The emerging machine learning technique called support vector machines is p
roposed as a method for performing nonlinear equalization in communication
systems. The support vector machine has the advantage that a smaller number
of parameters for the model can he identified in a manner that does not re
quire the extent of prior information or heuristic assumptions that some pr
evious techniques require. Furthermore, the optimization method of a suppor
t vector machine is quadratic programming, which is a well studied and unde
rstood mathematical programming technique,
Support vector machine simulations are carried out on nonlinear problems pr
eviously studied by other researchers using neural networks. This allows in
itial comparison against other techniques to determine the feasibility of u
sing the proposed method for nonlinear detection. Results show that support
vector machines perform as well as neural networks on the nonlinear proble
ms investigated,
A method is then proposed to introduce derision feedback processing to supp
ort vector machines to address the fact that intersymbol interference (ISI)
data generates input vectors having temporal correlation, whereas a standa
rd support vector machine assumes independent input vectors, presenting the
problem from the viewpoint of the pattern space illustrates the utility of
a bank of support vector machines. This approach yields a nonlinear proces
sing method that Is somewhat different than the nonlinear decision feedback
method whereby the linear feedback filter of the decision feedback equaliz
er is replaced by a Volterra filter. A simulation using a linear system sho
ws that the proposed method performs equally to a conventional decision fee
dback equalizer for this problem.