Search

In this correspondence, we investigate a hardware-impaired uplink cell-free massive multiple-input multiple-output (MIMO) system with intelligent reconfigurable surfaces (IRSs) under correlated Ricean fading channels. Based on the imperfect channel knowledge, we derive tight closed-form expressions for the lower-bound achievable rate and study the achievable rate under various settings, such as different channel conditions, degrees of hardware impairment, and numbers of APs/IRS/elements per IRS. We also formulate the design of the IRS phase-shifts and power control factors as an optimization problem that maximizes the weighted sum-rate (WSR) and develop algorithms to solve it. All theoretical analyses and the effectiveness of the proposed algorithms are verified via extensive simulations.

Wireless federated learning (FL) is a distributed learning framework, in which mobile devices (MDs) collaboratively train a model while protecting data privacy. Considering selfishness of MDs, they are unwilling to sacrifice their resources to assist model training without any reward, which may limit performance of FL. In this paper, we propose an incentive mechanism between a base station (BS) and several mobile devices (MDs) modeled as a Stackelberg game. During this game, BS first decides the total reward that can be distributed to MDs, and then each MD determines its number of local iterations according to its utility function. In addition, we derive the closed-form expressions of the optimal reward function of the BS and the number of local iterations of the MDs. Finally, the numerical results validate the efficacy of our proposed scheme.

Targeting the high-quality signal and transmission robustness of sensing and communications in future intelligent vehicles, this article proposes an ultra-wideband compact antenna which can work at left-hand circular polarization (LHCP) and right-hand circular polarization (RHCP). It consists of two cross-arranged dipoles, four parasitic elements, a feeding network, and a ground plane. The radiation arm of the dipoles is shaped in a truncated elliptical form to broaden the 3-dB axial ratio bandwidth. To stabilize the radiation pattern and improve the axial ratio at higher frequencies in the proposed cross-dipole antenna, four sequential rotating patches are strategically positioned on the two cross-arranged dipoles. The design incorporates a specialized and highly-integrated feed network, comprising a miniaturized 90° stripline coupler and two double-sided parallel lines. This feed network is meticulously engineered to excite the dipoles, resulting in the generation of dual-circularly polarized radiation. Through the implementation of this innovative approach, an ultra-wideband dual-circularly polarized antenna having stable radiation pattern is designed, manufactured, and measured. The proposed cross-dipole antenna achieves impedance bandwidth (Reflection coefficient ≤−10 dB) from 1.25 to 4.7 GHz and axial ratio bandwidth (Axial ratio ≤ 3 dB) from 1.3 to 4.7 GHz, indicating that the overlapped impedance and axial ratio bandwidth reaches 113.3%. Good radiation performance is obtained within the ultra-wide bandwidth.

Federated edge learning (FEEL) emerges as a privacy-preserving paradigm to effectively integrate edge computing for the implementation of deep learning-based vehicular applications. Nevertheless, the incentive mechanism for vehicles participating in varied learning tasks, has not been well explored yet. In this paper, software-defined network (SDN) technology is adopted for the training control among vehicles, and a novel FEEL framework, namely SDN-assisted semi-decentralized FEEL (SSD-FEEL) is investigated, where multiple edge servers collectively coordinate a large number of vehicular models from different learning tasks. By exploiting the low-cost and similar learning tasks among vehicles and edge servers, SSD-FEEL incorporates more training samples, while enjoying the flexibility of edge server assisted model aggregation. Aiming at motivating vehicles to actively participate in training while improving the model accuracy of multiple learning tasks, a joint contract design and task reorganization problem, combined with the evaluation of model convergence and contract performance, is formulated. Then, we propose a two-stage optimization algorithm incorporating iterative reward allocation and task matching, where the model parameters in different tasks are reconstructed according to the matching results with the mobility constraints. Extensive experiments conducted on multiple data sets validate that the proposed algorithm can achieve higher cluster utility and outperform the conventional multi-task FEEL schemes in terms of learning performance.

Electric vehicles (EVs) provide sustainable and eco-friendly transportation. However, long-range driving is challenging due to limited battery capacity. While charging stations can replenish the battery, they are not as widely deployed as gas stations, causing the battery depletion of EVs. This paper presents a novel optimal charging strategy for EV long-range driving. The strategy offers EV drivers the optimal charging instructions for a given trip. A new finite-horizon Markov decision process (FH-MDP) problem is formulated to minimize the travel time of an EV while preventing its battery depletion, by optimally selecting the charging stations and charging times along the trip. By confirming the monotonicity and subadditivity of the FH-MDP, we prove the existence of a monotone deterministic Markovian policy for the optimal charging decision and reveal the optimal charging time is monotone regarding the remaining battery level and driven distance. We also reveal that the optimal charging time only changes when either of two thresholds regarding the remaining battery level or driving distance is met. By comparing its state with the thresholds, the EV can make optimal decisions with linear complexity. Simulations corroborate that our algorithm can save travel time by at least 12.6% under our simulation settings, compared to alternative methods.

In this paper, a two-layer large scale fading precoding (LSFP) method is proposed in a downlink cell-free massive multiple-input multiple-output (mMIMO) system. In the first LSFP layer, the access points are responsible for designing distributed precoding and power control coefficients for each user equipment (UE). While for the second LSFP layer, the central processing unit performs the proposed zero forcing LSFP (ZF-LSFP) for the copilot UEs in order to mitigate the interference caused by pilot contamination. To address power constraints and enhance the system's spectral efficiency (SE), we introduce the copilot power scaling factors. Specifically, we derive closed-form expressions for the lower-bound achievable SE under distributed precoding schemes. These expressions are versatile and applicable to downlink SE irrespective of the availability of ZF-LSFP. Based on the derived closed-form results, we formulate a sum SE maximization problem with respect to copilot power scaling factors. Two approaches are designed to solve this problem, i.e. the weighted minimum mean square error (W-MMSE) based approach and the geometric programming (GP) based approach. While the W-MMSE-based approach is suitable for various systems, the GP-based approach is specifically tailored for systems with high signal-to-noise ratios, offering fast running speed. Finally, we validate all theoretical analyses and assess the effectiveness of the proposed two-layer LSFP method through numerical simulations.

In this paper, a recursive robust set-membership fusion estimator (RSMFE) is presented to track the moving targets treated as the mobile nodes in a mobile wireless sensor network (MWSN) system with unknown but bounded (UBB) anchor location uncertainty. Firstly, the accurate ellipsoidal set containing the Lagrange remainder of nonlinear distance measurement, which contains the effect of anchor location uncertainty, is estimated and adopted to improve the stability and accuracy of the moving target tracking. Secondly, the recursive RSMFE-based tracking formulae of reduced computation complexity, consisting of the prediction formulae and the fusion update formulae, are derived to estimate the robust minimized or optimal ellipsoidal set containing the actual moving target state by appropriately fusing the measured distances from the nearby anchors subject to the UBB anchor location uncertainty, where the computation complexity of the fusion update formulae is reduced substantially based on the decoupled techniques and the equivalent transformation. Finally, the numerical simulation and experiment of the WSN-assisted moving target tracking system with various anchor location uncertainty settings are performed to verify the effectiveness and superiority of the proposed recursive RSMFE-based moving target tracking or localization method.