DocumentCode :
1739458
Title :
Guard sequence optimization for block transmission over linear dispersive channels
Author :
Al-Dhahir, Naofal ; Diggavi, Suhas N.
Author_Institution :
AT&T Labs., Florham Park, NJ, USA
Volume :
2
fYear :
2000
fDate :
2000
Firstpage :
970
Abstract :
We show that the optimum length-ν guard sequence for block transmission over a linear, noisy, and dispersive channel with memory ν is a linear combination of the N information symbols of the block. The channel block throughput is maximized through a joint numerical optimization of Nν linear combination coefficients determining the guard sequence and the N-dimensional auto-correlation matrix of the information symbols, subject to a total average energy constraint on the information and guard symbols. A closed-form solution for the optimum guard sequence is derived for the special case of high SNR. The achievable channel block throughput with the optimum guard sequence is compared with that achievable with two common guard sequence types, namely zero stuffing and cyclic prefix
Keywords :
correlation methods; dispersive channels; matrix algebra; noise; optimisation; sequences; signal processing; DMT; auto-correlation matrix; block by block processing; block transmission; channel block throughput; closed-form solution; cyclic prefix; guard sequence optimization; guard symbols; high SNR; information symbols; linear combination coefficients; linear dispersive channels; memory; noisy channel; numerical optimization; optimum guard sequence; optimum length guard sequence; packet-based communication networks; total average energy constraint; vector coding; zero stuffing; Autocorrelation; Closed-form solution; Constraint optimization; Delay; Dispersion; Interference; Laboratories; Throughput; Transmitters; Wideband;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Global Telecommunications Conference, 2000. GLOBECOM '00. IEEE
Conference_Location :
San Francisco, CA
Print_ISBN :
0-7803-6451-1
Type :
conf
DOI :
10.1109/GLOCOM.2000.891282
Filename :
891282
Link To Document :
بازگشت