eduzhai > Applied Sciences > Engineering >

Low-complexity Architecture for AR(1) Inference

  • king
  • (0) Download
  • 20210507
  • Save

... pages left unread,continue reading

Document pages: 7 pages

Abstract: In this Letter, we propose a low-complexity estimator for the correlationcoefficient based on the signed $ operatorname{AR}(1)$ process. The introducedapproximation is suitable for implementation in low-power hardwarearchitectures. Monte Carlo simulations reveal that the proposed estimatorperforms comparably to the competing methods in literature with maximum errorin order of $10^{-2}$. However, the hardware implementation of the introducedmethod presents considerable advantages in several relevant metrics, offeringmore than 95 reduction in dynamic power and doubling the maximum operatingfrequency when compared to the reference method.

Please select stars to rate!


0 comments Sign in to leave a comment.

    Data loading, please wait...