Addressing phase-locking instability caused by mode interference in efficient phase-locked magnetrons, this paper proposes a design method for stable, efficient phase-locked MW-level non-relativistic S-band magnetrons based on differential regulation of intercoupled modes. By constructing a hybrid-waveguide coupling structure and regulating the relative position between the abrupt interfaces and the standing wave nodes, the field amplitude of interference modes in the anode cavities can be significantly weakened to suppress their excitation, with negligible impact on the operating mode. Multiple particle-in-cell simulations verify that at the optimal interface position Xc = 1/4 lambda gc (located in the single-mode operating region), the frequency peaks of the two interference modes disappear. The system achieves a stable locked state within 10-20 ns after saturation, representing about a 1/2 reduction in stable locked time compared to conventional coupling structures, while maintaining an electronic efficiency comparable to that of a single free-running magnetron. Field distribution tests further confirm the differential regulation capability, demonstrating that the frequency deviation amplitudes of interference modes are suppressed to approximately 50% of that of the operating mode. While maintaining the original high-efficiency phase-locking advantage, this design method significantly improves the operating stability of the intercoupled magnetron, demonstrating good application potential in large-scale phase-locked arrays.

This study proposes a method to enhance magnetron efficiency by synchronously adjusting the filament current, magnetic field strength, and operating voltage. In addition, a pulse high-voltage magnetron output characteristic analysis system with tunable multioperating parameters is established to validate the effectiveness of the proposed approach. Simulation results demonstrate that the magnetron efficiency increases from 75.3% to 91.4%, while the output power is significantly improved from 2.18 to 14.8 kW. Experimental results further confirm that, under pulsed operating conditions, the efficiency increases from 74.8% to 90.1%, and the output power rises from 1.8 to 16.6 kW.

This letter presents synchronous injection locking across multiple arrays enabled by high-power injection signals generated via a cascaded injection architecture. A frequency-stabilized signal generated by the solid-state microwave source and amplifier is employed to injection-lock a seed magnetron, thereby producing a high-power seed signal. The resulting signal is subsequently distributed via a power division network to realize injection locking of multiple arrays. Experimental results indicate that all arrays operate stably at an injection frequency of 2.468 GHz. The system achieves an injection ratio of -17.75 dB with a locking time below 200 ns. The proposed scheme obviates the need for circulators at the injection ports of the array modules and is particularly well suited for applications requiring synchronous locking of multiple magnetron arrays by low-level signals.

To investigate the locking performance of single-path injection on a multimagnetron array, a hybrid cascaded phase-locking system suitable for array applications is proposed. The five magnetrons are arranged as a cascaded coupled array using bidirectional couplers and coaxial line interconnections. A single external injection path is sufficient to achieve injection locking of the cascaded mutually coupled magnetron array. Simulation and cold measurement results indicate that different injection topologies have no significant influence on the locking characteristics of the array. The effect of the injection signal on cascaded magnetrons outside the main injection path is indirect. Other magnetrons in the array are locked through coupling with the injection-locked magnetron. High-power measurements demonstrate successful injection locking of the five-magnetron array at an injection ratio as low as -26.53 dB, operating at 2.465 GHz. Following injection locking, the array exhibits a 14.15-dBc improvement in the absolute spurious suppression ratio, an 82.67% reduction in the 3-dB bandwidth, and a pronounced enhancement in the spectral characteristics of the system. The proposed system achieves a maximum phase-locking efficiency of 92.57%. The hybrid cascaded phase-locking structure demonstrates strong applicability to phased-array systems.

An S-band racetrack slotted waveguide antenna (SWA) with low profile, high radiation efficiency (RE), low sidelobe levels (SLL), and high power capacity (PC) is presented. The antenna has only one feed port, eliminating the need for a complex T-type waveguide unequal-amplitude power divider network, and consists of a bent waveguide phase compensation structure (BWPCS) and a radiation array. The BWPCS is designed to ensure that each waveguide receives the residual microwave energy in phase, enabling a high gain (G) level. A new slot conductance formula is proposed to achieve high RE and low SLL. A ceramic radome was introduced to maintain the vacuum environment to enhance PC. Simulations at 2.458 GHz show a reflection coefficient below -26.6 dB, G of 28.09 dBi, RE of 99.73%, E-plane SLL of -29.4 dB, and H-plane SLL of -30 dB. The PC of the SWA reaches 1.56 GW.

This work proposes a method for controlling the relative orientation of the output ports in a phase-locked magnetron array and experimentally validates its feasibility. By bending a soft-waveguide-based coupling bridge, the proposed approach enables adjustable spacing between magnetrons while simultaneously providing continuous and wide-range tuning of the angle between the normal vectors of the output ports, with the phase difference between the two magnetrons maintained within a limited range. Experimental results show that, during soft-waveguide bending, the phase difference remains within 7.2 degrees-14.2 degrees, the spacing between the magnetrons can be adjusted over a range of approximately 1.4 wavelengths, and the angle between the normal vectors of the output ports can be continuously tuned from 0 degrees to 180 degrees. In addition, stable phase locking was achieved at 9.067 GHz, with a locking efficiency of up to 96.8%. These results demonstrate that employing a soft waveguide as the coupling bridge provides an effective means of controlling the angle between the normal vectors of the output ports of phase-locked magnetrons without the need for additional microwave transmission networks.

A phase-shifterless phase control method of magnetron arrays is proposed and experimentally demonstrated. By injecting a frequency-tunable signal into one magnetron within the mutual coupling magnetron arrays, the synchronized operating frequency can be adjusted, thereby modifying the phase shift in the phase-locking structure $\varphi _{c}$ and achieving a controllable phase difference $\Delta \varphi $ between the output signals. The experiment demonstrates that, within the effective injection-locking frequency from 9.0713 GHz to 9.0825 GHz, the phase difference varies quasi-linearly with the injection frequency, achieving a controllable range of over 80 degrees. Within this tuning range, the output power difference varies from -6.7 kW to 23.1 kW, corresponding to a maximum difference of 0.92 dB. To the best of our knowledge, this study is the first demonstration of phase control without phase shifters in a hundred-kilowatt-class system.

To further improve the output power of magnetron arrays while simplifying the system structure, the power combining of multiple magnetrons (>2) is essential. This article proposes a three-magnetron power combining system in which phase-locking is realized through an input-port, isolation-free, coaxial cavity radial power combiner (CCRPC). By precisely tuning the coupling phase between the CCRPC input ports with phase shifters, the magnetrons exchange energy through the input ports of the CCRPC, thereby enabling simultaneous phase-locking and efficient power combining. High-power experiments show that the proposed CCRPC can stably achieve phase-locking and power combining in a three-magnetron array, delivering a maximum combining efficiency of 95.39% and a total peak output power of 7325.51 W at 2.466 GHz. By replacing the phase shifters with equal-electrical-length coaxial lines, the system can be further simplified into a structure that requires neither an external injection source nor circulators and can ultimately operate without phase shifters. The uniformly distributed input ports along the cavity circumference of the CCRPC also provide good amplitude and phase consistency together with scalability, offering a practical route toward large-scale magnetron arrays.

This article proposes a slot antenna specifically designed for phase-locked magnetron arrays, where the spatial coupling strength and coupling phase are engineered to enable phase-locking while maintaining efficient radiation, extending the functional and application scope of the slot antenna from a conventional radiator to one that simultaneously enables magnetron phase-locking and radiation. The system has been fabricated and tested. The antenna exhibits gains of 21.21 and 21.10 dBi under low-power chamber testing and full-power magnetron operation, respectively, with sidelobe levels of -12.9 and -11 dB, and radiation efficiency of 92.68% and 90.36%. Analysis of the time-domain signals of the magnetron array reveals a stable zero-phase-difference phase-locked state. Additional power synthesizing tests show that the system achieves 96.91% spatial coherent synthesizing efficiency. The agreement between time-domain phase analysis and power synthesizing experiments provides mutual verification, collectively confirming the zero-phase-difference phase-locked state enabled by spatial mutual coupling injection of the slot antenna. The system operates stably at 2.467 GHz with an output power of around 8238.19 W.

This work tackles boundary controller design and analysis for input-saturated two-link flexible manipulators. The system dynamics, incorporating boundary force effects, are modeled using Partial Differential Equations (PDEs) defined within local body-fixed coordinate frames. To simultaneously achieve angular position regulation and suppress flexible link vibrations under actuator saturation limitations, we have presented a novel boundary control strategy. This strategy employs a smooth hyperbolic function to explicitly constrain the control input within a predetermined upper bound. Assuming a sufficiently small flexural displacement allows the approximation of its relationship with the axial rate of flexural displacement as linear. Utilizing LaSalle’s invariance principle and this approximation, the system’s stability is rigorously proved. Finally, Simulation results illustrate the efficacy of the controller in regulating the angular position and suppressing vibrations in the links.

In passive pulse compression systems, the reversing speed of the phase-inversion switch has a significant impact on the compression gain. To study the low-cost, miniaturized, and highly stable pulse compression array systems based on the vacuum electronic oscillators, such as magnetrons, a high-power, high-speed p-i-n waveguide diode phase shifter (PS) is proposed. Its high-power capacity of megawatt level has been experimentally verified. In this article, the time-domain characteristics of the PS are investigated to further verify the high-speed characteristics, as well as the stability of phase switching, which lays the foundation for applying it in the pulse compression array. The time-domain phase-switching process of the PS is extracted and analyzed in comparison with the theoretical recovery process of the p-i-n diode. The comparison results show that the phase-switching time of the PS is equal to the current change time in the diode recovery process and shorter than the storage time. The fastest switching time measured is less than 4 ns, and the phase-switching process is highly stable with a phase shift fluctuation of less than 3.1 degrees. The recovery features of the diode are simulated using the Advanced Design System (ADS) to verify the experimental conclusions and analyze the reasons for the errors between the experimental results and the theoretical model. The simulation results show that the forward current and reverse voltage mainly affect the recovery process of the p-i-n diode.

As people continue to pursue an enhanced cooking experience, this paper proposes a microwave heating uniformity improvement system that effectively integrates the benefits of injection-locking technology. This system facilitates microwave frequency and phase scanning by integrating an external injection structure. A microwave heating cavity optimized for uniformity improvement system requirements is designed by simulation of the voltage standing wave ratio within the heating cavity. Properly tuning the impedance matching between the load and the transmission line's characteristic impedance ensures that the standing wave ratio remains below 10, effectively mitigating arcing issues induced by excessive field intensity. Furthermore, uniformity experiments were conducted on the heating system comprising the newly designed heating cavity and the external injection structure. Experimental results demonstrate that, within a 30 MHz frequency range around the magnetron's center frequency and over a 360 degrees phase sweep, the heating uniformity was enhanced by up to 48.19% for thermal response paper and 92.08% for steamed buns. The uniformity enhancement effect is significant.

Mobile edge computing (MEC) provides a promising solution for Industrial Internet of Things (IIoT) tasks that require ultra-low delay. However, MEC faces greater challenges in task offloading and resource allocation due to the high dynamicity of edge networks and the stringent requirements of industrial tasks. Digital twin edge networks (DITEN) have emerged as a promising solution by mapping digital twins (DT) and physical devices in real-time. In this paper, we propose a DT-assisted optimization method for joint task offloading and resource allocation in multi-device collaborative (MDC) tasks within IIoT. Specifically, MDC tasks involve processing data from multiple sensors. Firstly, we establish a two-layer architecture for DITEN that includes physical and DT layers. The DT layer incorporates deviations in sensing, computing, and communication resources. Secondly, the problem of minimizing the end-to-end (E2E) delay of MDC tasks, subject to constraints such as energy consumption, is formulated by jointly optimizing computational frequency, transmit power, bandwidth, and offloading factors. Due to the non-convexity of the optimization problem, we decouple it into four subproblems and propose an alternating optimization algorithm that combines internal convex approximation and Lagrangian duality method for an iterative solution. Finally, simulation results verify that the proposed scheme can significantly reduce the E2E delay of MDC tasks, compared to other baseline schemes.

This study proposes a coaxial-line-based magnetron phase-locking system that enables flexible adjustment of the relative spacing between magnetrons by bending the coaxial line and validates its effectiveness through simulation. In this system, a coupling structure is established between the inner conductor of the coaxial line and the magnetrons, enabling cooperative operation of multiple magnetrons and allowing the phase difference between their output signals to be switched between 0 degrees and 180 degrees. Simulation results show that the magnetron's output characteristics in both phase-locked and free-running conditions are in excellent agreement with those of actual devices; phase-locking system performance is essentially unaffected by bending of the coaxial line, and the minimum phase-locking efficiency is 98.6%. Notably, this phase-locking system offers strong scalability for magnetron array implementations, enabling flexible adjustment of element spacing according to array output requirements and significantly enhancing the spatial compactness of magnetron array systems.

This paper proposes a method for real-time synthesis of reconfigurable pulses. The approach employs a high-voltage ultra-wideband (UWB) pulse source as the foundation, utilizing a genetic algorithm (GA) to optimize pulse synthesis parameters for targeted waveform generation. The method achieves not only peak power synthesis but also enables real-time arbitrary waveform synthesis through precise configuration of time delays and switching states of elementary pulses, thereby facilitating flexible switching between operational modes to meet diverse application scenarios. To address the insufficiency of the Pearson correlation coefficient in achieving high-fidelity waveform matching, a multi-objective fitness function is formulated for the GA, enhancing waveform consistency between synthesized and target pulses. Simulation results indicate that the waveform consistency between synthesized and target pulses exceeds 90% for positive Gaussian, negative Gaussian, and bipolar pulses, and surpasses 80% for double-exponential pulses. Experimental validation using a high-voltage UWB pulse source demonstrates that all synthesized positive Gaussian, negative Gaussian, and bipolar pulses achieve waveform consistency above 85% with targets, while double-exponential pulses exceed 75%, verifying the effectiveness of the proposed methodology.

A coupled mode theory (CMT) model based on odd-even mode decomposition is proposed to analyze the transient characteristics of detuning in the Stanford Linear Accelerator Energy Doubler (SLED). The primary objective of this analysis is to predict the detuning required to achieve low-distortion and low-loss mode switching in SLED pulse compression systems. The CMT calculations indicate that detuning the resonant cavity causes the output pulse of the SLED to exhibit a beat envelope at a frequency corresponding to the detuning frequency, resulting in pulse distortion. Numerical results show that waveform distortion and energy transmission efficiency are influenced by the resonant cavity's $Q$ -factor, coupling coefficient, and degree of detuning. Specifically, for fixed resonant cavity parameters, higher detuning leads to reduced distortion and improved energy transmission efficiency. A SLED system operating at 2.458 GHz was designed, using a choke piston to achieve stable tuning over a broad range of 86 MHz. Both simulation and experimental results demonstrate good agreement with the CMT model predictions. Experimental tests show that the designed SLED can switch to a low-loss, low-distortion noncompression mode with a detuning of 3 MHz.

Automatic modulation classification (AMC) plays an increasingly crucial role in intelligent spectrum management and dynamic spectrum access, which can effectively support the reallocation of low-utilization spectrum resources in wireless communication systems. While deep learning approaches have been widely employed in AMC, most deep learning-based AMC methods focus on signal classification as a singular task. Therefore, this paper proposes a multi-task learning-based method for radio signal recognition aimed at enhancing AMC performance. This method utilizes the designed multi-task collaborative learning network (MCLNet) model to achieve complementary gains across different tasks. By sharing parameters, it enhances the learning capability of crucial signal features, thereby acquiring more discriminative signal features and improving classification accuracy. Experimental results demonstrate that the proposed method outperforms other benchmark models on two benchmark datasets and exhibits greater performance gains in few-shot scenarios.

The adverse propagation environment in underground coal mine tunnels caused by enclosed spaces, rough surfaces, and dense scatterers severely degrades reliable wireless signal transmission, which further impedes the deployment of IoT applications such as gas monitors and personnel positioning terminals. However, the conventional power enhancement solutions are infeasible for the underground coal mine scenario due to strict explosion-proof safety regulations and battery-powered IoT devices. To address this challenge, we propose singular value decomposition-based Lagrangian optimization (SVD-LOP) to minimize transmit power at the mining base station (MBS) for IRS-assisted coal mine wireless communication systems. In particular, we first establish a three-dimensional twin cluster geometry-based stochastic model (3D-TCGBSM) to accurately characterize the underground coal mine channel. On this basis, we formulate the MBS transmit power minimization problem constrained by user signal-to-noise ratio (SNR) target and IRS phase shifts. To solve this non-convex problem, we propose the SVD-LOP algorithm that performs SVD on the channel matrix to decouple the complex channel coupling and introduces the Lagrange multipliers. Furthermore, we develop a low-complexity successive convex approximation (LC-SCA) algorithm to reduce computational complexity, which constructs a convex approximation of the objective function based on a first-order Taylor expansion and enables suboptimal solutions. Simulation results demonstrate that the proposed SVD-LOP and LC-SCA algorithms achieve transmit power peaks of 20.8dBm and 21.4dBm, respectively, which are slightly lower than the 21.8dBm observed for the SDR algorithm. It is evident that these algorithms remain well below the explosion-proof safety threshold, which achieves significant power reduction. However, computational complexity analysis reveals that the proposed SVD-LOP and LC-SCA algorithms achieve O(N3) and O(N2) respectively, which offers substantial reductions compared to the SDR algorithm’s O(N7). Moreover, both proposed algorithms exhibit robust convergence across varying user SNR targets while maintaining stable performance gains under different tunnel roughness scenarios.

To meet the phase control requirements of magnetrons in large-scale array applications, the advantages of injection locking and cascaded mutual coupling phase locking are fully integrated. A hybrid phase-locking system for magnetrons is proposed. By performing external injection on just one magnetron within the cascaded mutual coupling structure, simultaneous frequency locking of both magnetrons can be achieved. Simulation and experimental results indicate that two phase-locking regions emerge during one period of phase variation. By quantifying the power flow across the four ports of the phase locker, it is determined that the power necessary for phase locking is less than 1% of the total output power. External frequency tuning enables the stabilization of the frequencies of both magnetrons, resulting in a purer spectrum. With an injection ratio of -18.97 dB, the mutual coupling phase locker can be adjusted to vary the phase difference between the output signals of the two magnetrons within the range from -126(degrees) to 171(degrees). This facilitates phase control of the output signals from the two magnetrons. The phase-locking efficiency of this system is 88.2%. The hybrid phase-locking approach achieves high-efficiency phase locking and power output. This approach is particularly suited for coherent power synthesis and phase scanning in large-scale array applications.

The rapid development of deep learning has provided new solutions for radio signal recognition in complex electromagnetic environments, demonstrating excellent performance and being widely applied. The data-driven based radio signal recognition methods rely on large amounts of labeled data for training. However, in non-cooperative communication scenarios, the limited availability of labeled samples can lead to model overfitting, which in turn degrades the recognition performance of the model. Therefore, this paper first combines the variational mode decomposition with time series transformation to propose a mixed data augmentation method for improving the feature extraction efficiency of the model. Secondly, considering the implementation of lightweight network structures on devices with limited computing power and storage while ensuring model recognition performance, a sparse residual network (SRNet) is designed based on a hybrid attention mechanism and sparse coding structure. Finally, comparative experiments and performance analysis are conducted on simulated and real-measured data. The simulation and experiment results indicate that the proposed method can significantly improve recognition accuracy while achieving model lightweighting.