In Vehicular Crowdsensing (VCS) systems, task allocation is a critical process that connects sensing demands with distributed vehicular resources. Its effectiveness fundamentally relies on optimizing the spatio-temporal features of vehicles, where information freshness is quantified by the Age of Information (AoI). However, task allocation that aims to jointly optimize spatial coverage and AoI requires the collection of precise trajectory and region data, which raises serious privacy concerns and becomes a key barrier to practical deployment. To achieve privacy-preserving and spatio-temporally optimized task allocation, this paper proposes a novel AoI-aware privacy-preserving task allocation framework based on secret sharing. The framework encodes both task regions and vehicle trajectories into timestamp-associated arrays and splits them into lightweight random shares. Through collaborative computation between two fog servers and a cloud platform, it enables secure evaluation of vehicle-task spatio-temporal matching without exposing the original data. Furthermore, we design a budget-constrained, spatio-temporally near-optimal vehicle selection algorithm that simultaneously maximizes spatial coverage and minimizes AoI under budget constraints. Security analysis and experimental results demonstrate that our scheme effectively protects location privacy while achieving near-optimal spatio-temporal coverage. Moreover, it significantly reduces both communication and computational overhead compared with a homomorphic encryption baseline, confirming its practical effectiveness for time-sensitive VCS applications.

Large Language Models (LLMs) have been applied to various hardware design tasks, including Verilog code generation, EDA tool scripting, and RTL bug fixing. Despite this extensive exploration, LLMs are yet to be used for the task of post-synthesis metric reasoning and estimation of HDL designs. In this paper, we assess the ability of LLMs to reason about post-synthesis metrics of Verilog designs. We introduce MetRex, a large-scale dataset comprising 25,868 Verilog HDL designs and their corresponding post-synthesis metrics, namely area, delay, and static power. MetRex incorporates a Chain of Thought (CoT) template to enhance LLMs' reasoning about these metrics. Extensive experiments show that Supervised Fine-Tuning (SFT) boosts the LLM's reasoning capabilities on average by 37.0%, 25.3%, and 25.7% on the area, delay, and static power, respectively. While SFT improves performance on our benchmark, it remains far from achieving optimal results, especially on complex problems. Comparing to state-of-the-art regression models, our approach delivers accurate post-synthesis predictions for 17.4% more designs (within a 5% error margin), in addition to offering a 1.7x speedup by eliminating the need for pre-processing. This work lays the groundwork for advancing LLM-based Verilog code metric reasoning.

Backpropagation has been the cornerstone of neural network training for decades, yet its inefficiencies in time and energy consumption limit its suitability for resource-constrained edge devices. While low-precision neural network quantization has been extensively researched to speed up model inference, its application in training has been less explored. Recently, the Forward-Forward (FF) algorithm has emerged as a promising alternative to backpropagation, replacing the backward pass with an additional forward pass. By avoiding the need to store intermediate activations for backpropagation, FF can reduce memory footprint, making it well-suited for embedded devices. This paper presents an INT8 quantized training approach that leverages FF’s layer-by-layer strategy to stabilize gradient quantization. Furthermore, we propose a novel “look-ahead” scheme to address limitations of FF and improve model accuracy. Experiments conducted on NVIDIA Jetson Orin Nano board demonstrate 4.6% faster training, 8.3% energy savings, and $\mathbf{2 7. 0 \%}$ reduction in memory usage, while maintaining competitive accuracy compared to the state-of-the-art.

Integrated Sensing, Communication, and Computing (ISCC) combines sensing, communication, and computing functions to improve spectrum efficiency and reduce hardware costs. However, poor link quality in the presence of obstructions leads to high offloading latency. This paper explores the use of intelligent reflecting surface and non-orthogonal multiple access in ISCC systems to enhance link reliability and improve computation offloading efficiency. A latency minimization problem is formulated by jointly optimizing computing, sensing, and communication parameters. Due to the strong coupling of the optimization variables, the problem is decomposed into two modules: computational design and sensing/communication design, which are optimized alternately to achieve a high-quality, stable solution. Simulation results demonstrate that the proposed system significantly reduces task processing latency, and ensures both user sum rate and radar sensing performance, offering a promising approach to enhance ISCC systems under resource-constrained environments.

In the Internet of Vehicles (IoV), data privacy concerns have prompted the adoption of Federated Learning (FL). Efficiency improvements in FL remain a focal area of research, with recent studies exploring model pruning to lessen both computation and communication overhead. However, in the IoV, model pruning presents unique challenges and remains underexplored. Pruning strategy design is critical as it directly impacts each vehicle's learning latency and capacity to participate in FL. Furthermore, FL performance and model pruning are intricately connected. Additionally, the fluctuating number and mobility states of vehicles per round complicate determining the optimal pruning ratio, closely intertwining pruning with vehicle selection. This study introduces Vehicular Federated Learning with Adaptive Model Pruning (VFed-AMP) to tackle these challenges by integrating adaptive pruning with dynamic vehicle selection and resource allocation. We analyze the impact of pruning ratios on learning latency and convergence rate. Then, guided by these findings, a joint optimization problem is formulated to maximize the convergence rate concerning optimal vehicle selection, bandwidth allocation, and pruning ratios. Finally, a low-complexity algorithm for joint adaptive pruning and vehicle scheduling is proposed to address this problem. Through theoretical analysis and system design, VFed-AMP enhances FL efficiency and scalability in the IoV, offering insights into optimizing FL performance through strategic model adjustments. Numerical results on various datasets show VFed-AMP achieves superior training accuracy (e.g., at least 13.4% improvement for BelgiumTS) and significantly reduces training time (e.g., at least up to$1.8\times$for CIFAR-10) compared to traditional FL methods.

Multi-location task allocation is one of the most crucial issues in vehicular crowdsourcing (VCS). To ensure service quality, the VCS service provider prefers to assign multi-location tasks to the workers whose future trajectories have high spatial proximity with the task locations. However, this process requires workers and task owners to upload their precise locations to a not-fully-trusted service provider, thereby raising location privacy concerns. Although several privacy-preserving trajectory similarity evaluation schemes have been proposed, they either fail to match the multi-location task allocation scenario, or incur nontrivial computational costs due to homomorphic encryption. To address these challenges, we propose a novel efficient privacy-preserving multi-location task allocation scheme in fog-assisted VCS. Specifically, we design a lightweight secure Euclidean distance computation protocol based on arithmetic secret sharing (ASS), which can compute Euclidean distance without revealing the two input locations. Then, based on this protocol, we build our scheme that supports multi-location task allocation based on Hausdorff semi-distance (HSD). Our security analysis demonstrates the location privacy preservation of our scheme, and the experiment results on a real dataset also validate the efficiency of our scheme.

The flourishing of intelligent connected vehicles (ICVs) has fostered the emergence of vehicular crowdsourcing (VCS) applications, in which ICVs function as workers to execute diverse spatio-temporal critical tasks. As a vital service of VCS, task scheduling aims to assign tasks to the most suitable workers. To cope with the escalation of service scale, the service provider tends to outsource the service to powerful cloud servers, which however triggers the privacy concerns of workers, task owners, and the service provider. Previously reported privacy-preserving task allocation schemes can mainly be divided into single-attribute-aware and multiattribute-aware schemes. Nevertheless, the former suffers from practicality issues, while the latter either fails to achieve single-dimensional privacy and access pattern privacy or introduces substantial computational costs. To tackle the above challenges, we propose an efficient privacy-preserving outsourced task scheduling scheme (EPTS) for VCS, in which two cloud servers can cooperate to efficiently and securely conduct multiattribute-aware task scheduling. To this end, we devise five lightweight secure two-party protocols under the additive secret sharing (ASS) setting, which enable cloud servers to obliviously filter workers that meet multiple constraints and traverse the candidate worker set to obtain the optimal worker without revealing the input and output. Rigorous security analysis proves that our EPTS scheme effectively preserves user privacy, single-dimensional privacy, and access pattern privacy. Extensive experimental results validate its superior efficiency compared with the state-of-the-art scheme.

Vehicular crowdsensing (VCS) has emerged as a promising paradigm, in which spatio-temporal-based sensing tasks are outsourced to intelligent connected vehicles (ICVs) carrying sensor-equipped devices. A critical issue of VCS is to guarantee the spatio-temporal sensing coverage by assigning tasks to appropriate vehicles, which inevitably requires vehicles’ precise locations or trajectories and thus raises location privacy concerns. To address this problem, we propose a novel secret sharing-based efficient privacy-preserving task allocation scheme for VCS, which can select sensing vehicles with approximately optimal total spatio-temporal coverage based on their future trajectories while achieving strong location privacy preservation for users (customers and sensing vehicles). With a grid-based region encoding method, a user’s location information is encoded as a binary array, termed as the region code. Based on the idea of secret sharing, we design a bit-wise XOR-based secret splitting method to split a user’s region code into two random shares and separately transmit them to two fog servers, thereby perfectly hiding the original location information. With a carefully-designed code permutation mechanism and a greedy task allocation algorithm, the cloud server and fog servers can efficiently collaborate and complete task allocation based on permuted region codes without revealing users’ location information. Detailed security analysis shows that our proposed scheme effectively preserves users’ location privacy. Extensive experiments conducted on a realistic traffic scenario data set also demonstrate that it is efficient in communication and computation while achieving large total spatio-temporal coverage.

Approximate computing is an emerging computing paradigm that offers improved power consumption by relaxing the requirement for full accuracy. Since the requirements for accuracy may vary according to specific real-world applications, one trend of approximate computing is to design quality-configurable circuits, which are able to switch at runtime among different accuracy modes with different power and delay. In this paper, we present a novel framework RUCA which aims to synthesize runtime configurable approximate circuits based on arbitrary input circuits. By decomposing the truth table, our approach aims to approximate and separate the input circuit into multiple configuration blocks which support different accuracy levels, including a corrector circuit to restore full accuracy. Power gating is used to activate different blocks, such that the approximate circuit is able to operate at different accuracy-power configurations. To improve the scalability of our algorithm, we also provide a design space exploration scheme with circuit partitioning. We evaluate our methodology on a comprehensive set of benchmarks. For 3-level designs, RUCA saves power consumption by 43.71% within 2% error and by 30.15% within 1% error on average.

Deep Neural Networks (DNN) are widely deployed in resource-limited edge devices. Due to the limitation of computational resources, it is important to meet the timing and energy constraints while maintaining a high level of accuracy. To deploy the same DNN model on different edge devices, one challenge is to train a dynamic neural network with the flexibility of balancing the trade-off between accuracy and efficiency at runtime. In this paper, we present a novel methodology, dynamic Weight-enabling Network (WeNet), where the weights of neural network can be dynamically enabled or disabled to switch between different sub-networks, so that we are able to balance the trade-off between inference time, energy consumption and model accuracy. We extend the methodology to convolutional layers using group convolution and channel shuffling. We also propose a design space exploration approach to search for the optimal sub-network for different scenarios. We thoroughly evaluate our methodology using a number of DNN architectures on different hardware platforms, showing that WeNet provides a large number of energy-efficient operation modes, 73.2 % of which provide better accuracy-efficiency trade-off compared to other methodologies.

This paper proposes a secure beamforming scheme to enhance the physical layer security (PLS) in the intelligent reflecting surface (IRS) assisted cognitive radio networks (CRNs). Taking the statistical channel state information (CSI) error of eavesdroppers (Eves) related channels into account, we jointly optimize the transmit beamforming at the cognitive base station (CBS) and reflect beamforming at the IRS to minimize the transmit power subject to the quality of service of secondary user (SU), the limited interference on the primary users (PUs), the secrecy rate (SR) outage probability constraint of SU and unit modulus of the IRS. To tackle this mathematically intractable problem, we first transform the non-convex outage constraint into deterministic form by using the Bernstein-type inequality and then exploit the alternating optimization method with the help of semi-definite relaxation (SDR) and Gaussian randomization method. Simulation results show that our proposed algorithm can significantly reduce the transmit power compared with that without IRS, and allows us to use fewer antenna number of CBS.

Approximate computing is an emerging computing paradigm offering benefits in hardware metrics, such as design area and power consumption, by relaxing the requirement for full accuracy. In circuit design, a major challenge is to synthesize approximate circuits automatically from input exact circuits requiring minimal expert input. In this work, we present a method for approximate logic synthesis based on the Boolean matrix factorization, where an arbitrary input circuit can be approximated in a controlled fashion. Our methodology enables automatic computation of the dominant elements,bases, of the truth table of the circuit, and later combines the bases to approximate the original truth table. Such compression can reduce the complexity of the hardware implementation significantly, while introducing variable degrees of inaccuracy. Furthermore, in our approach, the factorization algorithm can be fine tuned as required by the application, to effectively improve control over degree of approximation. In this work, we provide a unified approach enabling the factorization algorithm to utilize semiring algebra, field algebra, and a combination of both for truth table factorization. In addition, we provide an automatic circuit partitioning approach and a design space exploration heuristic to navigate the search space. We implement our methodology using a full stack of open-source tools, and thoroughly evaluate our methodology on a number of representative circuits showcasing the benefits of our proposed methodology for approximate logic synthesis. Finally, we compare our methodology against an existing library of approximate designs and demonstrate state-of-the-art performance.

In vehicular ad hoc networks (VANETs), vehicles broadcast emergency messages and beacon messages, which enable drivers to perceive traffic conditions beyond their visual range thus improve driving safety. However, internal attackers can launch a false message attack for selfish purposes by reporting a non-existent traffic incident in emergency messages. Moreover, some collusion attackers may spread bogus beacon messages cooperatively to make the bogus traffic incident more deceptive. To improve the accuracy of false emergency message detection, we propose a novel intrusion detection system (IDS) based on time series classification and deep learning. Considering that traffic parameters are highly correlated with time, we collect time series of traffic parameters closely related to traffic incidents from messages of vehicles near reported traffic incidents as time series feature vectors. To recognize the pattern of traffic parameters changing over time more accurately, a traffic incident classifier based on long short-term memory (LSTM) is designed and trained using time series feature vectors from both normal and collusion attack scenarios. Based on the classification result, the authenticity of the emergency message can be determined. Finally, we evaluate the performance of the proposed LSTM-based IDS through extensive simulation. Simulation results validate that our IDS is more accurate in false message detection compared with some well-known machine learning-based schemes.