Noncoherent Frequency-Shift Keying for Ambient Backscatter Over OFDM Signals

Abstract

In this paper, we investigate frequency shift keying (FSK) over ambient orthogonal frequency division multiplexed (OFDM) signals. By cycling through a sequence of antenna loads providing different phase shifts at the tag, we are able to unidirectionally shift the ambient OFDM spectrum either up or down in frequency to disjoint subsets of the subcarriers allowing the implementation of FSK. We exploit the guard bands and the orthogonality of the OFDM subcarriers to avoid both direct-link and adjacent channel interference. Different from energy detection based techniques that suffer from asymmetric error probabilities and rely on signal-to-noise ratio (SNR) dependent detection thresholds, the proposed scheme has symmetric error probabilities and allows simple detection without the need for a threshold. We present both binary and four-ary schemes, and analyze the error performance of the optimal noncoherent detectors. For the binary scheme, we obtain exact expressions for the average probability of error, while for the four-ary scheme, a union bound is used to characterize the error performance. Single and multi-antenna receivers are considered, and their performance is analyzed. Finally, we present simulation results to corroborate our analysis and investigate the effects of multiple system parameters. The results show that the proposed scheme outperforms the baseline energy detection based schemes available in the literature in various scenarios by up to 5 dB.

Keywords

• Ambient backscatter
• Internet of Things
• green communications
• performance analysis
• OFDM