This paper studies federated learning (FL) in a classic wireless network, where learning clients share a common wireless link to a coordinating server to perform federated model training using their ...local data. In such wireless federated learning networks (WFLNs), optimizing the learning performance depends crucially on how clients are selected and how bandwidth is allocated among the selected clients in every learning round, as both radio and client energy resources are limited. While existing works have made some attempts to allocate the limited wireless resources to optimize FL, they focus on the problem in individual learning rounds, overlooking an inherent yet critical feature of federated learning. This paper brings a new long-term perspective to resource allocation in WFLNs, realizing that learning rounds are not only temporally interdependent but also have varying significance towards the final learning outcome. To this end, we first design data-driven experiments to show that different temporal client selection patterns lead to considerably different learning performance. With the obtained insights, we formulate a stochastic optimization problem for joint client selection and bandwidth allocation under long-term client energy constraints, and develop a new algorithm that utilizes only currently available wireless channel information but can achieve long-term performance guarantee. Experiments show that our algorithm results in the desired temporal client selection pattern, is adaptive to changing network environments and far outperforms benchmarks that ignore the long-term effect of FL.
We propose a new approach, which we term as scalar auxiliary variable (SAV) approach, to construct efficient and accurate time discretization schemes for a large class of gradient flows. The SAV ...approach is built upon the recently introduced IEQ approach. It enjoys all advantages of the IEQ approach but overcomes most of its shortcomings. In particular, the SAV approach leads to numerical schemes that are unconditionally energy stable and extremely efficient in the sense that only decoupled equations with constant coefficients need to be solved at each time step. The scheme is not restricted to specific forms of the nonlinear part of the free energy, so it applies to a large class of gradient flows. Numerical results are presented to show that the accuracy and effectiveness of the SAV approach over the existing methods.
The NOD-, LRR- and pyrin domain-containing protein 3 (NLRP3) inflammasome is a cytoplasmic supramolecular complex that is activated in response to cellular perturbations triggered by infection and ...sterile injury. Assembly of the NLRP3 inflammasome leads to activation of caspase-1, which induces the maturation and release of interleukin-1β (IL-1β) and IL-18, as well as cleavage of gasdermin D (GSDMD), which promotes a lytic form of cell death. Production of IL-1β via NLRP3 can contribute to the pathogenesis of inflammatory disease, whereas aberrant IL-1β secretion through inherited NLRP3 mutations causes autoinflammatory disorders. In this review, we discuss recent developments in the structure of the NLRP3 inflammasome, and the cellular processes and signaling events controlling its assembly and activation.
The NOD-, LRR- and pyrin domain-containing protein 3 (NLRP3) inflammasome, a critical component of the host innate immune system, has an important role in microbial infection, but its aberrant activation causes inherited disorders and contributes to sporadic inflammatory diseases.At steady state, the structure of NLRP3 is oligomeric and kept in an inactive form through interactions among the C-terminal LRR domains. In response to specific stimuli, NLRP3 forms a supramolecular complex, called the inflammasome, which activates caspase-1, leading to the release of interleukin (IL)-1β and IL-18.The NLRP3 inflammasome senses the disturbance of intracellular homeostasis induced by an array of stimuli that converge on K+ efflux, which is critical for NLRP3 activation.Localization of NLRP3 to the dispersed trans-Golgi network has been suggested to have an important role in NLRP3 activation.Post-translational modifications regulate the NLRP3 inflammasome at both the priming and activation steps.
This paper studies the degrees of freedom (DOFs) of a line-of-sight (LOS) communication system, employing orbital angular momentum (OAM)-based orthogonal multiplexing to enhance the spectral ...efficiency. Aperture and ring receive domains are considered, and analytical expressions of the DOF numbers are obtained. For both cases, the DOF numbers are found to be strongly dependent on the relative sizes between the transmitter and the receiver because of the divergence of the OAM-carrying beams. It is also found that the DOF number of an OAM-based system is less than that of a comparable system allowed to use any supported waves, suggesting that OAM-based multiplexing does not offer any additional performance benefits over traditional spatial multiplexing techniques. Some practical implications of the OAM-based radio are discussed. A numerical study is also presented to evaluate the analytical findings and to address the impacts of the assumptions adopted to reach these findings. The analytical DOF expressions are evidenced to have very good performance in both high-DOF and low-DOF scenarios.
In this paper, we study and predict flow observables in 2.76 and 5.02 A TeV Pb + Pb collisions, using the iEBE-VISHNU hybrid model with TRENTo and AMPT initial conditions and with different forms of ...the QGP transport coefficients. With properly chosen and tuned parameter sets, our model calculations can nicely describe various flow observables in 2.76 A TeV Pb + Pb collisions, as well as the measured flow harmonics of all charged hadrons in 5.02 A TeV Pb + Pb collisions. We also predict other flow observables, including
v
n
(
p
T
)
of identified particles, event-by-event
v
n
distributions, event-plane correlations, (normalized) symmetric cumulants, non-linear response coefficients and
p
T
-dependent factorization ratios, in 5.02 A TeV Pb + Pb collisions. We find many of these observables to remain approximately the same values as the ones in 2.76 A TeV Pb + Pb collisions. Our theoretical studies and predictions could shed light to the experimental investigations in the near future.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
Immunotherapies that target PD-1/PD-L1 axis have shown unprecedented success in a wide variety of human cancers. PD-1 is one of the key coinhibitory receptors expressed on T cells upon T cell ...activation. After engagement with its ligands, mainly PD-L1, PD-1 is activated and recruits the phosphatase SHP-2 in proximity to T cell receptor (TCR) and CD28 signaling. This event results in dephosphorylation and attenuation of key molecules in TCR and CD28 pathway, leading to inhibition of T cell proliferation, activation, cytokine production, altered metabolism and cytotoxic T lymphocytes (CTLs) killer functions, and eventual death of activated T cells. Bodies evolve coinhibitory pathways controlling T cell response magnitude and duration to limit tissue damage and maintain self-tolerance. However, tumor cells hijack these inhibitory pathways to escape host immune surveillance by overexpression of PD-L1. This provides the scientific rationale for clinical application of immune checkpoint inhibitors in oncology. The aberrantly high expression of PD-L1 in tumor microenvironment (TME) can be attributable to the "primary" activation of multiple oncogenic signaling and the "secondary" induction by inflammatory factors such as IFN-γ. Clinically, antibodies targeting PD-1/PD-L1 reinvigorate the "exhausted" T cells in TME and show remarkable objective response and durable remission with acceptable toxicity profile in large numbers of tumors such as melanoma, lymphoma, and mismatch-repair deficient tumors. Nevertheless, most patients are still refractory to anti-PD-1/PD-L1 therapy. Identifying the predictive biomarkers and design rational PD-1-based combination therapy become the priorities in cancer immunotherapy. PD-L1 expression, cytotoxic T lymphocytes infiltration, and tumor mutation burden (TMB) are generally considered as the most important factors affecting the effectiveness of PD-1/PD-L1 blockade. The revolution in cancer immunotherapy achieved by PD-1/PD-L1 blockade offers the paradigm for scientific translation from bench to bedside. The next decades will without doubt witness the renaissance of immunotherapy.
Mobile-edge computing (MEC) and wireless power transfer (WPT) have been recognized as promising techniques in the Internet of Things era to provide massive low-power wireless devices with enhanced ...computation capability and sustainable energy supply. In this paper, we propose a unified MEC-WPT design by considering a wireless powered multiuser MEC system, where a multiantenna access point (AP) (integrated with an MEC server) broadcasts wireless power to charge multiple users and each user node relies on the harvested energy to execute computation tasks. With MEC, these users can execute their respective tasks locally by themselves or offload all or part of them to the AP based on a time-division multiple access protocol. Building on the proposed model, we develop an innovative framework to improve the MEC performance, by jointly optimizing the energy transmit beamforming at the AP, the central processing unit frequencies and the numbers of offloaded bits at the users, as well as the time allocation among users. Under this framework, we address a practical scenario where latency-limited computation is required. In this case, we develop an optimal resource allocation scheme that minimizes the AP's total energy consumption subject to the users' individual computation latency constraints. Leveraging the state-of-the-art optimization techniques, we derive the optimal solution in a semiclosed form. Numerical results demonstrate the merits of the proposed design over alternative benchmark schemes.
This paper studies a new unmanned aerial vehicle (UAV)-enabled wireless power transfer system, where a UAV-mounted mobile energy transmitter is dispatched to deliver wireless energy to a set of ...energy receivers (ERs) at known locations on the ground. We investigate how the UAV should optimally exploit its mobility via trajectory design to maximize the amount of energy transferred to all ERs during a finite charging period. First, we consider the maximization of the sum energy received by all ERs by optimizing the UAV's trajectory subject to its maximum speed constraint. Although this problem is non-convex, we obtain its optimal solution, which shows that the UAV should hover at one single fixed location during the whole charging period. However, the sum-energy maximization incurs a "near-far" fairness issue, where the received energy by the ERs varies significantly with their distances to the UAV's optimal hovering location. To overcome this issue, we consider a different problem to maximize the minimum received energy among all ERs, which, however, is more challenging to solve than the sum-energy maximization. To tackle this problem, we first consider an ideal case by ignoring the UAV's maximum speed constraint, and show that the relaxed problem can be optimally solved via the Lagrange dual method. The obtained trajectory solution implies that the UAV should hover over a set of fixed locations with optimal hovering time allocations among them. Then, for the general case with the UAV's maximum speed constraint considered, we propose a new successive hover-and-fly trajectory motivated by the optimal trajectory in the ideal case and obtain efficient trajectory designs by applying the successive convex programing optimization technique. Finally, numerical results are provided to evaluate the performance of the proposed designs under different setups, as compared with benchmark schemes.
An e-nose or an e-tongue is a group of gas sensors or chemical sensors that simulate human nose or human tongue. Both e-nose and e-tongue have shown great promise and utility in improving assessments ...of food quality characteristics compared with traditional detection methods.
This review summarizes the application of e-nose and e-tongue in determining the quality-related properties of foods. The working principles, applications, and limitations of the sensors employed by electronic noses and electronic tongues were introduced and compared. Widely employed pattern recognition algorithms, including artificial neural network (ANN), convolutional neural network (CNN), principal component analysis (PCA), partial least square regression (PLS), and support vector machine (SVM), were introduced and compared in this review.
Overall, e-nose or e-tongue combining pattern recognition algorithms are very powerful analytical tools, which are relatively low-cost, rapid, and accurate. E-nose and e-tongue are also suitable for both in-line and off-line measurements, which are very useful in monitoring food processing and detecting the end product quality. The user of e-nose and e-tongue need to strictly control sample preparation, sampling, and data processing.
•The application of e-nose and e-tongue in determining food quality were introduced.•Future trends and limitation of e-nose and e-tongue were discussed.•Working principles of sensors and pattern recognition algorithms were introduced.
PDF
