在带宽有限的光学无线通信系统中部署神经网络均衡器的最小化部署与知识蒸
Yiming Zhu1, Yuan Wei1, Chaoxu Chen1
1Key Laboratory for Information Science of Electromagnetic Waves (MoE), Department of Communication Science and Engineering, Fudan University, Shanghai 200433, China.
Sensors (Basel, Switzerland)
|March 13, 2024
概括
一个新的1D卷积神经网络 (CNN) 均衡器,使用从循环神经网络 (RNN) 的知识蒸 (KD) 来训练,显著加快光通信系统的速度. 这种方法提高了97%的均等化速度,同时保持了高性能.
科学领域:
- 光学通信系统 光学通信系统
- 机器学习用于信号处理.
- 深度学习架构是一种深度学习架构.
背景情况:
- 循环神经网络 (RNN),特别是双向封闭循环单元 (biGRU),在光通信中擅长处理非线性损伤和符号间干扰 (ISI).
- 然而,双GRU固有的递归结构限制了计算并行化,导致较低的等分率.
- 需要更快的等分方法,而不会牺牲光通信系统的性能.
研究的目的:
- 提出和评估一个极简的1D卷积神经网络 (CNN) 均衡器.
- 通过从biGRU模型中使用知识蒸 (KD) 来提高光通信系统的等分速度.
- 调查KD在神经网络模型压缩和加速回归问题的有效性.
主要方法:
- 应用知识蒸 (KD) 来训练1D CNN等分器,使用预训练的biGRU模型作为教师.
- 将KD训练的1D CNN与原来的biGRU和没有KD训练的1D CNN的性能进行了比较.
- 关键性能指标包括Q因子 (信号质量) 和均等速度 (速度).
主要成果:
- 与没有KD的1D CNN相比,KD训练的1D CNN在Q因子上实现了1dB的增加.
- KD 提高了 1D CNN 的接收光功率 (RoP) 灵敏度,提高了 0.89 dB.
- 与biGRU相比,1D CNN等分器显示计算时间减少了97%,可训练参数减少了99.3%,只有0.5dB的Q因子处罚.
结论:
- 一个极简的1D CNN等分器,通过biGRU的KD进行优化,为高速光通信系统提供了一个有前途的解决方案.
- 这种方法在保持高信号质量的同时显著加快了等级,使其适合实际部署.
- 该研究强调了KD在光通信信号处理中的回归任务中的模型压缩和性能增强方面的有效性.
相关概念视频
01:18Reducing Line Loss
152
In a three-phase circuit, line loss is an indicator of energy dissipated as heat due to the resistance of transmission lines. To address this, incorporating transformers into the system—a step-up transformer at the source and a step-down transformer at the load—is a strategic solution. Two three-phase transformers are introduced to improve this.
With a step-up transformer at the source, the voltage is increased, thereby reducing the current in the transmission lines since power loss...
With a step-up transformer at the source, the voltage is increased, thereby reducing the current in the transmission lines since power loss...
152
01:26Linear Approximation in Frequency Domain
89
Linear systems are characterized by two main properties: superposition and homogeneity. Superposition allows the response to multiple inputs to be the sum of the responses to each individual input. Homogeneity ensures that scaling an input by a scalar results in the response being scaled by the same scalar.
In contrast, nonlinear systems do not inherently possess these properties. However, for small deviations around an operating point, a nonlinear system can often be approximated as linear....
In contrast, nonlinear systems do not inherently possess these properties. However, for small deviations around an operating point, a nonlinear system can often be approximated as linear....
89
01:21Linear Approximation in Time Domain
81
Nonlinear systems often require sophisticated approaches for accurate modeling and analysis, with state-space representation being particularly effective. This method is especially useful for systems where variables and parameters vary with time or operating conditions, such as in a simple pendulum or a translational mechanical system with nonlinear springs.
For a simple pendulum with a mass evenly distributed along its length and the center of mass located at half the pendulum's length,...
For a simple pendulum with a mass evenly distributed along its length and the center of mass located at half the pendulum's length,...
81
01:20Downsampling
155
When considering a sampled sequence with zero values between sampling instants, one can replace it by taking every N-th value of the sequence. At these integer multiples of N, the original and sampled sequences coincide. This process, known as decimation, involves extracting every N-th sample from a sequence, thereby creating a more efficient sequence.
The Fourier transform of the decimated sequence reveals a combination of scaled and shifted versions of the original spectrum. This...
The Fourier transform of the decimated sequence reveals a combination of scaled and shifted versions of the original spectrum. This...
155
00:59Propagation of Uncertainty from Random Error
682
An experiment often consists of more than a single step. In this case, measurements at each step give rise to uncertainty. Because the measurements occur in successive steps, the uncertainty in one step necessarily contributes to that in the subsequent step. As we perform statistical analysis on these types of experiments, we must learn to account for the propagation of uncertainty from one step to the next. The propagation of uncertainty depends on the type of arithmetic operation performed on...
682
01:25Network Function of a Circuit
290
Frequency response analysis in electrical circuits provides vital insights into a circuit's behavior as the frequency of the input signal changes. The transfer function, a mathematical tool, is instrumental in understanding this behavior. It defines the relationship between phasor output and input and comes in four types: voltage gain, current gain, transfer impedance, and transfer admittance. The critical components of the transfer function are the poles and zeros.
290