Blind Signal Detection for Asynchronous Multi-Tag Transmission in Ambient Backscatter Communications

Ambient backscatter communications, a promising technique to realize massive machine type communication (mMTC), has recently attracted great attentions, due to its spectrum-and-energy-efficient characteristics. In ambient backscatter communications, multiple tags will transmit signals asynchronously to the target reader, which however imposes huge challenges to the radio access and signal detection at the reader. To tackle these problems, we propose an independent component analysis (ICA) based blind signal separation, identification and detection scheme. Specifically, each of tag signals is randomly and independently encoded to reduce the collision of the tags. Then a novel ICA algorithm is applied at the reader to separate, identify and detect the signals. The results show that the proposed scheme, compared with existing schemes, provides higher detection accuracy with lower cost even under a large number of tags.