SCOPUS 정보 검색 플랫폼

Volumn 72, Issue 23, 1994, Pages 3634-3637

Separation of a mixture of independent signals using time delayed correlations

Author keywords

[No Author keywords available]

Indexed keywords

EID: 0000302959 PISSN: 00319007 EISSN: None Source Type: Journal
DOI: 10.1103/PhysRevLett.72.3634 Document Type: Article

Times cited : (653)

References (11)

2
- 84927808566
- Proc. Xeme Coll. GRETSI, 1017 (1985).
- Cristopher, F.¹ Morrisean, C.²

3
- 0024514737
- (1989) Nature (London) , vol.338 , pp. 334
- Gray, C.¹ Singer, W.²

4
- 0009412585
- edited by, S. Williamson, M. Hoke, G. Stroink, M. Kotani, Plenum, New York
- (1990) Advances in Biomagnetism

5
- 0001183650
- edited by, J. S. Decker, AIP Conf. Proc. No. 151, AIP, New York
- (1986) Neural Network for Computing , pp. 206
- Jutten, C.¹ Herault, J.²

10
- 84927808565
- J. Hirschberg and T. Szende, Pathologische Schreistimme–Stridor und Hustenton im Saïuglinsalter (Fischer, Stuttgard, 1985), p. 16 cry 3 and p. 19 cry 7.

11
- 84927808564
- Even for a situation where the number of signals m is larger than the number of detectors n, i.e., m > n but m <= n (n+1)/2, one could still determine the m times n mixing matrix Cij and the Nm time delayed correlations of the sources < ai(t) ai(t + l τ ) > (l = 0 ,..., N-1 ) from the Nn (n+1)/2 measured correlation functions < Ii(t) Ij( t + l τ ) > (l=0 ,..., N-1). One has only to choose the number N of delays large enough to ensure that the number of measurable variables Nn(n+1)/2 becomes larger than the number of unknowns m(n+N-1). However, the signals ai(t), i = 1 ,..., m, cannot be extracted from Ii(t), i = 1 ,..., n, because the m times n mixing matrix Cij cannot be inverted.

* 이 정보는 Elsevier사의 SCOPUS DB에서 KISTI가 분석하여 추출한 것입니다.