Acoustic Echo Cancellation Using Frequency-Domain Spline Identification

Zakharov, Yuriy; Tozer, T.C.; Pearce, D.A.J.

doi:10.1109/tsp.2006.887150

Cited by 4 publications

(2 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…ε > 0 is a small number, I is the identity matrix, andz b is a vector with elementsz b (n) defined in (20). The channel frequency response is then estimated as the BEM [39]:…”

Section: K=0mentioning

confidence: 99%

“…Find the fine Doppler estimatesâ 1,b andâ 2,b as in (39) and 40Make the fine Doppler correction of y b (n) to obtain the signalŷ b (n) as in (38) Find the SNR estimate SNR b as in (17) end Compute the weight coefficients w 1 and w 2 as in (18) Produce the combined signal y(n) fromŷ 1 (n) andŷ 2 (n) as in (16) Produce (ii) Refine {τ max , F max }, e.g., using the parabolic interpolation of the peak as described in [5] 4 (iii) Compute the coarse Doppler estimatesâ 1 andâ 2 as in (13) and 14, respectively Produce the combined fine Doppler estimatesâ 1 andâ 2 fromâ 1,b andâ 2,b as in (19) 11 end…”

mentioning

confidence: 99%

See 1 more Smart Citation

Data Packet Structure and Modem Design for Dynamic Underwater Acoustic Channels

Zakharov

Henson

Diamant

et al. 2019

IEEE J. Oceanic Eng.

Self Cite

View full text Add to dashboard Cite

In underwater acoustic (UWA) communications, the propagated signal undergoes severe Doppler and multipath distortions. The Doppler estimation/compensation and channel estimation/equalization techniques required to deal with these distortions contribute significantly to the overall complexity of UWA modems. In this paper, we propose a data packet structure for high data rate transmission in time-varying UWA channels with the channel dynamic modelled by velocity and acceleration between the transmitter and receiver. The data packet consists of superimposed data and periodic pilot sequences. The superposition allows achievement of a high spectral efficiency for data transmission. The repeated pilot symbols allow the use of a low-complexity multi-branch autocorrelation method for coarse estimation of Doppler parameters related to the velocity and acceleration. To refine these estimates, we further propose a lowcomplexity fine Doppler estimator based on dichotomous iterations. We also present a low complexity frequency domain channel estimator exploiting the channel sparsity. The proposed modem design has been evaluated using simulation and practical sea trials. The experiments demonstrate high detection performance of the proposed design, in particular in comparison with a more traditional design that ignores the acceleration between the transmitter and receiver.

show abstract

“…ε > 0 is a small number, I is the identity matrix, andz b is a vector with elementsz b (n) defined in (20). The channel frequency response is then estimated as the BEM [39]:…”

Section: K=0mentioning

confidence: 99%