A new broadband DC-bias included model for radio frequency (RF) power transistors is presented in this paper. The proposed method utilizes the canonical section-wise piecewise linear (CSWPL) model framework in order to incorporate both frequency as well as bias information simultaneously. Detailed descriptions of the fundamental theory of the developed model are provided, along with their validation through experiments. The model is implemented in commercial software and validated using DC and RF tests with measured load-pull data from a 10-W GaN device. In contrast to standard and bias-included CSWPL models, this model can predict transistor behavior under a variety of bias voltages and frequencies using only one set of parameters, thereby significantly reducing model complexity. Additionally, the proposed model is applied to a multi-octave PA design to provide further validation. Measurements made on the realized PA are compared to simulations based on the proposed model. Validity of the extracted model is confirmed by the agreement between measurements and simulations.

Recent advances in few-shot novel-view synthesis based on 3D Gaussian Splatting (3DGS) have shown remarkable progress. Existing methods usually rely on carefully designed geometric regularizers to reinforce geometric supervision; however, applying multiple regularizers consistently across scenes is hard to tune and often degrades robustness. Consequently, generating reliable geometry from extremely sparse viewpoints remains a key challenge. To overcome this limitation, we introduce SREGS, a framework tailored for few-shot reconstruction whose contributions focus on two aspects: explicitly consistent geometry and multi-scale depth-guided optimization. Specifically, to explicitly optimize reconstruction consistency, we initialize the point cloud with 2D Gaussians, thereby enhancing depth consistency for the same Gaussian observed from different views. Secondly, we employ region-adaptive rapid densificationn to fill under-covered regions with additional representations, while an opacity-aware noise term injects stochasticity into each Gaussian to boost exploration in under-observed areas. In addition, to strengthen geometric refinement of the radiance field, we impose multi-scale depth constraints based on a monocular depth prior, performing geometric refinement from global to local scales and ensuring highly accurate reconstruction. Extensive experiments on LLFF, MipNeRF360, and Blender show that SREGS achieves higher synthesis quality with lower computational cost and demonstrates robust performance. The code is available at:https://github.com/LeeXiaoTong1/SREGS.

Recently, various strategies for distributed training of large language models (LLMs) have been proposed. By categorizing them into basic strategies and composite strategies, we have discovered that existing basic strategies provide limited options in specific scenarios, leaving considerable room for optimization in training speed. In this paper, we rethink the impact of memory and communication costs on the training speed of LLMs, taking into account the impact of intra- and inter-group communication performance disparities, and then propose a new set of basic strategies named the \textbf{Pa}rtial \textbf{R}edundancy \textbf{O}ptimizer (PaRO). PaRO Data Parallelism (PaRO-DP) accelerates LLM training through refined model state partitioning and tailored training procedures. At the same time, PaRO Collective Communications (PaRO-CC) speeds up collective communication operations by rearranging the topology. We also propose a guideline for choosing different DP strategies based on simple quantitative calculations, which yields minimal ranking errors. Our experiments demonstrate that PaRO improves the training speed of LLMs by up to 266\% that of ZeRO-3 as basic DP strategies. Moreover, employing PaRO-CC independently for model parallel strategies, such as Megatron, can also boost the training speed by 17\%.

In this article, a multi-objective particle swarm optimization (MPSO) method is presented for the design of a sequential load modulation balanced amplifier (SLMBA). Based on the proposed method, the matching networks of the control amplifier (CA) and balanced amplifier (BA) are optimized separately in order to achieve optimal load modulation behavior. Furthermore, the effect of the phase offset line of the SLMBA is analyzed and optimized. In order to validate the proposed method, a SLMBA with a frequency range of 1.8 GHz to 2.1 GHz was implemented and measured. Consequently, it is capable of achieving a saturation drain efficiency (DE) of 70.1%-74.3%, a saturated output power of 43.7 dBm, and a DE of 52.3%-57.9% with 10.5-dB output back-off (OBO). In order to improve the linearity of the manufactured SLMBA, a digital pre-distortion method has been implemented, and a 20 MHz 5GNR signal has been used to test the device with good results.

In this article, an overall optimization strategy of power amplifier (PA) based on Bayesian algorithm is proposed to perform multiple multiobjective optimization design of sequential load modulated balanced amplifier (SLMBA). Specifically, by combining the programming language in MATLAB with commercial electronic design automation (EDA) software, such as advanced design system (ADS), the joint optimization process can be achieved. Then, the complex load modulation can be achieved and high-drain efficiency (DE) at various output power back-off (OBO) levels can be obtained by using the proposed overall optimization strategy, which proves the superiority of the combined optimization strategy in optimizing SLMBA compared with the optimization algorithms embedded in ADS. To verify the proposed optimization strategy, a prototype operating at 1.8–2.1 GHz was demonstrated and implemented using Gallium Nitride (GaN) transistors. A high-back-off efficiency SLMBA is simulated and measured, which the measured saturated total output power reaches 42.7–43.5 dBm with 75.8%–81.2% DE and 55%–62.5% DE at 10-dB power back-off. After that, digital pre-distortion (DPD) is implemented to further improve the linearity of the designed SLMBA with 20-MHz 5G NR signal, and good performance is achieved.

In the cloud datacenter, the resource utilization of different virtual machine (VM) and physical machine (PM) varies with time and it may lead to SLO violation and then degrade the application performance. In order to minimize the probability of SLO violation, load balancing is used to dynamically migrate VMs from overloaded PMs to underloaded PMs. Previous load balancing methods fail to achieve long term load balance. To address this problem, in this paper, we propose different load balancing methods and evaluate their performance on several metrics. We use the Fast Fourier Transform (FFT) method, an improved FFT method considering more frequencies in FFT and the long short term memory (LSTM) machine learning model to predict the resource utilization of VM and PM in the future. LSTM can always achieve the best prediction performance in the prediction. Taking advantage of the ML technique, we then propose a heuristic based method and a reinforcement learning (RL) based method relying on ML workload prediction to generate the VM migration plan in the datacenter. We conduct experiments in both trace-driven simulation (based on Google cluster trace, PlanetLab trace, Worldcup trace) and real implementation in terms of the SLO violation rate, the number of migrations and overhead. The experimental results show that the workload prediction helps reduce the SLO violation rate and/or the number of migrations, which improves the load balance performance in a datacenter. Also, the RL based VM migration method outperforms the heuristic based method in a heavily loaded system but does not show obvious advantages in a lightly loaded system.

Detecting and tracking objects in 3D scenes play crucial roles in autonomous driving. Successfully recognizing objects through space and time hinges on a strong detector and a reliable association scheme. Recent 3D detection and tracking approaches widely represent objects as points when associating detection results with trajectories. Despite the demonstrated success, these approaches do not fully exploit the rich appearance information of objects. In this paper, we present a conceptually simple yet effective algorithm, named AlphaTrack, which considers both the location and appearance changes to perform joint 3D object detection and tracking. To achieve this, we propose a cross-modal fusion scheme that fuses camera appearance feature with LiDAR feature to facilitate 3D detection and tracking. We further attach an additional branch to the 3D detector to output instance-aware appearance embedding, which significantly improves tracking performance with our designed association mechanisms. Extensive validations on large-scale autonomous driving dataset demonstrate the effectiveness of the proposed algorithm in comparison with state-of-the-art approaches. Notably, the proposed algorithm ranks first on the nuScenes tracking leaderboard to date.