Skip to main content
QUICK REVIEW

[论文解读] Through the Haze: a Non-Convex Approach to Blind Gain Calibration for Linear Random Sensing Models

Valerio Cambareri, Laurent Jacques|arXiv (Cornell University)|Oct 27, 2016
Sparse and Compressive Sensing Techniques参考文献 45被引用 15
一句话总结

该论文提出了一种非凸投影梯度下降算法,用于线性随机感知模型中的盲增益校准,其中需从噪声的乘性测量中恢复未知信号和未知正增益。该方法在未知数数量上实现线性(对数因子内)样本复杂度的可证明精确恢复,且在噪声下收敛解的误差表现平稳衰减。

ABSTRACT

Computational sensing strategies often suffer from calibration errors in the physical implementation of their ideal sensing models. Such uncertainties are typically addressed by using multiple, accurately chosen training signals to recover the missing information on the sensing model, an approach that can be resource-consuming and cumbersome. Conversely, blind calibration does not employ any training signal, but corresponds to a bilinear inverse problem whose algorithmic solution is an open issue. We here address blind calibration as a non-convex problem for linear random sensing models, in which we aim to recover an unknown signal from its projections on sub-Gaussian random vectors, each subject to an unknown positive multiplicative factor (or gain). To solve this optimisation problem we resort to projected gradient descent starting from a suitable, carefully chosen initialisation point. An analysis of this algorithm allows us to show that it converges to the exact solution provided a sample complexity requirement is met, i.e., relating convergence to the amount of information collected during the sensing process. Interestingly, we show that this requirement grows linearly (up to log factors) in the number of unknowns of the problem. This sample complexity is found both in absence of prior information, as well as when subspace priors are available for both the signal and gains, allowing a further reduction of the number of observations required for our recovery guarantees to hold. Moreover, in the presence of noise we show how our descent algorithm yields a solution whose accuracy degrades gracefully with the amount of noise affecting the measurements. Finally, we present some numerical experiments in an imaging context, where our algorithm allows for a simple solution to blind calibration of the gains in a sensor array.

研究动机与目标

  • 解决线性随机感知系统中因未知正增益导致的信号恢复挑战,这是实际实现中的常见问题。
  • 开发一种无需训练信号的(盲)校准方法,联合恢复未知信号与增益,且无需对任一者有先验知识。
  • 在增益满足温和假设且样本复杂度条件下,建立精确恢复的理论保证。
  • 将框架扩展至包含信号与增益的子空间先验,以减少所需测量数。
  • 分析噪声下的鲁棒性,表明随着噪声水平增加,恢复精度的误差衰减表现平稳。

提出的方法

  • 将盲增益校准建模为在联合信号-增益空间上的非凸优化问题,通过最小化数据拟合目标函数实现。
  • 采用精心设计的初始化点的投影梯度下降法,确保收敛至全局最小值。
  • 使用双线性感知模型,其中测量值为信号在子高斯随机向量上的投影,每个投影被未知正增益缩放。
  • 提出一种基于测量一阶矩的结构化初始化方法,利用子高斯向量的统计特性。
  • 在存在先验时,对增益子空间和信号子空间施加投影,以保持迭代点位于可行集合内。
  • 在增益与单位阵偏差受控的假设下,利用集中不等式与随机矩阵的谱性质推导收敛界。

实验结果

研究问题

  • RQ1在无训练信号的情况下,能否实现线性随机感知模型中未知增益的盲校准?在何种条件下可实现?
  • RQ2为保证信号与增益的精确恢复,所需的最小测量数(样本复杂度)是多少?
  • RQ3当信号与增益具有子空间先验时,对所需样本复杂度有何影响?
  • RQ4在噪声测量下,该算法性能如何?随着噪声增加,误差衰减速率如何?
  • RQ5在合理假设下,所提出的非凸优化方法能否以高概率收敛至真实解?

主要发现

  • 当测量数满足 mp = O((m + n) log²(mp)) 时(n = O(log mp) 条件下),所提投影梯度下降算法以高概率收敛至精确解。
  • 样本复杂度在未知数数量上呈线性增长(对数因子内),这在信息论下界意义下为最优。
  • 当对信号与增益均存在子空间先验时,所需测量数进一步减少,显著提升效率。
  • 在噪声情况下,算法的恢复误差随噪声水平增加而平稳衰减,误差界为 O(σ∥z∥),其中 σ 为噪声水平,∥z∥ 为信号范数。
  • 理论分析表明,在温和假设下收敛性可保证:增益在 ℓ∞-范数下有界,且与单位阵偏差不大(即 ∥g − 1m∥∞ ≤ δ,δ 较小)。
  • 成像场景中的数值实验验证了该方法的实用性,能够实现简单且精确的传感器阵列增益盲校准。

更好的研究,从现在开始

从论文设计到论文写作,大幅缩短您的研究时间。

无需绑定信用卡

本解读由 AI 生成,并经人工编辑审核。