To enhance the efficiency and quality of communications, it is crucial to employ digital predistortion (DPD) technology for linearizing the power amplifiers (PAs) in terahertz/millimeter wave (mmWave) transceivers. Previous studies have shown that training DPD models within the direct learning architecture (DLA) framework yields superior results, as the Transformer-based PA behavioral model in this framework can directly compute the inverse function. Owing to resource-constrained deployment environments, DPD models trained via DLA typically rely on alternative lightweight models-such as long short-term memory (LSTM) networks. However, in scenarios where only DLA is applicable, lightweight DPD models often suffer from limited linearization performance. To mitigate this limitation, drawing inspiration from related work on iterative learning control (ILC), we propose a simple yet effective cross-architecture knowledge distillation (KD) method in the DLA framework (dubbed CAKDDLA). In this method, the lightweight DPD model is trained using two sources: the PA behavioral model and input-output knowledge distilled from a teacher model. To fully verify the proposed method's effectiveness, extensive experiments were conducted on a 1-GHz dataset, showing that our approach outperforms baseline methods in terms of error vector magnitude (EVM) and adjacent channel leakage ratio (ACLR), while also raising the upper bound of model performance.

To design reliable and efficient terahertz/mmWave transceivers, accurate power amplifier (PA) behavioral modeling is essential. In terahertz/millimeter wireless communication, the bandwidth will be 1 GHz or even more than 1 GHz, where PAs exhibit strong non-linearity and strong memory effects. This necessitates a powerful model capable of capturing long-term signal dependencies while managing strong non-linearity. To this end, we propose multi-scale augmented Transformer (MSAformer), which combines the ability of long short-term memory (LSTM) to capture complex sequence patterns with the self-attention mechanism’s dynamic attention adjustment, allowing it to effectively capture the intricate relationships between PA input and output signals. To fully validate the methods’ effectiveness, we collected and analyzed signals from the physical platform of the D-band system and set the bandwidth from 1 GHz to 4 GHz. Extensive behavioral modeling experiments on the collected datasets demonstrate that our method outperforms the existing methods in terms of normalized mean square error (NMSE). Further application in DPD scenarios proves that our method can effectively help improve the linearization performance of lightweight DPD models in terms of error vector magnitude (EVM) and adjacent channel leakage ratio (ACLR).

The serving paradigm of large language models (LLMs) is rapidly shifting towards complex multi-agent workflows where specialized agents collaborate over massive shared contexts. While Low-Rank Adaptation (LoRA) enables the efficient co-hosting of these specialized agents on a single base model, it introduces a critical memory footprint bottleneck during serving. Specifically, unique LoRA activations cause Key-Value (KV) cache divergence across agents, rendering traditional prefix caching ineffective for shared contexts. This forces redundant KV cache maintenance, rapidly saturating GPU capacity and degrading throughput. To address this challenge, we introduce ForkKV, a serving system for multi-LoRA agent workflows centered around a novel memory management paradigm in OS: fork with copy-on-write (CoW). By exploiting the structural properties of LoRA, ForkKV physically decouples the KV cache into a massive shared component (analogous to the parent process's memory pages) and lightweight agent-specific components (the child process's pages). To support this mechanism, we propose a DualRadixTree architecture that allows newly forked agents to inherit the massive shared cache and apply CoW semantics for their lightweight unique cache. Furthermore, to guarantee efficient execution, we design ResidualAttention, a specialized kernel that reconstructs the disaggregated KV cache directly within on-chip SRAM. Comprehensive evaluations across diverse language models and practical datasets of different tasks demonstrate that ForkKV achieves up to 3.0x the throughput of state-of-the-art multi-LoRA serving systems with a negligible impact on generation quality.

Handover (HO) is one of the pivotal technologies for mobility management in highly dynamic mega LEO Earth Orbit satellite constellations (MLSCs). Due to the lack of channel reservation under random access (RA) and the prohibitive overhead incurred by centralized HO scheduling, intense competition among massive connections (e.g., IoRT nodes) results in significant degradation of service continuity. To solve the above challenges, we propose a distributed fairness-guided handover strategy (DHO-F) to dynamically select the best HO target and sub-channel. Specifically, the DHO-F optimization problem is formulated based on max-min egalitarian fairness and further modeled as a multi-objective Markov decision process (MOMDP), where fairness is expressed by the social welfare function (SWF). To address MOMDP, the analytical form of the policy gradient to maximize fairness is derived. Subsequently, Multi-agent Proximal Policy Optimization (MAPPO) with distributed cooperation is exploited to achieve long-term maximization. The fairness-guided MAPPO (FG-MAPPO) features a hybrid network architecture that simultaneously takes into account maximizing individual link rates and fairness among UEs. It reconciles these two conflicting objectives through the collaboration between a throughput-oriented (TO) network and a fairness-oriented (FO) network. Additionally, a distributed training framework is implemented to improve the sample efficiency and data diversity for on-policy FG-MAPPO. FG-MAPPO is fully compatible with 3GPP's conditional handover (CHO) framework, demonstrating that it can be implemented in real-world. Extensive evaluations demonstrate that DHO-F demonstrates superior performance even compared to centralized algorithms, achieving an average improvement of 3.48% in SWF metric. Moreover, DHO-F establishes new SOTA performance in balancing fairness and rate maximization across medium-to-high load scenarios compared to IDQN, ISAC, and MAPPO.

With the advancement of satellite resources and service capabilities, efficient inter-satellite scheduling strategies have become a key factor in meeting the growing demand for satellite tasks. The dynamics and limited resources of large-scale satellite networks, along with heterogeneous satellite states and capabilities, significantly complicate service scheduling and task offloading. Effectively and rapidly allocating large-scale satellite resources for inter-satellite scheduling and offloading remains a critical challenge. In this work, we abstract satellite services and formulate a multi-objective optimization problem that jointly considers task allocation and power control, with three optimization objectives: delay, energy consumption, and task load ratio. Leveraging the physical characteristics of large-scale satellite constellations, We divide heterogeneous constellations into multiple clusters and design a two-stage cluster-level scheduling strategy: one for intra-cluster optimization and another for inter-cluster matching. Within each cluster, we design a multi-objective heuristic particle swarm optimization algorithm to quickly generate efficient scheduling strategies. Across clusters, we extend the Kuhn-Munkres algorithm with weighted and virtual nodes to achieve efficient satellite matching for task offloading. Simulation results demonstrate that the proposed cluster-level scheduling algorithms improve scheduling performance compared to other approaches.

The extremely large-scale multiple-input multiple-output (XL-MIMO) technology is emerging as a leading solution to meet the growing demands for high spatial resolution and spectral efficiency in future wireless networks. By examining the spatial correlation between received steering vectors and coherent processing gain vectors in wideband settings, we demonstrate how precise sensing and communication can be seamlessly integrated under a location division multiple access (LDMA) mechanism enabled by wideband XL-MIMO. Our theoretical analysis reveals the disparity between range and cross-range resolutions, emphasizing the critical role of wideband techniques in addressing the range resolution limitations of XL-MIMO systems. To mitigate this issue, we propose an efficient algorithm to optimize transmit precoding for both spatial and frequency beamforming. Supporting our findings, numerical and analytical results confirm the essential synergy between wideband technologies and XL-MIMO for achieving superior spatial resolution. Additionally, these advancements enable enhanced ISAC capabilities with expanded performance boundaries.

This paper investigates an integrated communication and positioning (ICAP) system facilitated by rate-splitting multiple access (RSMA). We propose encoding part of users’ messages into a common stream using a public codebook, which simultaneously facilitates fingerprint-based positioning. To enhance positioning accuracy, we investigate the interplay between geometric and non-geometric spatial consistency, which intriguingly leads to improved quality of imperfect channel state information (ICSI). In particular, we establish a novel strategy to significantly reduce the minimum mean square error of channel estimation via the Bayes pooling principle. We also demonstrate the theoretical equivalence between ICSI enhancement and positioning accuracy. Moreover, we develop a progressive transmission protocol that minimizes training overhead alongside ICSI enhancement strategy while allowing for the reallocation of spare resources without compromising designed performance. Our cooperative ICAP system leverages both reconfigurable intelligent surface and transmit precoding techniques to reconfigure the spatial consistency of the propagation space. Numerical results underscore advantages of proposed system in achieving the broadest Pareto boundary among existing ICAP designs. Furthermore, we reveal that RSMA bolsters communication capabilities while the ICSI enhancement strategy predominantly augments positioning.

In non-terrestrial networks, satellite constellations based on the Low-Earth-Orbit (LEO) have become crucial for ensuring seamless global connectivity. The flexible continuity guarantee is demanded for task-oriented user connection requests in satellite networks. In this paper, we propose a task-oriented satellite conditional handover scheme, D-CHO, based on multi-agent game theory. For the problem formalization, this paper focuses on delay overhead, satellite utilization deviation, and load performance to model the multi-objective optimization. According to game theory, a Nash equilibrium exists among the multi-task game strategies that require satellite links. Through exploration and exploitation, the optimal satellite handover sequence scheme can be identified. This paper explores the optimal solution based on the MAPPO algorithm, which enables multiple tasks to make independent decisions based on their partial observations without requiring global information. This approach is beneficial for the adaptive expansion in response to dynamic changes in different satellite networks. The simulation results show that D-CHO improves performance by 22%, 26%, and 11% compared to the SCDP, G-CHO, and MADDPG-CHO algorithms, respectively, and exhibits better scalability while maintaining satisfactory performance.

This paper investigates the design of integrated detection and communication waveforms with a particular focus on the exploitation of constructive interference (CI) at both the system and symbol levels. We formulate a waveform design problem with the objective of optimizing the detection performance which is characterized by the Kullback Leibler distance (KLD) between the probability density functions (PDFs) under two hypotheses, subject to the constraints of the CI condition and constant modulus. To address the challenging non-convex problem, a low-complexity recursive penalty-based Riemannian conjugate gradient (RE-PRCG) algorithm is proposed. The numerical results demonstrate the effectiveness of the proposed algorithm in improving the detection performance and reducing the computational burden over the benchmark.

This paper proposes a novel reconfigurable intelligent surface (RIS)-aided integrated communication and positioning design for orthogonal frequency division multiplexing systems in indoor scenarios. A non-geometric strategy is employed to realize accurate positioning. Specifically, location-related information is embedded into channel frequency responses (CFR) and estimated through regular pilot subcarriers. The coefficients of RIS are optimized to maximize the norm of the CFR vector differences among users, exclusively considering physically adjacent users. To enhance positioning accuracy, we propose a two-stage framework that incorporates the prior information about the user in physical space. A unique feature, named “correlation dispersion”, within this framework is leveraged to enhance performance compared to geometric-based methods. By transforming the geometric prior information into the frequency domain capitalizing on Gaussian kernel method, we derive the Cramer-Rao Lower Bound (CRLB) of the proposed framework. A notable gain in CRLB is observed, highlighting the efficacy. Theoretical comparison with the CRLB of conventional methods validates the correlation dispersion property. Simulation results demonstrate a significant improvement in positioning accuracy when meticulously combining prior information with a non-geometric positioning method. Furthermore, our results unveil that the incorporation of rough positioning methods yields exceptionally high positioning performance, provided that the location information depicts different aspects.

To ensure high-quality communication, it’s of great value to use digital pre-distortion (DPD) to linearize the core component power amplifier (PA) of terahertz/mmWave transceiver. In this work, we propose transfer learning with Transformer and LSTM for DPD of terahertz/mmWave transceiver, which uses Transformer based PA behavioral model to train effective and lightweight LSTM model for DPD. To collect and analyze the signal data of terahertz/mmWave transceiver, we set up a physical platform for the D-band system. Through experiments, we show that the proposed method is capable of significantly reducing in-band and out-band distortion while avoiding the excessive complexity of DPD models.

Compared with geostationary orbit (GEO) satellites, low earth orbit (LEO) satellites offer advantages of lower communication latency, smaller size, and have thus been extensively deployed worldwide. However, due to the limited overpass time and limited payload capacity, it is difficult to provide stable and continuous transmission within the short service periods of LEO satellites. To support stable and continuous transmission, LEO satellites can offload data to GEO satellites. In this paper, we investigate a data offloading and resource allocation problem in the LEO-GEO satellite cooperative transmission networks, with the aim of optimizing the energy consumption of the network. To stress the problem, we decompose the original problem into three individual subproblems based on the Lyapunov optimization theory. In addition, we analyze the queue stability of the network. Simulations demonstrate the queue stability of the network and evaluate the impacts of system parameters on energy consumption and average delay. In addition, simulations validate the superiority of the proposed method.

To meet the demand for high-speed and high-quality communication in next 6G satellite communication, it is very necessary and urgent to study the behavioral modeling of 6G satellite communication power amplifiers (PAs). In satellite communication, PAs face the situation of high dynamic and wide bandwidth and exhibit strong non-linearity and strong memory effects. In this case, we need to study Transformer architectures that can better handle long sequence data and further explore the inherent characteristics of the PA signal data. In this article, we propose a behavioral modeling method of PAs named augmented real-valued time-delay Transformer (ARVTDform). ARVTDform is an augmented Transformer based method, which can capture long-range dependencies between the PA signal data and has powerful non-linear modeling capabilities. To simulate the working status of the satellite PAs, we set up two physical platforms and collect and analyze twelve datasets. To the best of our knowledge, this is the first time real satellite data has been used for behavioral modeling. Extensive experiments on the collected datasets further demonstrate that our Transformer based method is more suitable for handing PAs with strong non-linearity and strong memory effects in terms of normalized mean square error (NMSE). Finally, we discuss the major challenge and list the potential future work that may contribute to the sustained development of high-performance transceivers.

The deployment of low earth orbit (LEO) satellites megaconstellations presents a promising way for achieving global coverage and service, attributed to their comparatively low round-trip latency and launch costs. However, this surge in LEO satellite launches exacerbates the scarcity of the limited spectrum resources. Spectrum sharing between satellite constellations and terrestrial networks and beam hopping (BH) technology emerge as viable strategies to mitigate this spectrum shortage. To enhance spectrum efficiency and avoid serious inter-system interference, we investigate the beam hopping scheduling of satellites for interference avoidance. The beam hopping scheduling of the integrated satellite-terrestrial wireless networks system is formulated as throughput-driven beam hopping (TDBH) problem and satisfaction-rate-driven beam hopping (SDBH) problem, respectively. In particular, we decompose the TDBH problem into two sub-problems by relaxation, and a genetic algorithm (GA) is introduced to handle the SDBH problem. The impact of channel conditions and traffic load intensity on the satellite system throughput is analyzed in TDBH simulation. As for SDBH optimization problem, the simulation results show that the proposed GA algorithm improves the average traffic satisfaction rate by 16.96% at least, compared with other benchmarks and suits to scenarios with different traffic demands and fading channel conditions.

The performance of wireless communication systems is fundamentally constrained by random and uncontrollable wireless channels. Recently, reconfigurable intelligent surfaces (RIS) has emerged as a promising solution to enhance wireless network performance by smartly reconfiguring the radio propagation environment. While significant research has been conducted on RIS-assisted wireless systems, this paper focuses specifically on the deployment of RIS in a wideband millimeter wave (mmWave) multiple-input-multiple-output (MIMO) system to achieve maximum sum-rate. First, we derive the average user rate as well as the lower bound rate when the covariance of the channel follows the Wishart distribution. Based on the lower bound of users’ rate, we propose a heuristic method that transforms the problem of optimizing the RIS’s orientation into maximizing the number of users served by the RIS. Simulation results show that the proposed RIS deployment strategy can effectively improve the sum-rate. Furthermore, the performance of the proposed RIS deployment algorithm is only approximately 7.6% lower on average than that of the exhaustive search algorithm.

This paper investigates the cooperative integrated communication and positioning (ICAP) within multiple-input single-output (MISO) systems. A novel strategy that leverages positioning outcomes is presented to enhance the quality of imperfect channel state information (ICSI), thereby bridging these two functionalities. Initially, a fingerprint-based method is developed to establish positioning function with the channel frequency responses serving as identification features. We focus on the interplay of geometric and non-geometric spatial consistency to improve positioning accuracy. To this end, reconfigurable intelligent surface (RIS) and transmit precoding techniques are employed to promote the non-geometric spatial consistency while ensuring communication quality. Moreover, we refine the distribution of random channel uncertainty through the Bayes pooling principle by leveraging the reshaped spatial consistency, resulting in an updated ICSI covariance matrix. This refinement is proven to significantly enhance the quality of ICSI, as evidenced by a reduction in the minimum mean square error. Furthermore, the study introduces a progressive transmission protocol that reduces training overhead in line with the channel enhancement strategy. Numerical results validate that the proposed protocol enhances the accuracy positioning by adjusting the power allocation without compromising the performance of communication. Moreover, opting for overhead reduction rather than directly utilizing enhanced ICSI quality demonstrates superior communication performance.

Beam hopping (BH) scheduling is a crucial technology for future satellite communication systems. How to dynamically match the quite non-uniform traffic demands of cells with the limited satellite beam resources over time slots remains a challenge. Current deep reinforcement learning (DRL) based beam hopping methods do not take into account the fading channel and impact of demands distribution. In this paper, we formulate a BH scheduling problem which aims to maximize the traffic satisfaction rate of each served cell. We propose a DRL based approach to obtain the scheduling optimal policy through interactions with fading channel and differentiated traffic demands. In addition, evaluation results illustrate that the proposed method can achieve better offered-requested data match with other benchmarks, and suite to scenarios with different traffic demands and fading channel conditions.

Reconfigurable intelligent surface (RIS) has been expected to be a promising technology for next generation communication systems based on its capability of tailoring propagation environments. At the same time, hybrid analog and digital precoding is also a key technology for improving the energy efficiency of millimeter wave (mmWave) massive multiple-input multiple-output (MIMO) wireless systems. The requirements of low cost and power consumption for future wireless systems can be met by combining these two technologies. In this paper, we formulate the problem of jointly designing the hybrid precoding at the base station (BS) and the phase shifts at the RIS for both single-cell and multi-cell scenarios to maximize the achievable sum rate. For the single-cell scenario, we propose an improved sparse precoding (ISP) scheme for analog transmit precoding (TPC) design while considering the beam squint effect, and utilize the two-stage digital TPC scheme to reduce the computational complexity. Inspired by the two-stage digital TPC design, we propose the three-stage digital TPC design for the multi-cell scenario, which handles inter-cell and intra-cell interference, and power allocation respectively. Finally, simulation results demonstrate that the proposed scheme has better performance than existing hybrid precoding schemes.

In this paper, we propose a novel integrated communication and positioning design for orthogonal frequency division multiplexing system aided by a reconfigurable intelligent surface (RIS) in indoor circumstances. The channel frequency responses on pilots (CFROPs) of places of interest are used for online mapping with the offline CFROP database. We transform the objective of minimizing the similarity of different CFROPs into creating a differentiated database by optimizing the phase coefficients of RIS. Imperfect channel state information is considered due to time-varying caused by the two-stage mapping. We formulate a universal optimization problem for maximizing either the average or the minimum virtual distance of CFROPs. The communication service requirements are converted as constraints. A moderate case is discussed to reduce computational complexity with minor accuracy loss. A special property called correlation dispersion is analyzed. It is capable of eliminating the spatial consistency that incurs inaccuracy to traditional positioning methods. The property and the moderate case complement each other well with clear and logical physical interpretation. The particular characteristic makes our design outperform others especially in high-level-noise environments. It works even better when the prior information of user's potential location is available. The validity of our design is confirmed by numerical results.

Spectrum sensing is an effective method to improve spectrum utilization. In this paper, we propose a wideband spectrum sensing scheme based on the virtual grid model. Specifically, we firstly divide the network coverage area into grids with a certain accuracy according to the virtual grid model, and virtualize user nodes. And we only consider the impact of these virtual nodes on receiving nodes, while ignoring the actual transmitting nodes. Secondly, we transform the wideband spectrum sensing into a multi-dimensional joint compressive sensing problem according to the sparsity in the wideband frequency domain and the correlation in space domain. Finally, the sensing matrix is properly designed by minimizing the maximum column coherence in order to get better recovery accuracy, and the compressive sensing-based algorithm is proposed. Simulation results show that the proposed algorithm can effectively improve the performance of wideband spectrum sensing.