硅片单模光波导与弯道的损耗研究

需积分: 0 2 下载量 73 浏览量 更新于2024-09-09 收藏 839KB PDF 举报
"这篇论文详细探讨了在单模硅-on-绝缘体（SOI）条形波导和弯道中的损耗问题。作者Yurii A. Vlasov和Sharee J. McNab来自IBM T.J. Watson Research Center，他们在标准200mm CMOS制造线上实现了具有亚微米尺寸的单模硅波导，并对其传播和弯曲损耗进行了精确测量。在电信波长1.5微米处，对于TE极化，测得的传播损耗仅为3.6±0.1dB/cm，这表明了极低的损耗性能。90º弯道的损耗分别测量为1微米半径时0.086±0.005dB，而2微米半径时低至0.013±0.005dB，这些记录数值可作为硅微光子组件和电路进一步发展的基准。该研究对光子集成电路领域有重要意义。" 在本文中，作者关注的是硅-on-绝缘体（SOI）平台上的单模条形波导和弯道中的光学损耗。SOI是一种在硅基底上覆盖一层薄氧化硅绝缘层的半导体结构，常用于光电子学应用，因为它提供了良好的光子隔离和微小的光波导结构。单模波导只允许一个光模式传播，从而降低了模式色散和能量损失。 论文的核心发现是通过使用标准的200mm CMOS制造工艺，能够实现非常小的侧壁表面粗糙度，这导致了极低的传播损耗。在1.5微米的电信波长下，TE（电场平行于波导平面）极化的传播损耗仅为3.6dB/cm，这个数值对于高性能光通信系统来说是非常理想的。此外，研究还测量了波导弯道的损耗，特别是在1微米和2微米半径的90º弯道中，损耗分别为0.086dB和0.013dB，这表明了在微尺度下实现低弯曲损耗的技术进步。 这些成果对硅微光子学的发展至关重要，因为低损耗的波导和弯道是构建复杂光子集成电路的基础。这些组件可用于光通信、光计算、传感器和量子信息处理等领域。低损耗的硅微光子元件可以提高系统的整体效率，减少信号衰减，从而实现更远的传输距离和更复杂的光学网络。 参考文献提到了Y. Hibino的一篇文章，该文章讨论了基于二氧化硅的平面光波电路及其应用，进一步证明了SOI技术在光电子领域的广泛影响和重要性。这篇Vlasov和McNab的研究不仅提供了新的技术数据，也为硅微光子学的未来研究和设计提供了重要的实验基准。

Determination of the Radial Distribution of Attenuation in Single-Mode Optical Fibers

2021-02-08 上传
A new method to determine the radial distribution of attenuation in single-mode optical fibers is proposed. As an example, radiation-induced losses in gamma-irradiated germanosilicate fibers are ...

《Perceptual Losses for Real-Time Style Transfer and Super-Resol》

2023-04-14 上传
《Perceptual Losses for Real-Time Style Transfer and Super-Resolution》这篇论文提出了一种结合像素级损失和感知损失的新方法，以提高图像转换任务的效率和质量。 传统的图像转换方法往往依赖于逐像素损失（per...

Single-mode surface-emitting distributed feedback quantum-cascade lasers based on hybrid waveguide structure

2021-02-09 上传
Surface-emitting distributed feedback quantum-... The devices exhibit stable single-mode operation from 90 to 130 K with a side mode suppression ratio above 20 dB. A slope efficiency of 194 mW/A is obta

plt.figure(figsize= (10, 4)) plt.subplot(121) plt.plot(range(n_epoch), cbow_losses, 'r-o', label = 'CBOW Losses') plt.legend() plt.subplot(122) plt.plot(range(n_epoch), sg_losses, 'g-s', label = 'SkipGram Losses') plt.legend() plt.tight_layout()这段代码什么意思

2023-06-08 上传
- `plt.plot(range(n_epoch), cbow_losses, 'r-o', label = 'CBOW Losses')`：在当前子图中绘制CBOW算法的损失值曲线，并用红色圆点表示。其中`range(n_epoch)`表示x轴范围为0到n_epoch-1，`cbow_losses`表示y轴上的...

TverskyLoss = smp.losses.TverskyLoss( mode='binary', log_loss=False, alpha=alpha, beta=beta)

2023-05-02 上传
这是使用Segmentation Models PyTorch库中的Tversky Loss函数创建一个...- mode='binary'表示使用二进制分割模式。 - log_loss=False表示不使用log。 - alpha和beta是Tversky系数的参数，可以根据需要进行调整。

For Gradient Descent, we calculated the gradients for each sample individually. To improve the convergence behavior and to speed up the training process, we now consider batches of gradients. This stabilizes the convergence and prevents the parameters from being push in one direction and directly afterwards being pulled in the opposite direction. Therefore, we can use the gradient descent implementation and just swap the data loader to use the batched variant. The bad thing is that we need to iterate now multiple times over the dataset.帮我补全下面的代码。 net = ... epochs = 10 optimizer = GD(net.parameters(), 0.02) losses, states = ... fig = plot_losses(losses) conf_pltly() init_notebook_mode(connected=False) iplot(fig)

2023-06-12 上传
init_notebook_mode(connected=False) iplot(fig) ``` Note that in this code, we've added a data loader that batches the data into sets of `batch_size` samples. We then iterate over this data loader, ...

def draw_loss(n_epochs, losses, val_losses): epochs_range = range(n_epochs) fig1=plt.figure(figsize=(4,3)) plt.plot(epochs_range, losses, 'orange', label='train loss') plt.plot(epochs_range, val_losses, '-.k', label='test loss') # plt.title('loss') plt.xlabel('Epochs') plt.ylabel('Loss') plt.legend() plt.savefig('experiments/loss_figure.png')

2023-05-22 上传
它接受三个参数：n_epochs 表示训练迭代的总数，losses 表示每个迭代的训练损失，val_losses 表示每个迭代的验证损失。它使用 matplotlib 库来创建一个简单的折线图，其中包含训练和验证损失。每个损失都用不同的...

To do multiple steps, we implement a function the_loop that iterates over a dataloader. It should do a training step per batch for epochs. After one epoch, the loss on the validation set should be calculated.帮我根据这个要求补充完整下面代码中。。。的部分。 def the_loop(net, optimizer, train_loader, val_loader=None, epochs=None, swa_model=None, swa_start=5): if epochs is None: raise Exception("a training duration must be given: set epochs") log_iterval = 1 running_mean = 0. loss = Tensor([0.]).cuda() losses = [] val_losses = [] states = [] i, j = 0, 0 pbar = tqdm(train_loader, desc=f"epoch {i}", postfix={"loss": loss.item(), "step": j}) for i in range(epochs): running_mean = 0. j = 0 pbar.set_description(f"epoch {i}") pbar.refresh() pbar.reset() for j, batch in enumerate(train_loader): # implement training step by # - appending the current states to `states` # - doing a training_step # - appending the current loss to the `losses` list # - update the running_mean for logging ... if j % log_iterval == 0 and j != 0: pbar.set_postfix({"loss": running_mean.item(), "step": j}) running_mean = 0. pbar.update() if i > swa_start and swa_model is not None: swa_model.update_parameters(net) if val_loader is not None: # evaluate the current net on the validation data loader and # collect all losses in the ´val_loss´ list ... pbar.refresh() if val_loader is not None: return losses, states, val_losses return losses, states

2023-06-12 上传
return losses, states, val_losses return losses, states ``` Note that the training step implementation assumes that you have already defined a `loss_function` and that the batch data has an ...

tf.keras.losses.SparseCategoricalCrossentropy

2023-03-28 上传
tf.keras.losses.SparseCategoricalCrossentropy is a loss function that is used in multi-class classification problems where the target variable is integers representing the class labels. It measures ...

validation_losses.append(np.asscalar(val_loss.detach().numpy()))

2023-03-31 上传
As an AI language model, I cannot say for certain what the context and purpose of this code is without more information. However, based on my understanding of Python and PyTorch, here's a brief ...
anecdotist
  • 粉丝: 0
  • 资源: 1
上传资源 快速赚钱

最新资源