The recent appreciation of deep reinforcement learning (DRL) arises from its successes in many domains, but the applications of DRL in practical engineering are still unsatisfactory, including optimizing control strategies in cognitive electronic warfare (CEW). CEW is a massive and challenging project, and due to the sensitivity of the data sources, there are few open studies that have investigated CEW. Moreover, the spatial sparsity, continuous action, and partially observable environment that exist in CEW have greatly limited the abilities of DRL algorithms, which strongly depend on state-value and actionvalue functions. In this work, we use Python to build a three-dimensional space game named Explorer to simulate various CEW environments in which the electronic attacker is an unmanned combat air vehicle (UCAV) and the defender is an observation station, both of which are equipped with radar as the observation sensor. In our game, the UCAV needs to accomplish the task of detecting the target as early as possible to perform follow-up tracking and guidance tasks. To allow an “infant” UCAV to understand what “target searching” is, we train the UCAV’s manoeuvring strategies by means of well-designed reward shaping, simplified constant accelerated motion control, and a deep deterministic policy gradient (DDPG) algorithm based on a generative model and variational Bayesian estimation. The experimental results show that when the operating cycle is 0.2 s, the search success rate of the trained UCAV in 10,000 episodes is improved by 33.36% compared with the benchmark, and the target destruction rate is similarly improved by 57.84%.

In this paper, we investigate the problem of joint direction of departure (DOD) and the direction of arrival (DOA) estimation in bistatic MIMO radar. A new bistatic MIMO radar with the distributed nested array is proposed. Based on the new array, the DOD and DOA estimation algorithm called automatically paired DR-ESPRIT based on the angle disambiguation algorithm by using the range information of targets is proposed. By designing distributed nested array and using Khatri-Rao product processing, the long baseline and short baseline in the transceiver arrays are virtually extended simultaneously. The non-ambiguous and high-accuracy sine estimations of the DOD and DOA can be accomplished using DR-ESPRIT algorithm and the range information of targets and the geometry of bistatic MIMO radar. The virtual extension of array aperture and high angle estimation accuracy can be achieved for the MIMO radar with automatic pairing and without increasing the number of antennas compared with traditional uniform linear MIMO array. And, the proposed corresponding algorithm gives significant improvement in the DOD and DOA estimation performance. The simulation results validate the theoretical algorithm.

Electromagnetic warfare is the most extensive and most hidden theatre of battle in modern warfare. To enhance the jamming effectiveness of a cooperative jammer platform against the threat of a radar net, a combinatorial‐optimisation‐based threat evaluation and jamming allocation (COTEJA) system is proposed. This COTEJA system fully considers the confrontation analysis in the jammer‐radar process, including the interactions between radars, jammers, and jammer‐radar pairs, and emphasises the realisation of cooperative jamming strategies. The cooperative jamming strategies include the combination of jamming techniques and optimisation algorithms for the objective function. The performance of the COTEJA system is evaluated through a combat mission that considers a platform with four jammers attacking five threats. In addition, the extended permutation‐based differential evolution algorithm is used for the first time to optimise the jamming coding matrix, which effectively reduces the danger value of netted radar under multiple constraints. The numerical results reveal that the COTEJA system can make the optimal jamming decision within 1 s, which improves the survival ability of the platform in a complicated electromagnetic environment.