Local minima effects on the transient performance of non-linear blind equalizers
Tóm tắt
The computational requirements and the transient performance of several non-linear blind equalizers are compared in the case of transmission over linear and non-linear channels. The multilayer perceptron (MLP), the radial-basis-function network (RBF), the polynomial perceptron (PP) and two recently proposed non-linear structures (see Destro Filho, J.B., et al., Proc. GLOBECOM'96, p.196-200, 1996; Proc. GLOBECOM'99, 1999) are simulated. These equalizers are also compared to two classical benchmarks: the Volterra filter and Godard algorithm. A criterion for assessing the impact of parameter initialization (filter coefficients and synaptic weights) on the transient performance is proposed and evaluated. The results establish guidelines for choosing a particular non-linear blind equalizer when the trade-off between robustness to local minima problems and computational requirements must be satisfied.
Từ khóa
#Blind equalizers #Hydrogen #Multilayer perceptrons #Filters #Polynomials #Satellite communication #Neural networks #Steady-state #Quadratic programming #Computational modelingTài liệu tham khảo
10.1109/ICASSP.1998.679589
bouchired, 1998, Equalisation of satellite mobile channels with neural network techniques, Space Communications 15, 209
callender, 1994, A comparison of six different non-linear equalisation techniques for digital communications systems, Proc EUSIPCO 94, 1524
10.1109/ICASSP.1992.226422
0
10.1109/JSAC.1983.1145885
10.1109/TCOM.1980.1094608
xiang, 1994, Polynomial perceptrons and their applications to fading channel equalization and co-channel interference suppression, IEEE Trans Sig, 42, 2470, 10.1109/78.317868
10.1109/72.363440
10.1109/EURCON.2001.938173
10.1109/ICASSP.1999.760654
yates, 2000, Low-cost wireless data transmission, IEEE Signal Proc Magazine, 10.1109/79.841734
georgoulakis, 2000, Blind and semi-blind equalization using hidden Markov models and clustering techniques, Signal Processing 80, 1795, 10.1016/S0165-1684(00)00089-X
10.1109/72.728394
10.1016/0165-1684(94)90206-2
10.1109/72.238312
10.1109/GLOCOM.1999.827582
10.1109/GLOCOM.1996.594359
10.1109/72.774271
10.1109/23.736527