Janusz Murakowski

Time-reversal (TR) signal processing is a promising approach for mitigating multipath propagation effects in wideband wireless systems by leveraging channel reciprocity to pre-compensate for propagation-induced symbol delays, phase misalignment, and interference at the transmitter. This enables temporal compression at the receiver, resulting in the constructive combination of multipath components, reduced inter-symbol interference (ISI), and improved spatial focusing, while shifting the signal processing burden away from the receiver. However, the practical implementation of TR in distributed multi-antenna systems is fundamentally limited by the need for tight phase coherence and frequency synchronization across transmitters, which is difficult to achieve with conventional RF architectures that rely on independent oscillators. In this work, we present an RF-photonic distribution architecture that enables distributed coherent multi-transmitter TR operation by driving up to 10 transmit antennas from a single optical source, thereby ensuring phase coherence and frequency synchronization across all paths. Using experimentally measured channel states over the 6G carrier frequency of 20 GHz and an instantaneous bandwidth of 600 MHz, we demonstrate time-reversal transmission with strong temporal compression and spatial focusing. These results highlight the critical role of coherence in distributed TR systems and establish RF-photonic architectures as a scalable solution for multi-antenna time-reversal communication in emerging wireless networks.

Microring resonators are fundamental components in modern silicon photonic integrated circuits. Here, we present a reconfigurable silicon microring resonator that leverages a delayed back-coupling mechanism, turning intrinsic backscattering into a versatile design tool. By employing dual injections and tuning the phase between each injection arm, we lift modal degeneracy and achieve independent control of clockwise (CW) and counterclockwise (CCW) modes. This approach enables selective mode suppression, tunable spectral shifts, and bidirectional resonance control. Experimental results closely match simulations, demonstrating robust phase-controlled resonance engineering for advanced spectral signal processing.

This paper demonstrates concurrent spatial and spectral signal classification using two photonic integrated circuits (PICs) implemented in a 30-element Ka-band (26.5-40 GHz) antenna array. The received signals are coherently up-converted to the optical-domain using lithium niobate phase modulators. The up-converted RF signal is beamformed in a star-coupler PIC to instantaneously image angle-of-arrival to a single output channel that is amplified and input to an arrayed waveguide grating (AWG) PIC, which instantaneously channelizes the RF frequency into increments of approximately 1 GHz. A switching scheme is used to calibrate the AWG as well as alternate between frequency and beam location measurement. This work implements these devices in a radiometer that can localize RF beams both spatially and spectrally with minimal latency and computational requirements. Future applications will include advanced wireless communications and sensing.

For many years, microwave radiometers have contributed to the accuracy of atmospheric models which are used for weather forecasting and climate-change monitoring. In this work, a new microwave radiometer architecture is proposed that utilizes the advantages of photonics to measure a wide swath of microwave spectrum from 50 to 70 GHz. An optical-fiber-based arrayed waveguide grating is used to optically process the spectral content from a V-band radio front end. The result is a radiometer that measures the microwave spectrum near the 60 GHz oxygen absorption peak. The instrument described in this work is shown to produce brightness temperature spectra that fit with expected trends for given weather conditions and has a noise equivalent delta temperature (NEDT) of 0.68 K for an integration time of 8.8 seconds. The microwave-photonic approach to radiometry offers the capability to measure wider spectral bands than accessible using conventional means, and to probe currently unobserved frequencies, within a reasonable form-factor. Further development of the approach outlined in this work will lead to a space-worthy microwave radiometer for improved spectral coverage of atmospheric measurements.

Most of the energy exchange in the atmosphere occurs in the atmospheric boundary layer (ABL) with respect to solar heating and evaporation. Remote sensing in the ABL from space is challenging because of its proximity to the surface and potential sharp gradients in air properties. Passive microwave instruments can be a critical component to a ABL observing system since they provide thermodynamic information in both cloudy and clear air, including cloud and precipitation properties, and surface flux information such as temperature and wind speed. However, conventional passive microwave systems fall well short of being optimized for near surface sensing due to limited number of spectral channels and coarse spectral resolution covering only a small portion of the spectrum of interest for ABL sensing.The near surface thermodynamic structure information is encoded on the microwave spectrum between and on the shoulders of the water vapor absorption and oxygen lines near 183 GHz and 118 GHz respectively. The ultra-wideband photonic spectro-radiometer instrument is funded by NASA ESTO under ACT-20 program to combine low-noise wideband InP RF technology with a novel photonic integrated circuit (PIC) design for obtaining large bandwidth (>50 GHz) with enhanced channel resolution (

The Atmospheric Boundary Layer (ABL) is the portion of the troposphere that is directly influenced by the Earth’s surface and responds to combined action of mechanical and thermal forcing. Most of the energy exchange with respect to solar heating and evaporation that drive the atmosphere and the ocean occur within the ABL, yet it is one of the most poorly observed and modeled regions of the atmosphere. Conventional passive microwave systems fall well short of being optimized for near surface sensing due to limited number of spectral channels and coarse spectral resolution covering only a small portion of the spectrum of interest for ABL sensing. The so called “window regions” of the microwave spectrum between and on the shoulders of the strong oxygen and water vapor absorption lines carry the information on the near surface thermodynamic structure in the boundary layer. Sampling these regions requires new spectrometers capable of resolving >50 GHz spectral regions at modest spectral resolution (~1GHz). The ultra-wideband photonic spectro-radiometer instrument is funded by NASA ESTO under ACT-20 program to combine low-noise wideband RF technology with a novel photonic integrated circuit (PIC) design for obtaining large bandwidth (>50 GHz) with enhanced channel resolution (

We present a method for the recovery of two-dimensional angle-of-arrival information by beamspace processing using only a linear array of elements in the beam-forming processor. The method exploits the presence of grating lobes in beam patterns formed when using phased arrays with regularly spaced antenna elements. Periodic distribution of grating lobes in the reciprocal space allows, for some arrays, ‘unrolling’ a two-dimensional beamspace pattern into a linear distribution of beams, which is amenable to recovery using a linear array of elements in the beamspace processor. We develop mathematical underpinning of the approach and analyze several explicit examples of such array-beamspace (AB) mapping.

In this work, angle-of-arrival (AoA) detection using a two-dimensional (2D) antenna array coupled to a planar photonic integrated circuit (PIC) is demonstrated. To detect AoA in azimuth and elevation (az-el) the received signal from each antenna element, in a 2D array, is up converted to the optical domain using an electro-optic modulator and transmitted to a PIC via optical fiber. To preserve the two-dimensional phase information captured by the array, a special array-beam space (AB) mapping is used to route the fibers to a linear array of input waveguides on the PIC. This paper presents the steps to generate such a mapping for the antenna elements, as well as an equivalent mapping for the array's beam space. We also present experimental results for beam detection using a 30-element hexagonally spaced array. The PIC-based approach promises to reduce system size and reduce latency for AoA detection.

We describe an RF-photonic system architecture to measure microwave spectrum over a wide bandwidth with high resolution. The approach relies on the up-conversion of the electronic signal to the optical domain using a high-speed electro-optic modulator (EOM), and an arrayed-waveguide (AWG) grating for spectral analysis. The modulator is implemented in lithium niobate whereas the AWG takes the form of a photonic integrated circuit fabricated in a silicon-on-insulator material platform. We present experimental results showing high-frequency response, in excess of 200 GHz, for the modulator, and broadband response of the AWG with the free spectral range of 81 GHz. The system is suitable for the measurement of atmospheric radiation between the oxygen and water vapor absorption lines for the monitoring of the planetary boundary layer (PBL) to aid in weather prediction and the modeling of climate dynamics.

This paper presents a class of phased-array systems that function as video-rate imagers in the millimeter-wave (mmW) and sub-THz bands. While the systems presented operate in the Ka-band (35 GHz) and W-band (77 and 86 GHz), the approach is scalable to the THz regime. Their operation is based on the upconversion of incident mmW and sub-THz signals to the optical domain using high-speed electro-optic modulators (EOMs) that are connected to each antenna element in a phased-array antenna. The output optical fiber from each EOM is relayed to a fiber bundle, or optical fiber array, from which the upconverted mmW/sub-THz signals are launched into free space. Because the upconversion preserves both the temporal and spatial coherence, through a spatial phase-control loop (SPCL), the launched sideband signals re-form the beamspace of the incident mmW or sub-THz signals, but in the optical domain. At this point, a lens performs a 2D spatial Fourier transform, to produce a real-time image of the mmW or sub-THz signals from the environment on a short-wave infrared (SWIR) camera, which renders the scene at video rates. The fundamental operating principles of these systems are presented, along with the historical progress in their development, and experimental demonstrations.