As an important component of the Internet of Things (IoTs), underground localization based on magnetic induction (MI) communication with the received signal strength indicator (RSSI) has been a topic of considerable interest in wireless underground sensor networks (WUSNs). Existing works have widely exploited the active device localization based on wireless communications in underground space, where the transmission media remains air. However, achieving device-free localization (DFL) in an underground endogenous environment presents relatively large gaps. In this paper, we propose a novel MI-based communication method to achieve the DFL accuracy in complex underground endogenous environment. First, we establish an end-to-end MI link model to precisely characterize the RSSI and explicitly correlate MI antenna configurations with magnetic field attenuation in the underground endogenous environment. Subsequently, based on the magnetic-field propagation characteristics, we analyze the shadowing loss caused by the passive target in the MI link. A MI-DFL method is then formulated by incorporating the shadowing model. Finally, extensive numerical simulations across various environmental settings and target types demonstrate the effectiveness and robustness of the proposed method, indicating that MI-DFL attains high localization accuracy within the underground endogenous environments.

Wireless Powered Underground Sensor Networks (WPUSNs) extend the lifetime of underground Internet of Things (IoT) through Radio Frequency (RF) harvesting technology. However, severe transmission attenuation and dynamic Quality of Service (QoS) requirements often lead to energy imbalance and waste issues in WPUSNs. To address the challenges, we introduce a recycling mechanism that enables nodes to harvest energy from each other, thereby proposing the recycling WPUSN. Under this framework, the time allocation strategy is key to balancing energy harvesting and transmission performance. We take time slot as the optimization variable and formulate an energy efficiency maximization problem under heterogeneous QoS constraints. To solve this nonconcave fractional programming problem, we develop a Quadratic Transform-based Projected (QTP) Algorithm, which iteratively applies Karush–Kuhn–Tucker (KKT) conditions and projection operations to approximate the optimal solution. The simulation results demonstrate that the proposed method improves energy efficiency performance by 17.454% and converges 70 times faster compared with the commercial optimization solvers. The proposed recycling WPUSN opens up new prospects for sustainable underground environmental monitoring.

Low-altitude economy networking (LAENet) aims to deploy various aerial vehicles to support diverse services, where data collection from edge devices via unmanned aerial vehicles (UAVs) is a critical task. The key challenge lies in jointly optimizing energy consumption and data freshness in spectrum-constrained and eavesdropping-prone low-altitude environments during the data collection process. Although deep reinforcement learning (DRL) has become a viable solution for UAV-assisted data collection, the RL agent still has limited ability to obtain and utilize informative feedback from complex low-altitude environments. In this paper, we propose a large language model (LLM)-enhanced DRL framework for secure data collection in the LAENet, where we leverage an LLM to process environmental feedback for the RL agent. Specifically, we employ the LLM as (i) a state processor to transform basic environmental observations into task-aligned representations, (ii) a reward designer to generate enriched reward signals that guide the agent's actions toward the optimization objective, and (iii) a simulator to construct a virtual LAENet environment for evaluating enhanced state-reward pairs before policy training. Theoretical analysis and numerical results demonstrate that the proposed LLM-enhanced DRL framework achieves faster convergence, improved training stability, and superior performance compared with state-of-the-art baselines.

Utilizing device mobility to form virtual large-scale antenna arrays can provide accurate angle-of-arrival (AoA) information for robots. However, existing wireless localization systems that exploit device mobility are designed based on far-field channel assumptions and cannot directly provide range estimates. To address this problem, in this paper, we develop a novel near-field localization architecture for mobile robots by fusing the robot’s motion trajectory and channel state information (CSI) of a single antenna. Specifically, we first utilize channel reciprocity to multiply the uplink CSI and downlink CSI to eliminate the phase offset. Second, we further propose a two-stage localization algorithm that separates the line-of-sight (LoS) path from the multipath, and a multi-scale iterative scheme is employed to refine the estimation of AoA and distance of the LoS path. In addition, the range and AoA profiles for different trajectory shapes and the Cramer-Rao bounds for localization accuracy under squared channels are derived. Finally, the effectiveness of the proposed system is verified in a real environment. The simulation and experimental test results show that the proposed near-field localization system can operate in complex channel environments, and its localization accuracy outperforms the existing schemes.

Digital identity is fundamental for accessing mobile applications and managing user attributes. However, existing centralized identity management solutions rely on third-party operators, posing privacy risks and limiting user control. The decentralized solutions seek to address the issues but often fall short in privacy preservation, efficiency, and cross-application compatibility. In this paper, we propose PPUgIM, a user-governed identity management scheme with universally composable security, emphasizing privacy and data sovereignty in distributed mobile applications. PPUgIM introduces a DID-like account equipped with multi-attribute credentials, enabling users to autonomously manage and selectively disclose various identities without revealing sensitive information. An enhanced authenticated data structure is designed based on vector commitments, supporting short and constant-size proofs for efficient batch authentication of attribute credentials. Furthermore, a formal security analysis of PPUgIM is conducted, and a prototype implementation is developed for performance evaluation. Results show that credential generation takes 500 ms, verification 110 ms, with a constant size proof of 0.15 KB. Proof overhead for identities is reduced by 38.1% compared with existing schemes, demonstrating PPUgIM’s practicality in real-world distributed mobile applications.

In this paper, we propose the IFresher framework to improve the timeliness of multi-agent mobile augmented reality (MAR) systems. Existing works have made strides in accuracy-latency trade-offs, but fail to directly address realtime task responsiveness and multi-agent contention challenges. To bridge this gap, we introduce the concept of the age of analytics information (AoAI), which quantifies the combined impact of video analytics (VA) accuracy, transmission delay, and computational efficiency. By deriving a closed-form expression for AoAI, IFresher establishes a central control mechanism that jointly optimizes bandwidth allocation and video configuration to minimize AoAI while ensuring accuracy. Due to the mixed-integer nonlinear characteristics of the problem and the fact that each agent only has local observations, the problem is reformulated into a decentralized partially observable Markov decision process (Dec-POMDP). We propose a multi-agent reinforcement learning (MARL) algorithm, named convex-embedded transformer QMIX (CTQMIX), using the centralized training and decentralized execution (CTDE) framework for agent collaboration. Specifically, the convex optimization ensures optimal bandwidth distribution, and the transformer captures temporal dependencies between observations and actions across time steps to improve decisionmaking in dynamic environments. Evaluations with real-world experiments show that the CTQMIX outperforms state-of-theart (SOTA) algorithms.

This paper presents an edge-assisted massive mobile live streaming (MMLS) framework named LOCA, integrating chunk-level analysis and long-term optimization to design resource allocation, bitrate adaptation, and source selection strategies. The proposed method ensures sustained real-time video delivery while minimizing latency and communication costs. Firstly, a chunk-level analysis of the entire process of video streaming is introduced, aiming at modeling fetch queue waiting time and rebuffering duration in each time slot. By embedding this mathamatical model into consideration, a long-term optimization is formulated to minimize rebuffering and communication overhead while maintaining high video qualities for massive users. Leveraging Lyapunov optimization, we transform this problem into a computationally tractable form. Further simplification via linearization achieves near-optimal solutions by adopting the mixed-integer linear programming method with enhanced computational efficiency. Simulation results demonstrate superior stability and long-term performance compared to the state-of-theart and baseline methods, validating the framework's efficacy in MMLS scenarios

Cross-domain collaboration under wireless communication scenarios has gained traction in Internet-of-Things (IoT) applications. Authentication is essential for ensuring the security of wireless IoT devices (IoTDs). However, the existing cryptographic and physical layer authentication schemes are unappealing in cross-domain scenarios due to the presence of resource-limited IoTDs, privacy concerns of cross-domain sharing, and the negative effect of malicious attackers. This paper proposes FedScatter, a lightweight cross-domain authentication scheme for securing wireless IoTDs using backscatter communication. First, device identity signatures are constructed by harnessing passive signal features generated from feather-light backscatter tags, incurring negligible overhead. Subsequently, a federated learning model is designed to aggregate device identity information across domains while respecting device heterogeneity and data privacy. A novel parameter aggregation algorithm is proposed to bolster authentication resilience and against malicious attacks to avoid model pollution by powerful attackers with substantial hardware resources. A FedScatter prototype is implemented and evaluated, demonstrating significant improvements over state-of-the-art works in both true positive rate and false positive rate under various attacks.

The vision of backscatter communication always incorporates compatibility with active radios to enable low-cost and easy deployment. However, recent innovations lack the flexibility to communicate with heterogeneous wireless devices directly. In this paper, we design and implement a flexible backscatter system, i.e., Flexcatter, which can support various modulation schemes in a software-defined way, to be compatible with different kinds of active radios. The key technique is signal emulation, where the tag can vary the reflection coefficient in the time domain to emulate desired baseband signals. We first carefully design a cost-effective impedance network, which employs two radio frequency (RF) switches to provide up to 16 reflection coefficients. Next, we establish and model the emulation mapping between desired baseband signals and available reflection coefficients. Besides, the emulation frameworks for different modulation schemes are presented. After that, to face the emulation distortion for orthogonal frequency division multiplexing (OFDM), we introduce the oversampling method in the baseband modulation process. We further build the prototype hardware, and experiment results show that Flexcatter can flexibly generate various kinds of backscatter signals, including Wi-Fi, BLE, and LoRa. Especially the OFDM transmission generated by Flexcatter can achieve a throughput of 25.1 Mbps.

User-centric cell-free massive MIMO (CFmMIMO), where only partial access points (APs) are selected to serve a specific user equipment (UE), is a scalable extension of CFmMIMO. Existing works have studied the spectral efficiency of large-scale fading decoding (LSFD) aided user-centric CFmMIMO that includes local combining at each AP and statistical channel state information (S-CSI) based fusion in the central processing unit (CPU). However, few efforts have so far been paid to analyze the error probability bound, and existing detectors fail to balance the error probability and fusion complexity. In this paper, we analyze the symbol error rate (SER) and design low-complexity near-optimal detectors for uplink user-centric CFmMIMO systems. Considering non-identical large-scale fading coefficients and local channel estimation errors, we first leverage the pairwise error probability to derive an SER upper bound for optimal linear fusion (OLF) in the CPU, which is suitable to different local combining methods at the APs. Then, by combining local normalization methods and S-CSI based UE grouping or error correction, we design improved detectors for local maximum-ratio and local minimum mean squared error combining successively. Simulation results verify the correctness of the derived SER bound, and show that the proposed detectors are capable of approaching the SER performance of conventional OLF based counterparts with reduced fusion complexity even in scenarios with pilot contamination.

Ambient LoRa backscatter enables battery-free wireless communication with long-range connectivity for the Internet of Things. However, current efforts mainly focus on symbol-level modulation, which trades considerable data rates for long backscatter ranges. To enhance the data rate, this paper presents and prototypes Pacim, which fully explores the potential of long-period LoRa chirps and conveys additional information by varying symbol lengths. Specifically, we propose the chirp interval modulation scheme that modulates multiple data bits in each time interval between two chirp-based anchor symbols. Moreover, we design a twin-chirp cancellation method at the receiver that eliminates the frequency discontinuity within anchor symbols, and propose a fine-grained detection algorithm to measure the arrival time of anchor symbols in the frequency domain. We further propose three reliable methods to improve transmission reliability and analyze the symbol error rate (SER) performance. We also build a hardware prototype and perform comprehensive evaluations. Our experiment results show that Pacim can achieve up to$8.6 \times $throughput gain while keeping long-range, compared with the state-of-the-art ambient LoRa backscatter design.

In this paper, we design and implement a flexible backscatter system, i.e., Flexcatter, which can support various baseband modulation schemes to be compatible with active radios. The key technique employed by Flexcatter is signal emulation, where the tag can vary the reflection coefficient in the time domain to emulate desired baseband signals. To meet the low-power requirements of Flexcatter, we first carefully design a cost-effective impedance network, which can provide up to 16 reflection coefficients for different signal amplitudes and phases. Moreover, we present an effective emulation strategy for desired baseband signals. Next, we introduce the oversampling method in the baseband modulation pipeline to mitigate the effect of emulation errors caused by hardware deficiency. We further build the prototype hardware, including an FPGA platform and the radio frequency (RF) module. Experiment results show that OFDM transmissions from Flexcatter can achieve a throughput of 21.1 Mbps at the backscatter range of 12 m. To the best of our knowledge, we are the first to generate OFDM transmission with 64 subcarriers using low-power RF switches. Our design has the potential to accept other types of existing active radios as the receiver to reduce the deployment cost significantly.

Cloud-based cell-free massive multiple-input multiple-output (CFmMIMO) technology, which exploits a large number of distributed antennas to cooperatively serve multiple users, constitutes an appealing technique for B5G/6G wireless communications. However, the distributed nature of cloud-based CFmMIMO imposes great challenges in analyzing the error probability bounds, and very few efforts have so far been paid to optimize the detector design. In this paper, we try to add a stroke to this blank by analyzing the symbol error rate (SER) and design near-optimal detection algorithms. Specifically, considering non-identical large-scale coefficients and channel estimation errors, we first leverage the pairwise error probability to derive an asymptotic SER bound for uplink cloud-based CFmMIMO systems, which is verified by simulation results. Furthermore, motivated by the concepts of successive interference cancellation (SIC) and error correction mechanism (ECM), we design two distinct types of near-optimal detectors for cloud-based CFmMIMO systems and analyze their complexity and convergence performance. Finally, extensive simulation results show that our proposed SIC and ECM based detectors outperform conventional matched filtering (MF) and minimum mean squred error (MMSE) counterparts. In particular, the MMSE-SIC and MMSE-ECM detectors approach the derived asymptotic bound, and the MF-ECM detector strikes a balance between the SER and complexity in ultra CFmMIMO scenarios.