Semantic-Aware Virtual Movable Antenna Array With Pre-Blocking and Proactive Compensation for Emergency ISAC

In emergency Integrated Sensing and Communication (ISAC) systems, devices operate in infrastructure-limited environments with uncontrollable spatial interference, compounded by semantic redundancy and fixed devices' inflexibility. We propose a unified framework integrating task semantic-aware pre-blocking and proactive compensation, leveraging a key insight: sensing/communication streams are information-independent but semantically overlapping, creating avoidable interference. It first eliminates redundant emissions via beamforming nulling at the source, then uses adjustable nodes' mobility and beamforming agility to synthesize phase-opposite signals, canceling residual fixed-node interference, thus shifting from passive receiver suppression to source awareness. The non-convex joint optimization of positions and beamforming is solved via alternating optimization (AO). Simulations show this semantic-physical co-design, harnessing adjustable nodes to tackle both issues, outperforms conventional methods significantly, reducing interference power by up to 7 dB, suiting dynamic emergency ISAC.