History-Aware Trajectory k-Anonymization Using an FPGA-Based Hardware Accelerator for Real-Time Location Services

Prior FPGA-based trajectory anonymization relied exclusively on shortest-path computations, failing to capture realistic travel behavior and reducing data utility. This paper introduces a history-aware trajectory k-anonymization methodology with an FPGA architecture integrating parallel history-based searches, custom fixed-point counting, and shortest-path finding. Our system prioritizes behaviorally common routes, achieving real-time throughput exceeding 6,000 records/s, improving data retention by up to 1.2%, and better preserving major arterial roads.