First-hand information for everyone
This paper reviews recent progress in integrated microwave photonics, a field that uses light to generate, carry and process very high-frequency radio signals. The main message is that researchers are moving from individual devices — like modulators and filters — to complete chip- and module-level systems that can do real signal processing, radar functions, and wireless links on a small scale.
The authors summarize how different material platforms contribute different strengths. Silicon and silicon nitride are good for dense routing, low-loss delay lines, and programmable filters but they do not make light on their own. Indium phosphide can make lasers, amplifiers and high-speed detectors, but it tends to have higher passive loss. Thin-film lithium niobate gives very linear and wideband electro‑optic conversion, which is useful for clean analog links. Emerging ferroelectric materials promise strong electro‑optic effects but are still immature for reliable devices.
The review highlights a shift in demonstrations. Instead of testing single building blocks, teams are now showing integrated systems: signal processing engines, real-time spectrum sensors, full‑spectrum wireless links, fiber‑to‑wireless conversion, silicon beamforming networks, integrated radar, and photonic processors that include sources and detectors on the same chip. This system focus changes how progress is judged: engineers now care as much about packaging, stability, and calibration as they do about raw bandwidth or single‑device efficiency.
The paper also explains why this matters. Putting microwave photonic functions on chips can make devices smaller, more stable against the environment, less power hungry, and easier to scale. That could help radar, wireless communications (including future 6G bands), spectrum monitoring, and analog photonic computing become more practical in real applications.
Important caveats are discussed. No single material solves every need, so heterogeneous integration — combining several materials on one module — is a common path but brings its own complexity. Several platforms lack native light sources and require hybrid assembly. Thin‑film lithium niobate is sensitive to temperature and its packaging options are still maturing. Ferroelectric materials and some integration routes are not yet production‑ready. At the system level, unresolved issues include microwave packaging, long‑term calibration stability, and how well analog links perform in real environments.
Overall, the paper maps current capabilities and tradeoffs rather than claiming finished products. It frames the near‑term challenge as integrating complementary materials and solving practical engineering problems so that the promising demonstrations translate into reliable, application‑ready systems.