Title :
On the equivalence of blind equalizers based on MRE and subspace intersections
Author :
Gesbert, David ; Van der Veen, Alle-Jan ; Paulraj, A.
Author_Institution :
Inf. Syst. Lab., Stanford Univ., CA, USA
fDate :
3/1/1999 12:00:00 AM
Abstract :
Two classes of algorithms for multichannel blind equalization are the mutually referenced equalizer (MRE) method by Gesbert et al. (see ibid. vol.45, p.2307-17, 1997) and the subspace intersection (SSI) method by van der Veen et al. (see ibid., vol. 45, no.1, p.173-90, 1997). Although these methods seem, at first sight, unrelated, we show that certain variants of the SSI and the MRE methods both optimize a new blind criterion, which is referred to as maximum coherence and, thus, are equivalent
Keywords :
blind equalisers; optimisation; parameter space methods; telecommunication channels; MRE; algorithms; blind equalizers; data matrices; maximum coherence; multichannel blind equalization; mutually referenced equalizer; optimization; subspace intersections; Array signal processing; Blind equalizers; Delay; Digital communication; Information systems; Mobile communication; Optimization methods; Sensor arrays; Signal processing algorithms; Wireless communication;
Journal_Title :
Signal Processing, IEEE Transactions on