Hero Image
WSPMax Read

Requirement-Driven Magnetic Beamforming for MIMO Wireless Power Transfer Optimization Motivation For the general MIMO setup, MultiSpot doesn’t consider peak current/voltage constraints. Yang et al. studied the magnetic beamforming problem in an MIMO MRC-WPT system to find all the boundary points of the multi-user power region with the peak current and voltage constraints for each TX. They simplified the problem by ignoring the mutual inductance among RXs. These researches lack the ability to control the power distribution among receivers, and could not be applied to the scenario when different RXs have different requirements of power which can be denoted as weight factors. Contribution We formulate the requirement-driven magnetic beamforming design in the MIMO MRC-WPT system as a weighted sum-power maximization problem. We consider the peak current/voltage constraints, and provides a high efficient, provable optimal solution. We discuss the WSPMax problem with limited power budget where the peak current/voltage constraints are ignorable. We introduce a close-form theoretical bound of the WSPMax problem, and demonstrate that the power transfer efficiency maximization problem can be solved through TX-only solution without any feedback from the RXs, despite the fact that the beamforming is impacted by the interactions between TXs and RXs.

    Friday, December 13, 2024 | 3 minutes Read
    Hero Image
    Optimization Methods

    Lagrangian relaxation The explanation I provide is based on Lagrangian relaxation-Wikipedia and Chapter 12 Lagrangian Relaxation-ens-lyon Lagrangian relaxation is a relaxation method which approximates a difficult problem of constrained optimization by a simpler problem. The method penalizes violations of inequality constraints using a Lagrange multiplier, which imposes a cost on violations. These added costs are used instead of the strict inequality constraints in the optimization. In practice, this relaxed problem can often be solved more easily than the original problem.

      Thursday, December 12, 2024 | 11 minutes Read
      Hero Image
      Building issues

      Issues 1. Built-in Shortcode leads to Hugo timeout. Because of the networks, I can’t send requests to Twitter, YouTube, etc. So I delete the contents about Twitter and YouTube. They are in content\posts\category\sub-category\rich-content\index.md, content\posts\category\sub-category\rich-content\index.bn.md, hugo.yaml shareButtons:, data\en\sections\about.yaml socialLinks:. 2. Only readme when deploying to github pages. Need to set your deployment branch to gh-pages.

        Sunday, November 24, 2024 | 1 minute Read
        Hero Image
        Matrix Base

        0.1 Determinant Theorem: $ det(AB) = det(A)det(B) $ Let $ A, B $ be a square matrices of order n. Let $ det(A) $ be the determinant of $ A $. Let $ AB $ be the matrix product of $ A $ and $ B $. Then: $ det(AB) = det(A)det(B) $ Proof The proof I provide is based on Determinant of Matrix Product Consider two cases: $ A $ is singular. $ A $ is nonsingular.

          Wednesday, November 20, 2024 | 56 minutes Read
          Hero Image
          MagMIMOHotspot Read

          Magnetic MIMO: how to charge your phone in your pocket Motivation Current wireless chargers need to remember to regularly charge our mobile phones which is not the wireless charging we hoped for. We would like to have our cell phones charged in our pockets, and never again worry about forgetting to charge the phone. MIMO RF techniques power phones remotely, however, delivering a large amount of power via radiation can cause local heating inside the human body. Contribution They propose a novel design that focuses the magnetic flux from multiple coils in a steerable beam and points it at the phone, in a manner analogous to multi-antenna beamforming in wireless communications. The design can charge unmodified smart phones at distances up to 40 cm, and works regardless of the phone orientation with respect to the charging pad. Magnetic-Beamforming Figure 1: Illustration of the analogies between standard MIMO and MagMIMO. For a 2-antenna MIMO transmitter.

            Thursday, November 7, 2024 | 7 minutes Read
            Hero Image
            Circuit Base

            Basic circuit Charge It is defined in terms of the ampere by counting the total charge that passes through an arbitrary cross section of a wire during an interval of one second. In the SI system, the fundamental unit of charge is the coulomb (C). Current We define the current at a specific point and flowing in a specified direction as the instantaneous rate at which net positive charge is moving past that point in the specified direction. Current is symbolized by I or i, and so

              Thursday, November 7, 2024 | 3 minutes Read
              Hero Image
              First glance

              Scheduling Layer PTE Optimization Requirement-Driven Magnetic Beamforming for MIMO Wireless Power Transfer Optimization Motivation: In MIMO MRC-WPT system, we need to consider the scenario when different RXs have different requirements of power. Problem formulation: weighted sum-power maximization $WSPMax = max_{{i_n^{tx}}} \sum_q w_q |i^{rx}_q|^2 r^{rx}_q $ Subject to: $$ \sum_n \left|i_n^{tx}\right|^2 r_n^{tx} + \sum_q \left|i_q^{rx}\right|^2 r_q^{rx} \leq P_{\text{max}}, \ $$ $$ \left|i_n^{tx}\right|^2 \leq A_n^{tx}, \quad \forall 1 \leq n \leq N, \ $$

                Thursday, October 31, 2024 | 5 minutes Read
                Hero Image
                Multimodal-fusion research

                Fusion stage Figure 1: Mind map of Deep Multimodal Learning Inference mechanism for multi-sensors fusion Figure 2: Mind map of A comparative review on multi-modal sensors fusion based on deep learning

                  Wednesday, October 16, 2024 | 1 minute Read
                  Hero Image
                  Recommendation Progress

                  个人情况 本科： 重庆大学 专业： 计算机科学与技术 排名： 1.3% (rank 3) 英语： 六级 547 荣誉： 国家奖学金、重庆市三好学生 竞赛： 无竞赛 科研： 有 CV 领域相关经历，面试时有 NIPS 二作在投 最终去向： 中国科学技术大学-计算机科学与技术学院 夏令营情况 院校 类型 入营 offer 南京大学-计算机科学与技术学院 硕士 √ 候补 清华大学-深圳国际研究生院(数据与信息研究院人工智能项目) 硕士 × 清华大学-网络研究院 硕士 × 南京大学-人工智能学院 硕士 × 中国人民大学-高瓴人工智能学院 直博 √ 时间冲突，放弃参营 上海交通大学-人工智能学院 直博 √ 不合格（斩杀） 中国科学技术大学-计算机科学与技术学院 硕士 √ 学硕 南京大学-计算机科学与技术学院 线上测试过了，线下机试 0 分。。。 南京大学-人工智能学院 入了计院，不让入人工智能学院？ 上海交通大学-人工智能学院 没有提前联系好老师，科研积累太少。 预推免情况 院校 类型 入营 offer 清华大学-软件学院 专硕 √ × 浙江大学-计算机科学与技术学院 硕士 √ 专硕 总结：入营基本都能入，但自身实力差。南大机试报零，面试差导致上交直博不合格。

                    Thursday, October 10, 2024 | 1 minute Read
                    Hero Image
                    C language lecture 1

                    怎么和计算机打交道 Figure 1: 早期将穿孔纸带输入计算机 Figure 2: 使用汇编语言编程 Figure 3: 使用高级语言C语言编程 为什么要学 C 语言 基础性强：C 语言是一门底层语言，靠近计算机硬件。学习 C 语言能够帮助你理解计算机是如何工作的，如内存管理、指针操作和硬件交互等。很多高级语言（如 C++、Python）都基于 C 语言。 高效性能：C 语言生成的代码通常比很多高级语言效率更高，特别适合对性能要求高的场景，比如实时系统、游戏引擎等。 。。。 怎么开始学习 C/C++ 语言 首先需要一个写代码的好地方，推荐 VS Code 安装 VS Code 安装 C/C++ 扩展 安装 MinGW-w64 编译工具，并检查 向编程世界打声招呼 #include <iostream> using namespace std; int main() { cout << "Hello, World!" << endl; return 0; }

                      Friday, September 13, 2024 | 1 minute Read
                      Hero Image
                      transformer Read

                      Questions What’s the number of operations required to relate signals from two arbitrary input or output positions grows in the distance between positions ? Extended Learning Attention mechanisms with a recurrent network. Bahdanau, Dzmitry. “Neural machine translation by jointly learning to align and translate.” arXiv preprint arXiv:1409.0473 (2014). Use convolutional neural networks to compute hidden representations in parallel. Gehring, Jonas, et al. “Convolutional sequence to sequence learning.” International conference on machine learning. PMLR, 2017. Why is difficult to learn dependencies between distant positions. Hochreiter, Sepp, et al. “Gradient flow in recurrent nets: the difficulty of learning long-term dependencies.” (2001). Self-attention. Cheng, Jianpeng. “Long short-term memory-networks for machine reading.” arXiv preprint arXiv:1601.06733 (2016).

                        Thursday, September 12, 2024 | 1 minute Read
                        Hero Image
                        DINO Read

                        Questions What’s the function of centering $ C $ and sharpening parameter $ \tau_t $ ? Extended Learning Multi-crop training. Caron, Mathilde, et al. “Unsupervised learning of visual features by contrasting cluster assignments.” Advances in neural information processing systems 33 (2020): 9912-9924. Use a noise contrastive estimator (NCE) to compare instances instead of classifying them. Wu, Zhirong, et al. “Unsupervised feature learning via non-parametric instance discrimination.” Proceedings of the IEEE conference on computer vision and pattern recognition. 2018. Mean Teacher self-distillation. Tarvainen, Antti, and Harri Valpola. “Mean teachers are better role models: Weight-averaged consistency targets improve semi-supervised deep learning results.” Advances in neural information processing systems 30 (2017). knowledge distillation. Buciluǎ, Cristian, Rich Caruana, and Alexandru Niculescu-Mizil. “Model compression.” Proceedings of the 12th ACM SIGKDD international conference on Knowledge discovery and data mining. 2006. Extra learnable token in ViT. Dosovitskiy, Alexey. “An image is worth 16x16 words: Transformers for image recognition at scale.” arXiv preprint arXiv:2010.11929 (2020). Devlin, Jacob. “Bert: Pre-training of deep bidirectional transformers for language understanding.” arXiv preprint arXiv:1810.04805 (2018). Attention mechanism. Bahdanau, Dzmitry. “Neural machine translation by jointly learning to align and translate.” arXiv preprint arXiv:1409.0473 (2014).

                          Friday, September 6, 2024 | 1 minute Read