Abdulghani Amir M, Rodriguez-Villegas Esther
Department of Electrical and Electronic Engineering, Imperial College London SW7 2AZ, UK.
Annu Int Conf IEEE Eng Med Biol Soc. 2010;2010:1127-30. doi: 10.1109/IEMBS.2010.5627119.
In a traditional signal processing system sampling is carried out at a frequency which is at least twice the highest frequency component found in the signal. This is in order to guarantee that complete signal recovery is later on possible. The sampled signal can subsequently be subjected to further processing leading to, for example, encryption and compression. This processing can be computationally intensive and, in the case of battery operated systems, unpractically power hungry. Compressive sensing has recently emerged as a new signal sampling paradigm gaining huge attention from the research community. According to this theory it can potentially be possible to sample certain signals at a lower than Nyquist rate without jeopardizing signal recovery. In practical terms this may provide multi-pronged solutions to reduce some systems computational complexity. In this work, information theoretic analysis of real EEG signals is presented that shows the additional benefits of compressive sensing in preserving data privacy. Through this it can then be established generally that compressive sensing not only compresses but also secures while sampling.
在传统的信号处理系统中，采样频率至少是信号中发现的最高频率分量的两倍。这是为了确保稍后能够完全恢复信号。随后，采样信号可以进行进一步处理，例如加密和压缩。这种处理可能计算量很大，而且对于电池供电的系统来说，功耗大得不可行。压缩感知最近作为一种新的信号采样范式出现，受到了研究界的极大关注。根据这一理论，有可能以低于奈奎斯特速率对某些信号进行采样而不影响信号恢复。实际上，这可能提供多方面的解决方案来降低一些系统的计算复杂度。在这项工作中，对真实脑电图信号进行了信息论分析，结果表明压缩感知在保护数据隐私方面具有额外的优势。由此可以普遍确定，压缩感知不仅在采样时进行压缩，还能保障安全。
