TY - JOUR
T1 - Cluster-head based feedback for simplified time reversal prefiltering in ultra-wideband systems
AU - Soleimani, Hossein
AU - Tomasin, Stefano
AU - Alizadeh, Tohid
AU - Shojafar, Mohammad
N1 - Publisher Copyright:
© 2017 Elsevier B.V.
PY - 2017/12
Y1 - 2017/12
N2 - Time-reversal prefiltering (TRP) technique for impulse radio (IR) ultra wide-band (UWB) systems requires a large amount of feedback to transmit the channel impulse response from the receiver to the transmitter. In this paper, we propose a new feedback design based on vector quantization. We use a machine learning algorithm to cluster the estimated channels into several groups and to select the channel cluster heads (CCHs) for feedback. In particular, CCHs and their labels are recorded at both side of the UWB transceivers and the label of the most similar CCH to the estimated channel is fed back to the transmitter. Finally, the TRP is applied using the feedback CCH. The proposed digital feedback provides three main advantages: (1) it significantly reduces the dedicated bandwidth required for feedback; (2) it considerably improves the speed of transceivers; and, (3) it is robust to noise in the feedback channel since few bytes are required to send the codes that can be heavily error protected. Numerical results on standard UWB channel models are discussed, showing the advantage of the proposed solution.
AB - Time-reversal prefiltering (TRP) technique for impulse radio (IR) ultra wide-band (UWB) systems requires a large amount of feedback to transmit the channel impulse response from the receiver to the transmitter. In this paper, we propose a new feedback design based on vector quantization. We use a machine learning algorithm to cluster the estimated channels into several groups and to select the channel cluster heads (CCHs) for feedback. In particular, CCHs and their labels are recorded at both side of the UWB transceivers and the label of the most similar CCH to the estimated channel is fed back to the transmitter. Finally, the TRP is applied using the feedback CCH. The proposed digital feedback provides three main advantages: (1) it significantly reduces the dedicated bandwidth required for feedback; (2) it considerably improves the speed of transceivers; and, (3) it is robust to noise in the feedback channel since few bytes are required to send the codes that can be heavily error protected. Numerical results on standard UWB channel models are discussed, showing the advantage of the proposed solution.
KW - Clustering channels
KW - Digital feedback
KW - IEEE 802.15.SG3 channel model
KW - Machine learning
KW - Prefiltering
KW - Ultra wide-band (UWB)
UR - http://www.scopus.com/inward/record.url?scp=85030175017&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85030175017&partnerID=8YFLogxK
U2 - 10.1016/j.phycom.2017.09.005
DO - 10.1016/j.phycom.2017.09.005
M3 - Article
AN - SCOPUS:85030175017
SN - 1874-4907
VL - 25
SP - 100
EP - 109
JO - Physical Communication
JF - Physical Communication
ER -