I. Overview

Existing wireless systems lack the capability to support multiple functions simultaneously on a single device. More specifically, current radar systems do not have the ability to communicate with other nodes regarding their locations while in operation. As shown in Figure 1, for example, when a typical radar or communication system transmits over a 10-MHz channel, the receiver cannot operate within an adjacent bandwidth of approximately 1000 MHz due to interference. This not only severely reduces spectral efficiency but also prevents the receiver from detecting signals in that frequency range, as illustrated in Figure 2(a). The failure of ESM algorithms diminishes their ability to promptly identify intermittent signals, which can have catastrophic consequences in certain applications, such as the detection of improvised explosive device (IED) triggers.

Most existing systems allocate transmission and reception to separate time slots and/or frequency bands, which limits their flexibility in congested spectrum environments. Moreover, communication interruptions complicate the design of the medium access control (MAC) layer and can trigger timeouts at the network layer. In addition, ESM applications must be tailored to account for the intermittent nature of radar or communication operations, introducing exploitable vulnerabilities into the system. Another approach is to mitigate interference by physically isolating the two systems; however, this becomes impractical for small‑form‑factor devices, particularly in handheld applications and airborne platforms.

Figure 1: The application space of multifunctional arrays, highlighting the competing requirements for simultaneously performing radar, communications, and electronic support measures (ESM).

In contrast, next-generation platforms employ simultaneous transmit-and-receive (STAR) technology—also known as in-band full-duplex—which can concurrently support multiple functions, such as radar, communications, and surveillance; these capabilities are collectively referred to as integrated RF applications. The multi‑purpose utilization of a single system simplifies the requirements for task‑resource scheduling and typically imposes constraints on time and/or frequency usage, as shown in Figure 2(b). For the radar applications mentioned, a STAR‑enabled solution would significantly reduce the radar‑time‑line overhead associated with information exchange within established radar networks, thereby greatly facilitating coordination among mobile platforms. This is because the unique STAR functionality enables them to communicate directly with one another using a common interface, which not only facilitates the sharing of physical resources but also reduces equipment costs compared to deploying multiple function-specific systems.

Figure 2: (a) The relationship between frequency utilization and time for current radar and communication/ESM systems, and (b) the relationship between frequency utilization and time for future multifunctional systems with simultaneous transmit-and-receive (STAR) capabilities.

Before scaling up the multifunctional STAR phased‑array to achieve high‑performance systems, including high‑power radars, several critical research areas must be addressed. The following section presents measurement results from a prototype that demonstrates the STAR array architecture and the conceptual design of its digital processing components. Subsequently, we will discuss scalable array technologies.

II. The Concept of the STAR Array

Initially, STAR technology was employed in low-power radar systems that typically feature an omnidirectional radiation pattern and modest range requirements, such as vehicle sensors. As the radar’s range increases, systems often adopt antennas with directional radiation patterns to meet stringent link-budget constraints. By employing a phased-array system, it is possible to simultaneously support multiple functions while also leveraging STAR technology, as previously mentioned.

Figure 3(a) shows an example of a STAR‑capable phased‑array architecture that employs an all‑digital approach to mitigate interference and provide isolation. This method utilizes adaptive beamforming at both the transmit and receive elements, which can be dynamically reconfigured within the same aperture. Furthermore, a unique reference‑based multi‑channel digital cancellation scheme enables recovery of signals close to the receiver’s noise floor, as illustrated in Figure 3(b), where the evolution of the exemplary signal tracks the representative power levels of the transmitted signal and the noise relative to the receiver’s dynamic range and the signal of interest.

Overall, this approach is known as Aperture-Level Simultaneous Transmit and Receive (ALSTAR). It leverages the advantages of all-digital arrays, enabling dynamically reconfigurable apertures to optimize STAR operations. In this architecture, the phased array is divided into transmit and receive subarrays, which are selected based on the requirements of a given RF application or a combination of multiple functions. This reconfigurable subarray approach enables the system to perform conventional non-STAR operations without compromising performance. Once the array is partitioned into transmit and receive sections, several processing stages are employed to achieve a high degree of isolation.

The table below outlines the methods and functions of the three stages:

The multilayer structure shown in the table above gives rise to a novel STAR array, which exhibits measured isolation exceeding 140 dB over an instantaneous bandwidth of 100 MHz centered at 2.45 GHz.

III. Scalable Array Technology

3.1 All-Digital Architecture

Scanning beams in azimuth and elevation requires an all-digital, multi‑panel planar array architecture to extend the design across two dimensions. By leveraging integrated data conversion and processing techniques, high performance can be achieved through RFSoC. The RFSoC component enables direct connection of eight dual‑polarized antenna elements to the FPGA hardware, supporting the required beamforming and signal processing, as shown in the figure below.

Figure 2: Concept of an all-digital panel‑type array, illustrating inter‑element data sharing enabled by RFSoC‑based interconnects.

An array architecture based on RFSoC can be designed to deliver scalable data-processing capabilities directly on the array itself. This not only endows it with functionalities typically absent in other array systems but also reduces the need for off‑array processing. The distinctive features of the proposed RFSoC‑based processing stem from its data‑network architecture and the flexibility it enables. As shown in the figure above, each RFSoC (gray box) can establish physical connections via high‑speed links to its nearest unit modules, thereby forming a mesh network.

These connections can also extend to the edges of each array panel, ensuring continuity between panels—a critical feature when deploying scalable processing solutions. Data‑routing rules can then be used to construct data‑processing chains tailored to specific functions and configured on the fly to optimize particular use cases, such as the STAR beamforming/cancellation process mentioned earlier. Thanks to the flexibility of these interconnections, multiple processing chains can overlap within the network, enabling more efficient data handling than conventional approaches.

3.2 STAR’s T/R Module

Traditional transmit/receive modules were not designed for STAR data‑link operation; instead, they can only switch between transmit and receive functions to support time‑division half‑duplex (TDD) systems—exactly the approach employed by conventional pulse radar systems. However, a fully digital STAR array requires scalable transmit/receive modules that, by providing a precise transmit reference signal path, enable STAR‑specific operations.

To minimize the overall array volume, these T/R modules are sized to match the array lattice pitch; the conventional approach is to implement them as integrated circuits, as shown in Figure 3.

Figure 3: Conceptual architecture of the STAR T/R module, showing the desired (green) and unnecessary (red) coupling paths.

Figure (a) shows the STAR (ALSTAR) phased-array architecture at the aperture level, highlighting digital transmit/receive beamforming and multi-channel cancellation; Figure (b) presents an example of signal flow at each processing stage; and Figure (c) displays the measured isolation of 140 dB for an eight-element linear array prototype, tested over a 100-MHz bandwidth centered at 2.45 GHz.

3.3 Beamforming/Nulling Algorithm

To address several key challenges in digital beamforming and cancellation, one approach is to leverage partitioning to reduce complexity, as shown in Figure 6(a). By adjusting only a subset of array elements while keeping the others fixed, the number of adaptive filters in the processor—and the associated complex time-domain computations—can be significantly reduced. Furthermore, the symmetry inherent in the array‑element coupling correlation matrix can be exploited, enabling fast adaptive algorithms that shorten the time required to compute filter weights, as illustrated in Figure 6(b). Finally, the scalability of these methods should be investigated by designing distributed algorithms tailored to modular processing architectures.

The implementation of the STAR algorithm highlights (a) the feasibility of subarray beamforming using adaptive elements at the T/R boundary, and (b) a fast adaptive solver designed to accelerate tuning.

4 Conclusion

Future radar systems will be capable of supporting multiple functions simultaneously, in addition to their primary radar operations. These ancillary modes include communications for information sharing and network‑node coordination, as well as ESM required for enhanced situational awareness and spectrum sensing in the environment. High‑performance implementations of these integrated RF applications will be designed using STAR technology based on phased‑array radars, which enables the shared utilization of a single, flexible resource.

5. Afterword

As noted above, the key research areas of the scalable STAR array architecture are:

• STAR T/R Module: Compatible with array lattice spacing, addressing unintended coupling through electromagnetic modeling and other techniques.
• Beamforming/Nulling Algorithms: Reduce complexity through subarray partitioning, matrix‑based beamforming, and fast adaptive algorithms.
• All-digital architecture: Designed based on RFSOC, it integrates data conversion and processing, enables high-speed interconnectivity among components via a mesh network, supports scalable data processing and multi‑function data‑link configurations, and reduces reliance on external processing.

The heterogeneous integration capabilities of RFSoC perfectly align with the core requirements of the STAR array architecture. Moreover, its SD-FEC (soft-decision forward error correction) module can address potential link‑loss issues in STAR array signal transmission, enhancing transmission reliability through error‑correcting coding.

Meanwhile, the unique characteristics of the STAR array architecture confer distinct advantages in specific application domains. For instance, in integrated sensing‑and‑communication systems, the STAR‑RIS array can achieve 360-degree, full‑space signal coverage, overcoming the coverage limitations inherent to conventional RISs that are restricted to reflecting signals. Coupled with RFSOC’s high‑bandwidth processing and low‑latency operation, it enables the deployment of highly efficient edge‑computing communication systems. This synergistic combination also aligns seamlessly with the requirements of integrated space‑air‑ground networks for precise signal control and long‑distance transmission.

Through flexible hardware resource allocation, the RFSoC’s software‑hardware co‑customization capabilities seamlessly accommodate the STAR array’s scaling requirements, from small‑scale units to large‑scale deployments. This eliminates the need for extensive hardware reconfiguration; instead, modifications to the RFSoC’s firmware and algorithmic routines suffice, significantly reducing the iteration costs associated with the STAR array architecture.

To address the development of next-generation STAR systems, Birei Electronics is adopting a new technological approach. While conventional systems are limited to distinct radar modes, this approach integrates simultaneous transmit-and-receive (STAR) technology operating on the same frequency.

As one of the earliest domestic companies to adopt RFSoC as its core technological roadmap, INTERWISER has developed over 40 RFSoC‑based module variants. In real‑time mode on a single board, its synchronization accuracy reaches ±2 ps, while for ultra‑large‑scale arrays, it achieves 200 ps—making INTERWISER the only commercial team in China to deliver millisecond‑level synchronization across thousands of nodes. Its solutions are tailored to 20 distinct application domains, including radar and quantum measurement and control. Keeping pace with advances in the semiconductor industry, INTERWISER boasts end‑to‑end system‑level R&D capabilities, spanning module design, board‑level testing, logic development, data‑flow management, RF integration, and host‑software development. Bringing cutting‑edge technologies from the lab to real‑world deployment has been INTERWISER’s mission since its inception.