现代电子技术
現代電子技術
현대전자기술
MODERN ELECTRONICS TECHNIQUE
2013年
9期
40-42
,共3页
正交基神经网络%非线性%卫星信道%预失真
正交基神經網絡%非線性%衛星信道%預失真
정교기신경망락%비선성%위성신도%예실진
orthogonal basis neural network%nonlinear%satellite channels%predistortion
针对卫星信道存在非线性失真严重影响系统性能的问题,提出了一种基于正交基神经网络的预失真补偿算法.该算法利用正交基神经网络的特点,能够一步确定权值,无需迭代,不存在局部极小值问题,计算速度快,误差小,补偿效果明显.
針對衛星信道存在非線性失真嚴重影響繫統性能的問題,提齣瞭一種基于正交基神經網絡的預失真補償算法.該算法利用正交基神經網絡的特點,能夠一步確定權值,無需迭代,不存在跼部極小值問題,計算速度快,誤差小,補償效果明顯.
침대위성신도존재비선성실진엄중영향계통성능적문제,제출료일충기우정교기신경망락적예실진보상산법.해산법이용정교기신경망락적특점,능구일보학정권치,무수질대,불존재국부겁소치문제,계산속도쾌,오차소,보상효과명현.
Since the satellite channel nonlinear distortion seriously affects the system performance,a pre?distortion compen?sation algorithm based on orthogonal neural network is proposed to solve this problem. The algorithm which uses the characteris?tics of the orthogonal basis neural network can determine the weight value directly in one step without iteration. It has the advan?tages of fast calculation and small error,and has no local minimum.