Terahertz (THz) communication is emerging as a pivotal enabler for 6G and beyond wireless systems owing to its multi-GHz bandwidth. One of its novel applications is in wireless data centers, where it enables ultra-high data rates while enhancing network reconfigurability and scalability. However, due to numerous racks, supporting walls, and densely deployed antennas, the line-of-sight (LoS) path in data centers is often instead of fully obstructed, resulting in quasi-LoS propagation and degradation of spectral efficiency. To address this issue, Airy beam-based hybrid beamforming is investigated in this paper as a promising technique to mitigate quasi-LoS propagation and enhance spectral efficiency in THz wireless data centers. Specifically, a cascaded geometrical and wave channel model (CGWCM) is proposed for quasi-LoS scenarios, which accounts for diffraction effects while being more simplified than conventional wave-based model. Then, the characteristics and generation of the Airy beam are analyzed, and beam search methods for quasi-LoS scenarios are proposed, including hierarchical focusing-Airy beam search, and low-complexity beam search. Simulation results validate the effectiveness of the CGWCM and demonstrate the superiority of the Airy beam over Gaussian beams in mitigating blockages, verifying its potential for practical THz wireless communication in data centers.

Driven by the rapid development of deep learning approaches, many novel WiFi sensing based applications have emerged, such as human activity recognition, pose estimation and indoor localization. However, due to the limited richness of collected WiFi data, the performance of WiFi sensing based models still lags behind conventional vision based models in terms of recognition accuracy and generalization. To break through the bottleneck of insufficient WiFi data, we propose a diffusion model based data augmentation scheme for human activity recognition task, in which the training dataset is composed of both real data and synthetic data. In particular, to reduce training overheads of the diffusion model, it is trained by taking activity classes as input conditions. Therefore, a single model is able to generate multiple types of WiFi data corresponding to activities, thereby avoiding the need to train separate models for each individual activity. Simulation results show that the generated WiFi data samples are visually indistinguishable from real ones, even when the model is trained on a small-scale dataset. Moreover, it also shows that adding an appropriate amount of synthetic data into training dataset can indeed improve the performance of WiFi sensing in most cases.

Facing the new era of the innovations and integration between virtual reality (VR) and the Internet of Things (IoT), the characteristics and model of VR traffic need to be studied to provide better quality of service guarantee. In this work, we simulate a testbed where the wearable VR glasses is connected into the wireless network provided by the server. And the VR traffic is studied from the two aspects of deterministic and stochastic component respectively. At the same time, the arrival curve in network calculus is used to build the traffic model and a novel method for VR traffic modeling is proposed. The theoretical curves in our proposed model are compared with existing theoretical methods and measured statistical data to verify the feasibility of our proposed model. Simulations indicate that our proposed method outperforms the existing method.

Nowadays, the integration of conventional 5G communication systems and artificial intelligence (AI) has become one of the important trends for the evolution of future advanced systems. With the employment of AI, many works have verified that significant performance gains can be achieved for varieties of tasks. However, these works mainly focus on the performance improvement and ignore the practical feasibility. In this paper, we investigate the performance of a convolutional neural network (CNN) based channel estimation scheme to run on two powerful mobile terminal chips with different quantization types. In particular, the accuracy of channel estimation and the computation capability of terminal chips for supporting model inference are evaluated. Simulation results show that the accuracy losses caused by both quantization types are small in the mobile scenario of speed at 120 km/h. However, in the case of speed at 350 km/h, the INT8 quantization type leads to a performance degradation while the $\Gamma$P16 quantization type can still maintain a satisfying performance. In addition, the results also show that further optimization for AI based computing is essential for guaranteeing a promising pratical deployment more reliably. Finally, the achievable demodulation performance gain is much smaller than the channel estimation gain, which should be explored fully in the future works.

The Terahertz (0.1-10 THz) band wireless data center networks (DCNs) are promising to provide high data rates and low latency for next-generation cloud applications. However, one research gap that is still existed is the lack of measurement data and thorough characterization of the THz wave propagation in data centers. To address this problem, in this paper, two sets of measurement campaigns are conducted in a data center scenario at 130–140 GHz band, by using a vector network analyzer (VNA)-based channel sounder system with different receiver heights. The measured data is further processed to extract the multi path components (MPCs) and classify the MPCs into clusters. Furthermore, the channel characteristics, including the path loss, shadow fading, K-factor, delay and angular spreads are calculated and analyzed. Clustering results and MPCs propagation are analyzed and examined in light of the real geometry in the data center. Interestingly, comparison with measured results in meeting room scenarios at 140 GHz shows that the reflections and scattering from metal racks in the data center are more significant, resulting in lower path loss, smaller K-factor, and larger delay spreads. The measured results in this work substantiate guidelines for system design of THz wireless DCNs.

Spatial beamforming is an efficient way to realize dual-functional radar-communication (DFRC) for integrated sensing and communications towards 6G network. In this paper, we study the DFRC design for a general scenario, where the dual-functional base station simultaneously detects the target as a MIMO radar while communicating with multiple multi-antenna communication users (CUs). This necessitates a joint transceiver beamforming design for both MIMO radar and multi-user MIMO communication. In order to avoid iterative optimization with high complexity, two low-complexity beamforming designs based on CU-selection and zero-forcing are proposed, where the closed-form expressions of the low-complexity beamforming designs are derived. Simulation results are provided to verify the effectiveness of the proposed low-complexity designs.

In this paper, the training time minimization problem is investigated in a quantized FEEL system, where the heterogeneous edge devices send quantized gradients to the edge server via orthogonal channels. In particular, a stochastic quantization scheme is adopted for compression of uploaded gradients, which can reduce the burden of per-round communication but may come at the cost of increasing number of communication rounds. The intrinsic trade-off between the number of communication rounds and per-round latency is characterized. Specifically, we analyze the convergence behavior of the quantized FEEL in terms of the optimality gap. Constrained by total bandwidth, the training time minimization problem is formulated as a joint quantization level and bandwidth allocation optimization problem. To this end, an algorithm based on alternating optimization is proposed, which alternatively solves the subproblem of quantization optimization via successive convex approximation and the subproblem of bandwidth allocation via bisection search. With different learning tasks and models, the validation of our analysis and the near-optimal performance of the proposed algorithm are demonstrated by the experimental results.

Training a machine learning model with federated edge learning (FEEL) is typically time consuming due to the constrained computation power of edge devices and the limited wireless resources in edge networks. In this study, the training time minimization problem is investigated in a quantized FEEL system, where heterogeneous edge devices send quantized gradients to the edge server via orthogonal channels. In particular, a stochastic quantization scheme is adopted for compression of uploaded gradients, which can reduce the burden of per-round communication but may come at the cost of increasing the number of communication rounds. The training time is modeled by taking into account the communication time, computation time, and the number of communication rounds. Based on the proposed training time model, the intrinsic trade-off between the number of communication rounds and per-round latency is characterized. Specifically, we analyze the convergence behavior of the quantized FEEL in terms of the optimality gap. Furthermore, a joint data-and-model-driven fitting method is proposed to obtain the exact optimality gap, based on which the closed-form expressions for the number of communication rounds and the total training time are obtained. Constrained by the total bandwidth, the training time minimization problem is formulated as a joint quantization level and bandwidth allocation optimization problem. To this end, an algorithm based on alternating optimization is proposed, which alternatively solves the subproblem of quantization optimization through successive convex approximation and the subproblem of bandwidth allocation by bisection search. With different learning tasks and models, the validation of our analysis and the near-optimal performance of the proposed optimization algorithm are demonstrated by the simulation results.

Federated edge learning (FEEL) is a popular distributed learning framework that allows privacy-preserving collaborative model training via periodic learning-updates communication between edge devices and server. Due to the constrained bandwidth, only a subset of devices can be selected to upload their updates at each training iteration. This has led to an active research area in FEEL studying the optimal device scheduling policy for communication time minimization. However, owing to the difficulty in quantifying the exact communication time, prior work in this area can only tackle the problem partially and indirectly by minimizing either the iteration rounds or per-round latency, while the total communication time is determined by both metrics. To close this research gap, we make the first attempt in this paper to formulate and solve the communication time minimization problem. We first derive a tight bound to approximate the remaining communication time through cross-disciplinary effort that combines the learning theory for convergence rate analysis and communication theory for per-round latency analysis. Building on the novel analytical result, an optimized probabilistic device scheduling policy is derived in closed-form by solving the approximate communication time minimization problem. It is found that the optimized policy gradually turns its priority from suppressing the remaining communication rounds to reducing per-round latency as the training process evolves. Extensive experiments based on real-world dataset and a use case on collaborative 3D objective detection in autonomous driving are provided to verify the superiority of the proposed policy over three benchmark policies based on the indirect solution approaches.

Federated learning provides a promising paradigm to enable network edge intelligence in the future sixth generation (6G) systems. However, due to the high dynamics of wireless circumstances and user behavior, the collected training data is non-independent and identically distributed (non-IID), which causes severe performance degradation of federated learning. To solve this problem, federated learning with non-IID data in wireless networks is studied in this paper. Firstly, based on the derived upper bound of expected weight divergence, a federated averaging scheme is proposed to reduce the distribution divergence of non-IID data. Secondly, to further harmonize the distribution divergence, data sharing is associated with federated learning in wireless networks, and a joint optimization algorithm is designed to keep a sophisticated balance between the model accuracy and the cost. Finally, the simulation results based on a common-used image data set are provided to evaluate the performance of our proposed schemes, which can achieve significant performance gains with a small price of latency and energy consumption.

To adapt to the complicated interference and the high dynamics of wireless circumstances, deep reinforcement learning (DRL) has been considered as a potential solution for beam management in the massive multiple-input and multiple-output (MIMO) systems. However, due to the extremely high dimensions of both action and state spaces, the existing DRL-based schemes are with high computation costs, and the practical performance is still unknown. To provide some insights, DRL-based beam management in the massive MIMO systems is studied in this paper. First, a DRL-based beam management scheme has been designed for beyond the fifth generation and the sixth generation (B5G/6G) systems, which can support the collaborative beam selections of multiple panels with low complexity and fast convergence. Second, a system-level simulation platform is developed to evaluate the performance of our proposed scheme in B5G/6G systems. Finally, the system-level simulation results are provided, which show that our proposed scheme can achieve much higher spectrum efficiency than the referred evaluation results given by international telecommunication union (ITU).

Driven by the vision of intelligent connection of everything and digital twin towards 6G, a myriad of new applications, such as immersive extended reality, autonomous driving, holographic communications, intelligent industrial internet, will emerge in the near future, holding the promise to revolutionize the way we live and work. These trends inspire a novel technical design principle that seamlessly integrates two originally decoupled functionalities, i.e., wireless communication and sensing, into one system in a symbiotic way, which is dubbed symbiotic sensing and communications (SSaC), to endow the wireless network with the capability to “see” and “talk” to the physical world simultaneously. Noting that the term SSaC is used instead of ISAC (integrated sensing and communications) because the word “symbiotic/symbiosis” is more inclusive and can better accommodate different integration levels and evolution stages of sensing and communications. Aligned with this understanding, this article makes the first attempts to clarify the concept of SSaC, illustrate its vision, envision the three-stage evolution roadmap, namely neutralism, commensalism, and mutualism of SaC. Then, three categories of applications of SSaC are introduced, followed by detailed description of typical use cases in each category. Finally, we summarize the major performance metrics and key enabling technologies for SSaC.

Energy harvesting is necessary to make the wireless network self-sustaining and self-organizing regardless of the traditional power grid. How to allocate the limited radio resources in the energy harvesting wireless network is a challenging work. This paper focuses on optimizing the time and power related resources in the multi-cell scenario to maximize the throughput constraining of the changeable energy in the base station. The optimal off-line resource allocation strategy is proposed, based on analysis of structural properties of the optimal total power sequences. To decrease the complexity of the optimal solution, a low complexity suboptimal off-line algorithm is presented based on the nature of the concave function. Furthermore, inspired by the off-line algorithm, an on-line resource allocation algorithm is proposed as well. Simulation results show the suboptimal offline algorithm closely tracks the performance of the optimal solution. And the proposed on-line algorithm also has brilliant performance compared with several kinds of algorithms under different system settings.

Device-to-Device (D2D) technology provides a potential way to improve cellular network throughputs. However, severe interference is introduced with universal frequency reuse, which significantly degrades the network performance. In this paper, we investigate distributed power control strategies in a D2D underlaid cellular network. An enhanced single-leader-multiple-followers Stackelberg game model is presented, where the quality-of-service (QoS) constraints of both the macro base station user (leader) and the D2D users (followers) are considered simultaneously in the price update by using a discount factor. The conditions for the uniqueness of Stackelberg equilibrium (SE) are proposed, and a distributed power control algorithm related to the price update for the game model is proposed to achieve SE. The simulation results show that besides the reasonability and fairness in power allocation, our proposed scheme provides commendable QoS protection, and there exists a tradeoff between the QoS of the leader and followers by adopting different price policies.

Conventional study of green communication mainly focuses on the transmission power adjustment to minimize the total power consumption while guaranteeing a target system capacity. However, for the energy efficient design the dynamic transmission mode is an effective way to reduce total transmission power in multiuser networks. In this paper, an energy efficient traffic transmission scheme based on user convergence behavior (UCB) is proposed which characterizes the phenomenon of similar/convergent users' traffic requests during a certain timewindow. First a system model is built to study the relations of user convergence, length of time-window and transmission power consumption. Specifically in each time-window the transmitter analyzes the similarity of users' traffic requests and the similar traffics will be transmitted by multicast mode while the other traffics will be transmitted using unicast mode. To analyze the performance of our scheme, we establish a simple stochastic model in which locations and density of users, wireless channel conditions and transmitting mode are considered. Analytical results, such as power reduction ratio and energy efficiency (EE) of the proposed scheme, are developed, from which the quantitative relationship between UCB and the energy conservation can be obtained. Simulation results validate the theoretical analysis and demonstrate that our scheme can potentially lead to 35% power consumption deduction compared with the conventional transmission scheme.

Distributed antenna systems (DAS) are deployed not only for enhancing coverage, but also for reducing the transmission power consumption, which have emerged as the promising candidate for energy-efficient wireless communications. In this paper, orthogonal and co-channel resource allocation schemes considering energy efficiency in single-cell DAS are analyzed. For orthogonal resource allocation scheme, the interference is avoided by sacrificing the available resources. While for the co-channel resource scheme, the spectrum efficiency improves under the expense of increasing the interference. The close-form expressions of spectrum and energy efficiency of these two schemes are derived, by which the impact of transmission power on spectrum and energy efficiency are theoretically compared. Both the theoretical analysis and simulation results indicate that increasing transmission power beyond a reasonable level will seriously impair energy efficiency. Furthermore, the DAS with co-channel resource allocation scheme is recommended to utilize in the green communication systems.

Channel estimation (CE) is a challenging problem and one of the key technologies in Orthogonal Frequency Division Multiplexing (OFDM) systems. The Linear Minimum Mean Square Error (LMMSE) proposed for OFDM systems has excellent Mean Square Error (MSE) performance. However, the conventional method requires the statistical knowledge of the channel information in advance and the complicated inverse operation of large dimension matrices. In this case, an enhanced fast LMMSE channel estimation using Fast Fourier Transform (FFT) operation and the most significant taps (MST) algorithm for OFDM systems is proposed. This method has the advantage of low computational complexity and no requirement on the detailed channel information. In addition, simulation results demonstrate that the proposed algorithm outperforms other LMMSE methods in terms of bit error rate (BER) without the loss of MSE performance.