A generalized adaptive code acquisition scheme is proposed by using a multi-layer structure. The backpropagation network (BPN) is used in modeling the multilayer system. The timing offset of the received code can be detected by combining the hidden layer weights after an adaptation period. The characteristics of mean squared error (MSE) and weight distribution for the generalized system are analyzed, and the effects of input and hidden nodes are investigated. The generalized system provides a high detection probability and flexibility in acquisition system design.
Published in:
Communication Systems, 2002. ICCS 2002. The 8th International Conference on
(Volume:1
)
Date of Conference: 25-28 Nov. 2002