systems, enabling high-bandwidth utilisation and ultra-low-latency processing. Conventional wired links suffer from high resistivity and latency, while radio-frequency and millimetre-wave wireless solutions face limitations such as bandwidth congestion, interference and power inefficiency. Terahertz (THz) plasmonic communication, utilising surface-plasmon polaritons (SPPs), is shown to provide broad bandwidth and high data rates for wireless network-on-chip (WiNoC) links, while remaining compatible with nanophotonic architectures. A novel Binary Field-Driven Meta-Routing Method is proposed, supported by a semi-analytical framework that models the interaction between graphene’s tunable electromagnetic properties and THz plasmonic phenomena. Graphene impedance modulation is exploited to dynamically couple localized surface-plasmon resonances (LSPRs) and guide them across a meta-network, enabling controlled beam steering within chip-scale architectures. Analytical conductivity models are combined with coupled-mode theory and algorithmic control to predict and configure LSPR-based beam steering in graphene metasurfaces. Four reconfigurable graphene meta-pixel antenna configurations — Y-MetaRouter, MetaSwitcher, Penta-MetaEmitter and CP-MetaCore — are designed and analysed; they enable unidirectional radiation, bi-directional meta-steering, frequency-driven multidirectional transitions and circular polarization, respectively. Real-time beam steering is enabled via chemical-potential modulation, thereby forming configurable LSPR pathways and creating virtual SPP channels. A theoretical formulation of the Coupled-Mode Theory of Field-Driven LSPR Meta-Networks is developed to model the current distribution of virtual SPPs and path-dependent LSPR coupling for prediction of far-field characteristics. Theoretical results show excellent agreement with full-wave numerical simulations. A point-to-point meta-wireless link is analysed by both theoretical and numerical methods, thereby demonstrating scalability for low-latency, high-performance THz communication in WiNoC and nanophotonic platforms. System-level metrics — such as link-budget, data-rate and reconfiguration energy — are estimated to validate feasibility for applications including chiplet communication, intra-core data transfer, heterogeneous computing, and compact transceivers in space-constrained environments.