最新刊期
卷 47 ， 期 2 ， 2026
- 摘要：The evolution of communication networks is intensifying spectrum scarcity due to the growing number of frequency-dependent devices. Native-intelligence in 6G, which leverages machine learning for spectrum management, offers an effective way to enhance spectrum utilization. The research on key technologies for native-intelligence-driven spectrum management in 6G was surveyed. Firstly, emerging trends in spectrum evolution were analyzed. Secondly, a three-layer spectrum management analysis framework was established. Within this framework, native-intelligence-driven decision-making applications in cooperative spectrum sensing and dynamic spectrum access were compared, and security solutions based on blockchain and machine learning were summarized. Finally, potential challenges and future research directions were discussed.关键词：spectrum management;cooperative spectrum sensing;dynamic spectrum access;machine learning;blockchain242|275|0更新时间：2026-02-25
- 摘要：The performance degradation and decision rigidity of conventional spectrum sensing in dynamic electromagnetic environments, arising from reliance on fixed acquisition and offline training under distribution shifts, were addressed. An embodied spectrum sensing modeling and policy computation approach was proposed. An unified system model integrating environment state, observation, knowledge representation, and embodied actions was constructed, and spectrum sensing was formulated as a sequential decision process over knowledge states. An immediate reward was designed with recognition of correctness, latency, and resource cost as the core factors. The value function and the state action value function were defined, the Bellman optimality equations were derived, the existence form and structural characteristics of the optimal policy were characterized, and a policy computation method based on value function updates was developed. An embodied decision framework for modulation recognition was established, and adaptive matching to diverse environments was enabled through multi model selection. Simulation results show that, compared with a static single model baseline, the proposed scheme significantly improves overall recognition accuracy, environmental adaptability, and error recovery capability. The proposed scheme provides an effective approach for robust spectrum sensing in dynamic electromagnetic environments.关键词：embodied intelligence;spectrum sensing;dynamic environment;closed loop decision;Bellman structure139|734|0更新时间：2026-02-25
- 摘要：To address weak echo submersion and multipath interference in digital terrestrial multimedia broadcast (DTMB) external illuminator, a UAV passive sensing method based on hierarchical cross-attention network (HCANet) was proposed. A dual-branch architecture was constructed to extract temporal dependencies via time-domain multi-scale convolution and micro-Doppler features via frequency-domain amplitude-phase joint encoding. Hierarchical cross-attention modules were utilized to adaptively align and fuse time-frequency features in the feature space. Experimental results show that the method maintaines robustness within -10 to 10 dB signal-to-noise ratios. The detection accuracy under 10 dB additive white Gaussian noise, Rician, and Rayleigh channels reaches 95.75%, 99.50%, and 71.25%, respectively, and the accuracy of measured data in complex urban environments reaches 99.65%. These results validate the effectiveness in low signal-to-noise ratio and strong-scattering environments.关键词：UAV passive sensing;DTMB external illuminator;hierarchical cross-attention;weak target detection;time-frequency feature fusion136|193|0更新时间：2026-02-25
- 摘要：To address the problems of prominent long-tailed data distribution and high-dimensional sampling features in specific emitter identification (SEI), an improved SEI algorithm based on temporal attention was proposed. By constructing a multi-scale feature fusion module to integrate feature information under different temporal granularities, and designing an identity-aware temporal attention module that generates and embeds identity-adaptive attention weights, the model was encouraged to focus on individuals with a small number of tail samples and identity-related temporal features. Experiments on the self-collected AIS dataset and public ADS-B dataset demonstrate that the proposed algorithm achieves recognition accuracies of 94.89% and 93.23% respectively, and the accuracy decreases by less than 2% under a low signal-to-noise ratio (SNR) of 3 dB.关键词：specific emitter identification;long-tailed distribution;temporal attention;IQ signal;feature learning117|238|0更新时间：2026-02-25
- 摘要：Traditional wireless communication architectures with hierarchical optimization exhibit limited performance and often lack dedicated designs for specific task scenarios. Consequently, task scenario-driven cross-layer joint optimization has become an inevitable trend. However, it faces challenges such as complex coupling of multi-level parameters, large differences in response times at different levels, and severe conflicts in autonomous decision-making among multiple nodes. Therefore, inspired by the fast conditioned reflex and efficient decision-making of the human brain through deep thinking, a brain-like intelligent cross-layer joint decision-making architecture driven by task scenarios was proposed. The key technologies of integrated intelligent adaptation for link-level cross-layer decision-making and efficient cognitive adaptation for multi-node collaboration at the network level were systematically expounded. Some key technologies were verified through specific cross-layer optimization cases.关键词：wireless communication;cross-layer optimization;brain-like decision architecture;task-driven;intelligent adaptation122|589|0更新时间：2026-02-25
- 摘要：To enhance the blind signal separation performance of independent component analysis with reference (ICA-R) in communication spectrum-sharing applications and to reduce its hardware implementation complexity, a simplified non-random separation vector initialization method based on the minimum mean square error (MMSE) criterion was proposed, aiming at a lightweight design of a robust ICA-R algorithm. In this method, the observation signal matrix was directly multiplied by the reference signal to compute the initial value of the separation vector. Theoretical analysis and hardware simulation results show that, compared with existing methods, the proposed approach offers comprehensive advantages including high separation accuracy, strong robustness, fast convergence, and low hardware overhead. Experimental results on the separation and extraction of spectrally overlapped communication signals using the ICA-R algorithm demonstrate that the proposed method can still achieve efficient and robust target signal separation and extraction under strong co-channel interference, while effectively reducing the algorithm runtime. Thus, it is suitable for resource-constrained communication devices with stringent requirements for real-time performance, robustness, and computational efficiency.关键词：spectrum sharing;blind signal separation;independent component analysis with reference;separation vector initialization85|113|0更新时间：2026-02-25
- 摘要：To address the limited situational awareness and high decision coupling of traditional routing and spectrum access methods in dynamic network topologies, a joint optimization framework that integrates graph attention networks (GAT) with deep reinforcement learning (DRL) was proposed. The distributed path establishment process was formulated as a partially observable Markov decision process (POMDP), enabling hop-by-hop decentralized decisions via DRL. GAT was implemented to aggregate local observations to capture irregular topologies and inter-node interference, improving adaptability to complex environments. During training, prioritized experience replay enhances sample efficiency. Extensive simulations under random, clustered, and multi-flow scenarios demonstrate the method’s effectiveness: in random topologies, it achieves approximately 10% higher bottleneck throughput while reducing both channel switching frequency and path hop count. In clustered topologies, it reduces channel switches by about 10% and hop count by about 13%, and in multi-flow scenarios, its performance is comparable to baseline approaches.关键词：graph attention network;deep reinforcement learning;routing;spectrum access89|844|0更新时间：2026-02-25
-
Intelligent jamming decision-making algorithm based on HRDQN in complex electromagnetic environments AI导读
摘要：To address the issues of slow convergence speed and poor energy efficiency of existing intelligent jamming decision-making in communication countermeasures scenarios, a hierarchical Rainbow deep Q-network (HRDQN) algorithm was proposed. Firstly, a communication system model subject to non-cooperative intelligent jamming was formulated, and the jamming decision-making process was modeled as a Markov decision process (MDP), deriving the suppression coefficient threshold to quantify jamming effectiveness. Secondly, the action space and decision-making method of the agent were designed to improve decision-making efficiency based on a hierarchical structure. Finally, the reward function was designed to combine the suppression coefficient threshold with the estimated jamming-to-signal ratio (JSR) to guarantee stable convergence of the algorithm. Simulation results demonstrate that the proposed algorithm rapidly generates ideal jamming decisions while reducing power consumption, and outperforms traditional algorithms in convergence speed, thereby corroborating the merits of the proposed algorithm.关键词：communication countermeasure;intelligent jamming decision-making;Markov decision process;hierarchical rainbow-DQN algorithm;energy efficiency75|704|0更新时间：2026-02-25
Topics\: Key Technologies and Applications for Intelligent Spectrum Sensing\, Management and Utilization
- 摘要：To address the long control loop and hybrid traffic congestion issues caused by cross-domain interconnection scenarios of remote direct memory access (RDMA) technology, a congestion control method for computing power networks, named WRCC (WAN RDMA congestion control), was proposed. A fair rate computing strategy based on input rate was employed, enabling switches to accurately calculate the port fair rate of congested queues. Combined with dual control loops on the near-source switch and in-band network telemetry technology, it achieved end-network collaboration rate control and rapidly responded to congestion. Simulation experiments demonstrate that compared with existing commercial methods, WRCC reduces the average and tail flow completion time by 8%~47% and 10%~70%. Prototype system tests indicate that compared with NVIDIA CX7, WRCC reduces the tail latency by 7%~49% in short-distance scenarios. In long-distance scenarios of 640 kilometers, WRCC reduces the average and tail latency by 2%~7% and 45%~49%, while achieving the average throughput improvement of 26%~90%.关键词：congestion control;remote direct memory access;computing power network;end-network collaboration105|177|0更新时间：2026-02-25
- 摘要：To address the trade-off between inference performance and cost in Token-level collaboration within the AI-model network, a self-efficacy-based Token-level multi-model collaboration method named ConfiPara was proposed. Firstly, a Token-level collaborative method with an exit mechanism was designed to mitigate the high overhead of exis-ting approaches. Secondly, a self-efficacy assessment algorithm integrating the base model’s confidence and reliability was introduced to determine the optimal exit timing. By leveraging self-efficacy to guide the base model in switching to independent inference at appropriate moments, redundant collaboration was skipped, thereby maintaining accuracy while reducing Token overhead. Experimental results demonstrate that the proposed ConfiPara method achieves a substantial reduction in Token consumption and inference latency with only a minor accuracy loss. In a single collaborative model scenario, the method reduces Token cost by approximately 21% and cuts per-Token generation latency by up to 75%, at the cost of only a 2.5% drop in accuracy.关键词：large model;AI-model network;token-level model collaboration;exit mechanism;self-efficacy118|677|0更新时间：2026-02-25
- 摘要：To address the inefficiency of intersections from independent control of vehicles and traffic signals in traditional transportation, a cooperative optimization method for connected vehicles and traffic signals via C-V2X (cellular vehicle-to-everything) was proposed. A closed-loop architecture integrating perception, communication, decision-making, and control was designed, leveraging edge intelligence to empower real-time traffic decision-making. A cooperative optimization algorithm, comprising a traffic pressure-based signal phase adjustment strategy and a state machine-based vehicle speed guidance strategy, was developed to achieve the collaborative optimization of signal phases and vehicle speeds. Hardware experiments and traffic flow simulation results demonstrate that the proposed method meets the requirements for hardware deployment and real-time response, effectively reducing vehicle travel time at intersections and improving traffic efficiency.关键词：Internet of vehicles;C-V2X;edge intelligence;cooperative optimization116|137|0更新时间：2026-02-25
- 摘要：As mobile networks evolve, the user plane function (UPF) faces demands for high bandwidth, low latency, and massive connectivity. However, skewed traffic distribution and diverse services lead to a surge in fine-grained rules. The resulting rule dependencies have become a critical bottleneck for UPF forwarding performance. Moreover, existing algorithms for resolving dependencies suffer from high computational latency and fail to meet strict timing requirements. To address this, a data processing unit (DPU) driven software-hardware hyper UPF architecture was proposed. This architecture separated elephant and mice flows based on traffic features. Elephant flows were offloaded to programmable hardware for acceleration. In contrast, the CPU handled mice flows and employed a fast independent rule generation algorithm to eliminate rule dependencies. Additionally, a rule storage structure and multi-pipe cooperative update mechanism were proposed. Experimental results show that under a high concurrency test environment, the system achieved a throughput of 97 Gbit/s with an end-to-end latency below 500 µs. The computational overhead for rule dependency resolution was reduced by over 65%, with rule storage and update efficiency improved by approximately 50%.关键词：mobile network;programmable network;data processing unit;in network computing;user plane function56|663|0更新时间：2026-02-25
- 摘要：Coded distributed computing (CDC) has been recognized as a classic method to mitigate stragglers via erasure codes. Nevertheless, previous studies ignored the work completed by stragglers and incurred additional coding overheads. Thus, two coded computation methods based on hierarchical bundle matrix codes (HBMC) were proposed, fixed-rate hierarchical bundle coded distributed computing (F-HB) and rateless hierarchical bundle coded distributed computing (R-HB). F-HB was designed to solve the straggler problem and reduce coding latency. Based on this, R-HB further exploited the work completed by stragglers. The effectiveness of the proposed schemes was verified through theoretical analysis and experimental simulation. Numerical results show that F-HB and R-HB reduce task time by approximately 68% and 74% respectively, compared to uncoded distributed computing (UDC) methods. F-HB and R-HB based on HBMC reduce the task time of distributed matrix-vector multiplication computing systems significantly by lowering the latency of coding and exploiting the work completed by stragglers. The proposed schemes provide an efficient and feasible new approach to solving the straggler problem in distributed computing systems.关键词：distributed computation;matrix-vector multiplication;hierarchical bundle matrix code;F-HB;R-HB51|101|0更新时间：2026-02-25
- 摘要：In order to improve the energy efficiency of chaotic signals generated by chaotic shape-forming filters, a chaotic communication system was proposed based on time indexed chaotic shape-forming filters. In this system, time index modulation was applied in the process of generating chaotic signals, and additional information bits were transmitted by the vacancy of the basis function pulses. This improved the energy efficiency of chaotic signals generated by traditional chaotic shape-forming filters, while the transmission of chaotic reference signals was avoided. At the receiver, a chaotic matched filter was used to demodulate the information bits based on the minimum absolute value detection and polarity decision. Theoretical bit error rate formulas were derived for both the additive white Gaussian noise (AWGN) and multipath Rayleigh fading channels. The simulation results show that the proposed system can improve not only the energy efficiency but also the bit error rate performance.关键词：chaotic communication;time index;chaotic shape-forming filter;time index chaotic shape-forming filter;energy efficiency;bit error rate49|115|0更新时间：2026-02-25
- 摘要：In large language model inference accelerators integrating multiple compute and memory chiplets via 2.5D packaging, cross-chiplet communication during the decoding stage is bursty and highly unbalanced, causing traffic to concentrate on a small number of links and form hotspot queuing, which tends to make the network-on-package a performance bottleneck. To mitigate this bottleneck, T²-CHIP was proposed, a collaborative optimization method that characterized the distribution of decoding-stage cross-chiplet traffic over the interconnect, identified hotspot links, reallocated bandwidth resources, and adjusted task mapping to reduce hotspot cross-chiplet interactions, thereby effectively relieving decoding-stage communication congestion. Cycle-accurate network simulations show that the proposed method improves decoding-stage tail performance and overall throughput while reducing dynamic power consumption and maintaining low implementation overhead.关键词：large language model;2.5D chiplet architecture;die-to-die interconnect;heterogeneous co-optimization75|137|0更新时间：2026-02-25
- 摘要：In order to evaluate the ability of LBC-IoT algorithm to resist linear cryptanalysis, the set of linear approximations with the longest number of rounds was solved based on the MILP automated search technique using both direct search and iterative linear approximation loop construction methods and obtained the initial key guessing basis for each linear approximation with the longest possible number of extended rounds. The initial key guessing basis was further compressed by combining the minimum guessing basis technique, which filtered out the optimal linear approximation for key recovery attack. The results show that the LBC-IoT algorithm had a total of 6 linear approximations of 23 rounds with linear bias of 2-15, among which there existed the only optimal linear approximation with a minimum guessing basis size of only 52 bit. The first key recovery attack of up to 30 rounds was launched against the LBC-IoT algorithm based on the upward and downward expansion of this distinguisher by 3 and 4 rounds, respectively. The data, time, and storage complexity of this attack was 230 KP, 277.9 30 rounds of encryption, and 252, respectively. Compared with the existing results, the number of rounds of the attack has been increased by 4 rounds as a whole, which leads to less than 7% of the security redundancy rounds of the LBC-IoT algorithm, which is not recommended to use in practical communication data encryption anymore.关键词：lightweight block ciphers;LBC-IoT algorithm;linear cryptanalysis;minimum guessing basis50|670|0更新时间：2026-02-25
- 摘要：To address the joint optimization of dynamic channel state information acquisition and resource allocation in reconfigurable intelligent surface (RIS)-assisted non-orthogonal multiple access (NOMA) communication systems, an implementation scheme of RIS-NOMA communication scheme based on environment-aware codebook was proposed. The proposed scheme adapted the existing environment-aware codebook scheme to the RIS-NOMA scenario. In the offline phase, a virtual channel set was generated based on statistical channel state information, and RIS phase shifts, beamforming, and power allocation were jointly designed using an alternating optimization algorithm to create an offline codebook. In the online phase, configurations were selected from the codebook to maximize the achievable rate, enabling low-complexity dynamic resource allocation. Additionally, the theoretical performance of the environment-aware codebook model was analyzed under channel estimation errors. Numerical simulation results demonstrated that the proposed scheme achieved a high sum rate while ensuring user fairness and effectively managing multi-user interference. Compared to environment-aware codebook-based RIS-assisted multiple-input single-output communication, the proposed scheme achieves a sum rate improvement of 20% to 40%, particularly highlighting the advantages of NOMA systems in multi-user multiplexing and power allocation. The study provides theoretical support and design references for the practical deployment of RIS-NOMA systems.关键词：RIS;NOMA;channel training;alternating optimization;environment-aware codebook84|175|0更新时间：2026-02-25
Papers
- 摘要：Cognitive security is recognized as an emerging direction in cyberspace security, with collective cognitive security identified as its academic frontier. To systematically summarize and analyze the main methods, collective cognitive security was divided into two subtasks group detection and group anomaly detection. The former primarily utilizes community detection techniques, while the latter mainly relies on graph anomaly detection methods. Community detection was categorized into local community detection and global community detection, whereas graph anomalies were classified into node/edge-level anomalies and subgraph/graph-level anomalies. Main concepts, fundamental theories, and representative methods were systematically reviewed, with comparative analyses conducted to evaluate the strengths and limitations of existing approaches. Future research directions were also prospected.关键词：collective cognitive security;group detection;group anomaly detection44|655|0更新时间：2026-02-25
Comprehensive Reviews
- 摘要：Considering the impact of tweakeys on tweakable block cipher algorithms, achieved integral attacks on Deoxys-BC and RAIN algorithms by combining zero-correlation linear cryptanalysis with integral attacks and utilizing partial-sum technique. 176 types of 5.5-round zero-correlation linear distinguishers for Deoxys-BC-256 and 176 types of 6.5-round zero-correlation linear distinguishers for Deoxys-BC-384 were constructed by taking the mask propagation rules of tweakeys into consideration. Based on the relationship between zero-correlation linear distinguishers and integral distinguishers, achieved 10-round and 12-round integral attacks on the two versions of Deoxys-BC respectively by combining equivalent key technique. Then, 48 types of 6-round zero-correlation linear distinguishers for RAIN algorithm were constructed, and converted them into 6-round integral distinguishers. Without considering the whitening key, 10-round integral attacks were achieved on both versions of RAIN algorithm. The results show that the complexities of the proposed attack scheme are significantly reduced compared with those of the existing ones.关键词：tweakable block ciphers;Deoxys-BC;RAIN;integral attack68|156|0更新时间：2026-02-25
- 摘要：Traditional honeypot deployment schemes suffer from issues such as poor dynamic adaptability and insufficient trapping capability when confronting increasingly complex network environments. Based on the CIC-IDS-2017 attack dataset, a multi-type dynamic honeypot deployment scheme was proposed integrating the Stackelberg game and deep Q-network (DQN). First, by conducting time-state modeling on attack behaviors to capture their temporal evolutionary characteristics, and combining this modeling with Markov prediction, the prediction of unknown attacks was achieved. Secondly, considering the differences in deployment costs and trapping capabilities among different types of honeypots (low-interaction, medium-interaction, high-interaction, and mimic honeypots), a comprehensive utility function that integrated offensive and defensive benefits was designed. Finally, through the dynamic switching of the leading role in the Stackelberg game and DQN-based strategy optimization, optimal deployment under fixed resource constraints was realized, which further enhanced the dynamic adaptability of the strategy. Simulation results demonstrate that the proposed scheme can effectively cope with the temporal evolution of attack behaviors, provide an optimal honeypot deployment scheme under fixed resource constraints, and improve the adaptability of the defense system. Specifically, the scheme achieves a trapping success rate of 96% for temporal attacks (in the case of mimic honeypots), the defense utility is 35% higher than that of traditional schemes, and it can dynamically adapt to multi-type attack scenarios.关键词：network security;honeypot trapping;dynamic deployment;dynamic gaming;DQN70|136|0更新时间：2026-02-25
Correspondences
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备09064830号-19
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
- Total Visits：665640 Visits Today：300
0