This paper proposes an adaptive anti-jamming resource allocation scheme to address both external malicious jamming and internal interference in dynamic multi-user communication environments. The scheme assesses jamming levels through a module based on average channel quality indicator (CQI) and adaptively selects the appropriate anti-jamming policies by utilizing the resouces of power-domain, frequency-domain, or a combination of power-frequency domain. This approach overcomes the limitations of fixed-domain policies in dynamic and complex jamming scenarios. To optimize resource allocation, a discrete soft actor-critic (DSAC) algorithm with an adjustable temperature coefficient is employed, addressing sample inefficiencies in traditional actor-critic methods and enhancing policy robustness. Simulation results show that the proposed scheme improves convergence speed under intelligent jamming (IJ) by 55%, 62%, 65%, 75%, and 60% compared to trusted region policy optimization (TRPO), deep deterministic policy gradient (DDPG), proximal policy optimization (PPO), deep reinforcement learning with dual action network (DRL-DAN) and soft actor critic with fixed temperature coefficient (SAC-FTC)-based schemes, respectively. The scheme achieves evasion rates of 99% against periodic comb jamming (PCJ), 98.7% against sweep jamming (SJ), 95% against combined jamming (CJ), and 90.8% against IJ. When the jammer alternates between regular jamming (RJ) and IJ, the proposed scheme responds quickly, demonstrating robustness in dynamic jamming environments.

In complex jamming environments with multi-user spectrum sharing, existing distributed anti-jamming strategies are constrained by significant communication overhead, limited efficiency in knowledge dissemination, and low collaborative effectiveness. To address these challenges, a distributed anti-jamming strategy based on local knowledge diffusion and differential weighted fusion mechanisms (LKD-DWF-M) is proposed. In this strategy, a local knowledge diffusion mechanism is introduced to facilitate knowledge sharing among communication nodes, enabling each node to gain a comprehensive understanding of its neighbors’ behavior. Subsequently, a knowledge contribution measurement method based on mutual information is proposed, and a differential weighted fusion (DWF) mechanism is designed to effectively integrate the policy and value parameters of neighboring nodes. This integration enables accurate global value estimation while optimizing individual anti-jamming strategies. Additionally, the existence of the Nash equilibrium (NE) for each node’s policy and value parameters is theoretically established using Kakutani’s fixed-point theorem. Furthermore, through the construction of a Lyapunov function, it is demonstrated that the proposed strategy can stabilize and converge to the NE in the long-term jamming counteraction process. Simulation results indicate that, in comparison to anti-jamming strategies employing global knowledge diffusion and differential weighted fusion mechanism (GKD-DWF-M), global knowledge diffusion and average fusion (GKD-AF-M), and local knowledge diffusion and average fusion (LKD-AF-M), the proposed distributed anti-jamming strategy achieves respective improvements of 4%, 17%, and 20% in system normalized throughput under statistical jamming (SJ). Under dynamic sweeping jamming (DSJ), the system normalized throughput improves by 8%, 11%, and 11.5%, respectively; under intelligent comb jamming (ICJ), it increases by 10%, 10.5%, and 19%, respectively; and under intelligent block jamming (IBJ), it increases by 5%, 16%, and 21%, respectively. Moreover, the proposed strategy exhibits superior convergence speed compared to other strategies. When the jammer alternates between SJ, DSJ, ICJ, and IBJ, the proposed distributed anti-jamming strategy responds quickly, demonstrating robustness in dynamic jamming environments.

As essential radiating elements in RF and microwave microsystems, microstrip antennas require sufficient bandwidth to ensure stable operation, integration flexibility, and overall microsystem performance. From a microsystem optimization perspective, this paper proposes a bandwidth extension method for microstrip antennas that combines swarm intelligence and reinforcement learning. The proposed ICOA-TD3 framework is designed to enhance antenna bandwidth within target frequency bands and thus improve the performance robustness of compact RF microsystems. In the proposed method, an improved crayfish optimization algorithm (ICOA) is first used to explore the global design space and achieve global bandwidth enhancement, followed by the Twin Delayed Deep Deterministic Policy Gradient (TD3) algorithm for local refinement and further exploitation of the antenna structure’s bandwidth potential. In Experiment 1, the impedance bandwidth ($S_{11}\le -10\mathrm{dB}$) is increased by up to 200%. In Experiment 2, the impedance bandwidth ($S_{11}\le -10\mathrm{dB}$) and axial-ratio (AR) bandwidth ($\mathrm{AR}\le 3\mathrm{dB}$) are improved by up to 27% and 250%, respectively. The results indicate that the proposed method is a feasible solution for bandwidth-oriented optimization of microstrip antennas and is promising for the intelligent design of high-performance RF microsystems.