University of Houston

Federated learning (FL) is an appealing paradigm for learning a global model among distributed clients while preserving data privacy. Driven by the demand for high-quality user experiences, evaluating the well-trained global model after the FL process is crucial. In this paper, we propose a closed-loop model analytics framework that allows for effective evaluation of the trained global model using clients' local data. To address the challenges posed by system and data heterogeneities in the FL process, we study a goal-directed client selection problem based on the model analytics framework by selecting a subset of clients for the model training. This problem is formulated as a stochastic multi-armed bandit (SMAB) problem. We first put forth a quick initial upper confidence bound (Quick-Init UCB) algorithm to solve this SMAB problem under the federated analytics (FA) framework. Then, we further propose a belief propagation-based UCB (BP-UCB) algorithm under the democratized analytics (DA) framework. Moreover, we derive two regret upper bounds for the proposed algorithms, which increase logarithmically over the time horizon. The numerical results demonstrate that the proposed algorithms achieve nearly optimal performance, with a gap of less than 1.44% and 3.12% under the FA and DA frameworks, respectively.

By utilizing intelligent mobile terminals, mobile crowdsensing (MCS) can realize the sensing data collection effectively and economically. However, the privacy security and freshness quality of the obtained sensing data are two major concerns to be addressed in MCS, as they directly impact the system security and timeliness performance. In this regard, we focus on improving the data freshness performance and protecting sensing data content, sensing terminals' identification, and location information simultaneously. Accordingly, based on the artificial noise (AN)-based differential privacy and covert communication technologies, we aim to jointly minimize the Age of Information (AoI) metric and weighted privacy preservation budget in the single terminal scenario. Besides, we achieve the goal of average AoI optimization with data computing requirements in multiple terminal systems, where the privacy preservation budget is treated as the critical constraint. Furthermore, by using the backward induction (BI) method and block successive upper-bound minimization (BSUM) approach, we solve the above two optimization problems, respectively. Finally, compared with the listed baselines, the results evaluate the proposed schemes' effectiveness under various simulation settings.

Using mmWave radar to conduct gesture recognition is a promising solution for human-computer interaction. Although many studies have shown initial success, two-fold problems still remain unsolved, namely, the high-strength human activity interference and the difficulty in handling similar gestures. In light of these, we develop a robust mmWave radar based gesture recognition system, Rodar, to achieve accurate recognition of similar gestures under high-strength human activity interference, where a Multi-view De-interference Transformer (MvDeFormer) network is proposed. Specifically, to deal with the strong human activity interference, we design a DeFormer module to capture the useful gesture features by learning different patterns between gestures and interference, thereby reducing the impact of interference. Then, we develop a hierarchical multi-view fusion module to first extract the enhanced features within each view, and effectively fuse them across various views for final recognition. To evaluate the proposed Rodar system, we construct a dataset with seven similar gestures under three common human activity interference scenarios. Experimental results show that the accuracy can achieve up to 93.01%. The code implementations are available at https://github.com/Xlab2024/MvDeFormer .

Device-free vital signs monitoring is an emerging technology that utilizes the unique influence of chest vibrations on surrounding wireless signals to achieve vital signs monitoring in a device-free and contact-free manner. Existing methods could achieve good monitoring performance when high-quality reflected signals can be obtained. However, in daily vital signs monitoring at home, the received reflected signals are often very weak due to factors such as obstruction and attenuation, resulting in a sharp decrease in the monitoring performance. To address the aforementioned challenges, in this paper, we develop a diversity-enhanced robust device-free vital signs monitoring system using mmWave signals. Specifically, inspired by the concept of diversity in the field of communications, we propose a diversity-enhanced wireless sensing strategy that comprehensively utilizes multi-dimensional physical layer resources, including antennas, chirps, and space, to improve the signal-to-noise ratio of vital signs. Additionally, inspired by cameras that achieve clear images by prolonging exposure time, we propose an accumulation-enhanced localization method to lock onto the chest of the human body in complex scenarios. Extensive experiments on a 60GHz mmWave testbed demonstrate that our developed system could guarantee robust vital signs monitoring performance in various challenging scenarios, even at distances of up to 40m.

In this paper, based on martingale theory, we investigate the problem of maximum throughput in hybrid energy harvesting wireless communication systems (EH-WCS) under energy storage and delay (or backlog) constraints. Specifically, the energy supply and data transmission of the hybrid EH-WCS are modeled as two queuing systems. For the first energy supply queueing system, we construct corresponding martingales for each type of EH process and the system’s energy consumption (EC) process. Leveraging the multiplicativity of martingales, the stochastic characteristics of the hybrid EH process are described in the martingale domain. On this foundation, a closed-form expression for the energy depletion probability bounds (EDPB) under the various energy storage constraints is derived. In the second data transmission queueing system, to capture the impact of channel fading on the system’s service, we map the arrival and service processes to the signal-to-noise ratio (SNR) domain and construct the corresponding martingales. A martingale parameter is proposed that connects the martingales of the arrival and service processes with the system’s EDPB. Based on this, the closed-form expressions for the delay violation probability bounds (DVPB) and backlog violation probability bounds (BVPB) are derived. Utilizing these derived performance bounds, we address the maximum throughput optimization problems under the energy storage and delay (or backlog) constraints. Furthermore, we instantiate a scenario and provide guidance on the impact of resource allocation on maximum throughput through simulation and validation, offering insights for achieving green communication networks.

Intergrating multi-access edge computing (MEC) with the Internet of Things (IoT) is able to provide IoT sufficient computational resources in addition to its capabilities of sensing and communication. In this paper, given the limited computational and energy resources, IoT devices (IDs) are allowed to offload computational tasks to MEC severs for execution. However, as the number of IDs increases dramatically, jointly optimizing the usage of sensing, communication and computational resources becomes challenging due to the exponential growth in interactions among the IDs. In this paper, we address the energy-efficient joint optimization problem for sensing and computation in the MEC-assisted IoT system, aiming to ensure the freshness of the status update and minimize the energy consumption of IDs. To reduce the computation complexity, we introduce the concept of the general mean field N-player Markov game (GMFG), and reformulate it as a mean-field game (MFG) with teams, leveraging the network structure of states. Considering the advantages of reinforcement learning (RL) for solving dynamic problems, we propose a MFG based actor-critic algorithm (MFGAC) to minimize the long-term average system cost. Through extensive simulations, we demonstrate that the proposed method is effective and can outperform other schemes under different scenarios.

With the proliferation of cellular vehicle-to-everything (C-V2X), connected and automated vehicles (CAVs) are gradually being commercialized. CAVs can interact with road infrastructure and human-driven vehicles (HDVs) to acquire relevant traffic information, thereby altering the characteristics of traditional traffic flow. The emergence of CAVs is widely believed to bestow benefits to the traffic system in terms of safety, efficiency, and energy consumption. Nevertheless, as with most phenomena, there are two sides to the coin. Further exploration is necessary to determine whether the emergence of CAVs will trigger adverse effects and the underlying factors that may induce adverse effects. To be specific, this paper first delves into how selfish driving behaviors (egoism CAV control strategy) can have an unfavorable impact on the performance of traffic systems, thereby lowering the traffic efficiency. Subsequently, we develop an unselfish (altruism) CAV control strategy that aims to achieve global optimization and improve the overall road operational capacity. Based on the simulation results obtained at different inflow and outflow rates on highway, it is evident that egoism driving behavior leads to a 11.55% decrease in average speed performance as compared to the non-control strategy, while altruism driving behavior results in a 20.14% improvement. Furthermore, we compare the proposed strategy with the current road infrastructure control, which only improves the average speed performance by 11.6%. This indicates that controlling CAVs has the potential to replace the deployment of traditional road infrastructure, thereby optimizing social and economic benefits. This paper can provide insightful guidance for future policy formulation in the transportation authorities, wherein the emergence of CAVs needs to be effectively regulated based on altruism, thus fostering the establishment and development of a safe and efficient mixed traffic ecosystem.

The growing use of smart devices requires improving privacy and security. Conventional biometrics confront false positives and unauthorized access, stressing cautious user input. We enhance security by analyzing distinctive human physiological characteristics rather than relying on conventional methods susceptible to spoof attacks. Drinking, a common physiological activity, can provide continuous authentication. SipDeep , proposed innovative system, utilizes bone-conducted liquid intake sound, incorporating unique biometrics from bone and pharyngeal characteristics. The system captures these elements in the external auditory canal, offering a novel transparent authentication applicable to a diverse user range. Our noise filtering system eliminates environmental and anatomical interferences during drinking, including subtle body movements. The study introduces a hybrid event detection technique integrating wavelet transform with start/end points detection. Next, we extract physiological features from bone structure, liquid intake sound, and liquid intake pattern. We used the physiological features to train a deep learning algorithm based on a Triplet-Siamese network to classify authentication. The proposed model has been thoroughly compared with advanced models such as DenseNet169, ResNet18, and VGG16. Following extensive experimentation involving multiple users across various environments, SipDeep demonstrates 96.5% authentication accuracy, coupled with a 98.33% resistance to spoof attacks.

We develop inverted n-p junction arrayed vertical-cavity surface-emitting lasers (VCSELs) with 875 devices operating at ~940 nm, optimized for high optical output power in sensing applications. Employment of an n-type GaAs substrate prevents performance degradation caused by defects in p-type GaAs substrates. A tunnel junction enables polarity inversion. The inverted n-p VCSEL arrays, which are preferred for circuit design and packaging, are compared with conventional p-n junction VCSEL arrays on an n-type substrate using three-dimensional device modeling and experimental measurements. The optical output power of large-area 25×35 VCSEL arrays shows ~5.5 W at I ≈ 6 A. The threshold current density and slopes of the L - I curve of the inverted n-p VCSEL arrays are ~1.2 kA/cm ² and 0.98 W/A, respectively, which are similar to those of reference p-n VCSELs. The inverted n-p arrays demonstrate slightly better electrical performance, higher output power, and power conversion efficiency than p-n, enhancing their potential in voltage-controlled sensing systems. This is the first demonstration of the large-area inverted n-p VCSEL arrays, achieving the highest light output power critical for emerging 3D sensing and LiDAR applications.

In previous work, the authors established a generalized version of Schmidt’s subspace theorem for closed subschemes in general position in terms of Seshadri constants. We extend our theorem to weighted sums involving closed subschemes in subgeneral position, providing a joint generalization of Schmidt’s theorem with seminal inequalities of Nochka. A key aspect of the proof is the use of a lower bound for Seshadri constants of intersections from algebraic geometry, as well as a generalized Chebyshev inequality. As an application, we extend inequalities of Nochka and Ru–Wong from hyperplanes in 𝑚-subgeneral position to hypersurfaces in 𝑚-subgeneral position in projective space, proving a sharp result in dimensions 2 and 3, and coming within a factor of 3 / 2 3/2 of a sharp inequality in all dimensions. We state analogous results in Nevanlinna theory generalizing the second main theorem and Nochka’s theorem (Cartan’s conjecture).

Purpose To address the issue of RF‐induced heating for partially in and partially out (PIPO) medical devices during 1.5 T MRI scans by proposing a method of minimizing the external portion. Methods A method of tightly winding the external segment of the PIPO device is proposed to minimize the overall device effective reception length during MRI scans to mitigate the RF‐induced heating. Two commercially available PIPO medical devices and simplified solid wires were used to demonstrate the concept. RF heating results are compared between typical and minimized‐length trajectories under the American Society for Testing and Materials (ASTM) testing procedure. In addition, 16 scaled and validated device models were used in conjuncture with human body numerical simulations within three virtual human models to estimate clinically relevant heating. Results The wound segments in PIPO devices functioned as a lumped element rather than a receiving antenna, reducing induced energy/heating as compared to the original PIPO devices under typical straight or loop configurations. Minimizing the lead's external portion can reduce the RF‐induced heating by significant factors for all studied cases during ASTM phantom measurements and in human body simulations. Conclusion Our findings show a significant reduction in RF heating by minimizing the external segment, thereby enhancing patient safety during 1.5 T MRI procedures. Although limited to four devices at 1.5 T across two applications, the extent of heating reduction may vary for others. Nonetheless, tightly winding the external segment of PIPO electrodes holds promise for improving device safety under MRI.

Microbes like bacteria and fungi are crucial for host plant growth and development. However, environmental factors and host genotypes can influence microbiome composition and diversity in plants such as industrial hemp (Cannabis sativa L.). Herein, we evaluated the endophytic and rhizosphere microbial communities of two cannabidiol (CBD; Sweet Sensi and Cherry Wine) and two fibers (American Victory and Unknown). The four hemp varieties showed significant variations in microbiome diversity. The roots had significantly abundant fungal and bacterial endophyte diversity indices, whereas the stem had higher fungal than bacterial diversity. Interestingly, the soil system showed no significant diversity variation across CBD vs. fiber genotypes. In fungal phyla, Ascomycota and Basidiomycota were significantly more abundant in roots and stems than leaves in CBD-rich genotypes compared to fiber-rich genotypes. The highly abundant bacterial phyla were Proteobacteria, Acidobacteria, and Actinobacteria. We found 16 and 11 core-microbiome bacterial and fungal species across genotypes. Sphingomonas, Pseudomonas, and Bacillus were the core bacteria of fiber genotypes with high abundance compared to CBD genotypes. Contrarily, Microbacterium, and Rhizobium were significantly higher in CBD than fiber. The Alternaria and Gibberella formed a core-fungal microbiome of fiber-genotype than CBD. Contrarily, Penicillium, and Nigrospora were significantly more abundant in CBD than fiber genotypes. In conclusion, specific hemp genotypes recruit specialized microbial communities in the rhizosphere and phyllosphere. Utilizing the core-microbiome species can help to maintain and improve the growth of hemp plants and to target specialized traits of the genotype.

Despite the proliferation of rooftop solar in the United States, its deployment and associated benefits have not been distributed equitably. Many states have adopted targeted incentives to improve access to rooftop solar and increase its uptake among low- and moderate-income (LMI) communities. This article examines the policy feedback effects of energy efficiency policies and electricity-sector portfolio standards on the adoption and diffusion of LMI solar incentives across states. Event History Analyses indicate that between 2010 and 2019, the adoption and diffusion of the incentives have been conditional on a state’s portfolio standards but independent of energy efficiency policies. Feedback effects from the portfolio standards in neighboring states are found to have a regressive impact on the likelihood of adoption. Hence, the feedback effects of previously adopted renewable energy policies are helping states to better serve vulnerable communities. However, there is no evidence of geographic clustering in the diffusion of incentives.

For decades, the synthesis of 2‐quinolones, a crucial structural motif in pharmaceuticals and agrochemicals, has relied heavily on costly noble metal complexes and structurally complex ligands. Despite considerable efforts from synthetic chemists, a mild, metal‐free, environmentally friendly, and cost‐effective approach has remained elusive. This study introduces a robust, metal‐free synthetic platform that leverages an innovative organoiodine‐catalyzed electrophilic arene C(sp²)−H amination strategy to efficiently produce a wide range of new and modifiable 2‐quinolones. Moreover, this study allows ready synthetic access to novel 8‐aryl‐substituted 2‐quinolones, uncovering new chemical spaces with significant potential for medicinal applications.

Infertility was reported in approximately 15% of all heterozygous couples, with the male factor accounting for nearly half of the cases. This typically occurs due to low sperm production, sperm dysfunction, and sperm delivery obstruction. In this randomized controlled single-blind clinical trial, 90 infertile male subjects diagnosed with oligospermia, hypospermia, asthenozoospermia, or necrozoospermia were recruited. Semen samples were obtained with the masturbation method and an assessment of semen volume, sperm count, and motility was performed. Five milliamps of electrical shock was delivered to the participants through the fertility improvement device. Semen analysis was collected 4 months post-intervention from all subjects. Data were collected and an analysis of pre- and post-intervention results was performed. There was an improvement in the count, volume, and motility of the patient’s sperm after electrical shock treatment compared with the control group. By using the analysis of variance (ANOVA) test, there were statistically significant differences between the first and the second seminal analysis results (<.05). All other results were found to be independently correlated. This study demonstrated that using a painless, convenient at-home device, which is designed to contain all the testis tissue as a cup and then extend to include the scrotal roots reaching the penile root to include the epididymis, could significantly improve sperm motility and count. This device can be utilized to tackle the significant issue of infertility in a cost-effective, safe, and efficacious manner. An ultrasound was done before and after using the device as well as years after with no changes noted. Clinical Trial’s Registration Number: NCT04173052

Background Accumulating evidence indicates that coronary microvascular dysfunction (CMD) caused by hypercholesterolemia can lead to myocardial ischemia, with or without obstructive atherosclerotic coronary artery disease. However, the molecular pathways associated with compromised coronary microvascular function before the development of myocardial ischemic injury remain poorly defined. In this study, we investigated the effects of hypercholesterolemia on the function and integrity of the coronary microcirculation in mice and the underlying mechanisms. Methods and Results Mice were fed a hypercholesterolemic Paigen's diet for 8 weeks. Echocardiography data showed that Paigen's diet caused CMD, characterized by significant reductions in coronary blood flow and coronary flow reserve, but did not affect cardiac remodeling or dysfunction. Immunofluorescence studies revealed that Paigen's diet–induced CMD was associated with activation of coronary arterioles inflammation and increased myocardial inflammatory cell infiltration. These pathological changes occurred in parallel with the upregulation of lysosomal signaling pathways in endothelial cells (ECs). Treating hypercholesterolemic mice with the cholesterol‐lowering drug ezetimibe significantly ameliorated Paigen's diet–induced adverse effects, including hypercholesterolemia, steatohepatitis, reduced coronary flow reserve, coronary endothelial cell inflammation, and myocardial inflammatory cell infiltration. In cultured mouse cardiac ECs, 7‐ketocholesterol increased mitochondrial reactive oxygen species and inflammatory responses. Meanwhile, 7‐ketocholesterol induced the activation of transcriptional factor EB and lysosomal signaling in mouse cardiac ECs, whereas the lysosome inhibitor bafilomycin A1 blocked 7‐ketocholesterol–induced transcriptional factor EB activation and exacerbated 7‐ketocholesterol–induced inflammation and cell death. Interestingly, ezetimibe synergistically enhanced 7‐ketocholesterol–induced transcriptional factor EB activation and attenuated 7‐ketocholesterol–induced mitochondrial reactive oxygen species and inflammatory responses in mouse cardiac ECs. Conclusions These results suggest that CMD can develop and precede detectable cardiac functional or structural changes in the setting of hypercholesterolemia and that upregulation of transcriptional factor EB–mediated lysosomal signaling in endothelial cells plays a protective role against CMD.

Interleukin-15 (IL-15) promotes the survival of T lymphocytes and enhances the antitumour properties of chimeric antigen receptor (CAR) T cells in preclinical models of solid neoplasms in which CAR T cells have limited efficacy1–4. Glypican-3 (GPC3) is expressed in a group of solid cancers5–10, and here we report the evaluation in humans of the effects of IL-15 co-expression on GPC3-expressing CAR T cells (hereafter GPC3 CAR T cells). Cohort 1 patients (NCT02905188 and NCT02932956) received GPC3 CAR T cells, which were safe but produced no objective antitumour responses and reached peak expansion at 2 weeks. Cohort 2 patients (NCT05103631 and NCT04377932) received GPC3 CAR T cells that co-expressed IL-15 (15.CAR), which mediated significantly increased cell expansion and induced a disease control rate of 66% and antitumour response rate of 33%. Infusion of 15.CAR T cells was associated with increased incidence of cytokine release syndrome, which was controlled with IL-1/IL-6 blockade or rapidly ameliorated by activation of the inducible caspase 9 safety switch. Compared with non-responders, tumour-infiltrating 15.CAR T cells from responders showed repression of SWI/SNF epigenetic regulators and upregulation of FOS and JUN family members, as well as of genes related to type I interferon signalling. Collectively, these results demonstrate that IL-15 increases the expansion, intratumoural survival and antitumour activity of GPC3 CAR T cells in patients.