The objective of this research is to develop advanced millimeter-wave (mmWave) systems that improve the range, efficiency, and speed of next-generation IoT devices. Many current IoT systems struggle with limited range, power constraints, and slow data rates, making large-scale deployment difficult. This work explores the use of dielectric lenses to enhance signal strength, expand coverage, and enable fully-passive operation. To achieve this, a novel mmWave identification (mmID) system that uses a frequency-diverse pixel antenna array and machine learning to achieve precise localization and two-axis orientation detection is developed. To further extend the reading range, a harmonic retrodirective mmID is constructed, which combines a bi-convex dielectric lens with a fully-passive frequency doubler circuit to enable long-range communication with wide spherical coverage. Building on these advancements, two systems are proposed to address challenges in mmWave energy harvesting and ultra-low power, high speed communication: a 3D lens-enabled energy harvester that efficiently captures RF power from multiple angles and a broadbeam gigabit backscatter mmID designed to support high-speed data transmission over long distances.