⚡ electrical engineering

Stepped frequency comb advanced sub-THz bandwidth spectrum sensing

This paper presents a photonic spectrum sensing scheme based on a stepped frequency comb that synergistically combines parallel channelization and serial sweeping to achieve real-time, high-resolution sub-THz bandwidth sensing (0.3–150 GHz) using low-bandwidth receivers and low-sampling-rate digitizers.

Original authors: Zhiqiang Fan, Yang Yang, Jun Su, Yunxiang Wang, Shuangjin Shi, Qi Qiu

Published 2026-09-01

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Zhiqiang Fan, Yang Yang, Jun Su, Yunxiang Wang, Shuangjin Shi, Qi Qiu

Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer

The air around us is filled with invisible waves, a constant hum of radio signals carrying everything from your morning news to the data that guides autonomous vehicles. As our technology grows more demanding, pushing into higher and faster frequencies, this invisible landscape is becoming dangerously crowded. To manage this traffic and prevent signals from crashing into one another, engineers need a way to listen to the entire spectrum in real time, spotting every transmission as it happens. The challenge is that the frequencies we now need to monitor are so high—reaching into the sub-terahertz range—that the electronic tools we have traditionally used simply cannot keep up. They are too slow to catch the fleeting signals or too narrow to see the whole picture at once.

A team of researchers at the University of Electronic Science and Technology of China has developed a new way to solve this problem by using light instead of electricity to do the listening. They created a system that can scan a massive slice of the radio spectrum, from 0.3 gigahertz all the way up to 150 gigahertz, with a speed and clarity that previous methods could not achieve. By combining the speed of light with a clever stepping mechanism, they built a sensor that can identify signals across a bandwidth of 149.7 gigahertz in just 35.5 microseconds. This means the system can see the entire spectrum almost instantly, identifying not just where a signal is, but exactly what frequency it is using with a precision of 12 megahertz.

The core of their invention is a device that generates a "stepped frequency comb." Imagine a ladder where each rung represents a specific frequency, but instead of standing still, the ladder moves up and down in tiny, precise steps. In this system, the researchers use light to create three parallel channels, each acting as a separate ladder. Inside each channel, the light steps through 250 different frequencies, moving in increments of 200 megahertz. This creates a total of 750 distinct steps across the entire system. When a radio signal enters the device, it is converted into light and mixed with these stepping ladders. The system then measures how the signal interacts with each step, effectively translating the high-speed radio wave into a pattern that can be read by standard, slower electronic sensors.

This approach solves a long-standing dilemma in sensor design. Previous methods had to choose between speed and detail. Some systems could scan a wide area quickly but missed the fine details of the signals, while others could see the details clearly but took too long to scan the whole range. The new system manages to do both simultaneously. It achieves a measurement bandwidth of 149.7 gigahertz, which is nearly three times wider than the best previous photonic systems, while maintaining a frequency resolution of 12 megahertz. This level of detail allows the system to distinguish between signals that are very close together in frequency, a capability essential for untangling the complex web of modern communications.

To prove their system works, the researchers tested it with various types of signals. They successfully identified single tones at frequencies like 10.5, 60.5, and 130.5 gigahertz, mapping them to specific time slots with an error margin of less than 2.95 megahertz. They also tested the system's ability to track signals that change frequency rapidly, such as those used in radar or agile communications. The system reconstructed the time-frequency patterns of these signals with high fidelity, even when the signals were changing shape or jumping between different bands. It could also detect multiple signals happening at once, separating them cleanly even when they were packed closely together.

The beauty of this design lies in its efficiency. Because the system uses light to do the heavy lifting of scanning, it does not require the expensive, high-speed electronic components that usually limit these kinds of sensors. The researchers were able to use relatively simple receivers with a bandwidth of only 200 megahertz and standard analog-to-digital converters running at 1.25 gigasamples per second. This makes the system much more practical and potentially cheaper to build than previous solutions that required specialized, high-end electronics for every part of the spectrum. The entire setup fits into a compact architecture that could be scaled up or integrated into future photonic chips.

This work represents a significant step forward in our ability to manage the electromagnetic spectrum. By demonstrating that it is possible to monitor a sub-terahertz bandwidth with high speed and high resolution using low-bandwidth receivers, the researchers have opened a new path for spectrum sensing. Their system provides a clear view of a complex electromagnetic environment, offering a tool that could help prevent signal conflicts in future 6G networks, improve radar systems, and support the growing number of wireless devices that rely on these high-frequency bands. The ability to see the whole picture in real time, without sacrificing detail, changes the game for how we manage the invisible waves that power our modern world.

Technical Summary: Stepped Frequency Comb Advanced Sub-THz Bandwidth Spectrum Sensing

Problem Statement
The rapid evolution of next-generation communication and radar systems, operating increasingly in the sub-THz and THz regimes, has created a highly congested and dynamic electromagnetic environment. Effective real-time spectrum sensing in this regime requires simultaneous ultrawide bandwidth (extending toward sub-THz), high frequency resolution, and high temporal resolution. However, existing electronic approaches are fundamentally limited by component bandwidth, typically struggling to exceed 40 GHz. Established photonic techniques face inherent trade-offs:

Frequency-to-Power Mapping (FTPM): Offers speed but lacks simultaneous multi-frequency measurement and suffers from limited Figure of Merit (FOM) due to amplitude fluctuations.
Dispersion-based Frequency-to-Time Mapping (DFTTM): Provides wide coverage but is limited by dispersion to coarse frequency resolutions (typically >100 MHz) and requires high-bandwidth receivers.
Parallel Channelization: Achieves high FOM but demands a large number of receiver units, increasing hardware complexity.
Serial Sweeping: Reduces receiver requirements but sacrifices temporal resolution due to sequential interrogation and is often constrained by the tuning range of the source (typically <51 GHz).

Consequently, a system capable of delivering high frequency resolution, high temporal resolution, and ultrawide bandwidth simultaneously remains elusive.

Methodology
The authors propose and demonstrate a photonic spectrum sensing scheme based on a Stepped Frequency Comb (SFC). This approach synergistically integrates the ultra-broadband advantage of parallel channelization with the hardware efficiency of serial sweeping.

System Architecture: The system utilizes two continuous-wave (CW) lasers. One path generates the Signal Under Test (SUT) in the optical domain via a Mach-Zehnder Modulator (MZM). The second path generates the SFC.
SFC Generation: The SFC is generated by combining a stepped frequency module (using a recirculating frequency-shifted fiber loop with an acousto-optic modulator) and an electro-optic comb module. This creates an optical signal with $N$ sub-channels, where each channel contains $M$ frequency steps.
Operation Principle: The SFC is demultiplexed into $N$ sub-channels. Within each sub-channel, the stepped-frequency local optical signal is coupled with the optical SUT. Through parallel channelization, the SUT is mapped to specific time bins in the electrical domain via serial sweeping.
Detection: The system employs low-bandwidth (200 MHz) image-reject receiver (IR-Rx) arrays and low-sampling-rate (1.25 GSa/s) Analog-to-Digital Converters (ADCs). The SUT frequency is identified by determining the sub-channel and time bin of the mapped signal, calculated as $f_{SUT} = (n-1) \times f_{ch} + m \times f_{b} - f_{Rx}$ , where $n$ is the channel, $m$ is the step, and $f_{Rx}$ is the detected frequency.

Key Contributions

Novel Architecture: The introduction of the SFC architecture, which overcomes the fundamental trade-off between sweep rate and resolution found in conventional swept-frequency approaches.
Hardware Efficiency: The system achieves ultrawideband sensing using only three receiver units and low-sampling-rate ADCs, avoiding the need for high-speed electronic components typically required for such bandwidths.
Scalability: The design is compatible with photonic integration and can be scaled by increasing the number of comb lines ( $N$ ) or frequency steps ( $M$ ) to extend bandwidth further into the THz regime.

Experimental Results
The authors experimentally validated the system using an SFC comprising three sub-channels ( $N=3$ ) with 250 frequency steps ( $M=250$ ) per channel at 200 MHz intervals ( $f_b$ ).

Measurement Bandwidth: The system successfully performed real-time spectrum sensing across 0.3 to 150 GHz, achieving a total measurement bandwidth of 149.7 GHz.
Resolution and Accuracy: The system achieved a frequency resolution of 12 MHz and a root mean square (RMS) frequency measurement error below 2.95 MHz across 100 repeated measurements.
Temporal Resolution: The temporal resolution was approximately 35.5 μs.
Figure of Merit (FOM): The system achieved a large FOM of 12,475 (defined as measurement bandwidth divided by frequency resolution).
Signal Reconstruction: The system demonstrated high-fidelity reconstruction of various signal types, including frequency-agile signals (triangular, "UESTC" text format, and Costas 7 waveforms), multi-tone signals, and multiband frequency-agile signals.

Significance
The paper claims that this work represents the first photonic real-time spectrum sensing demonstration extending the measurement bandwidth to the sub-THz regime (149.7 GHz) while maintaining high resolution and temporal resolution. By balancing the requirements for receiver unit count, receiver bandwidth, and ADC sampling rate against performance metrics, the proposed SFC-based scheme offers a new framework for ultrawideband spectrum sensing. The authors suggest this approach provides a low-cost route toward compact, real-time spectrum management systems capable of operating in complex electromagnetic environments, potentially supporting future 6G and integrated sensing and communication applications. The work highlights that the balance between resolution, speed, and hardware requirements can be tuned by adjusting SFC parameters, and that recent advances in photonic integration (e.g., on-chip AOMs and modulators) support the feasibility of realizing integrated versions of this system.

More like this