TY - GEN
T1 - The application of neural network algorithm in ATM buffer management
AU - Yin, XH
AU - Sun, BL
AU - Zhou, HB
PY - 2003/9/26
Y1 - 2003/9/26
N2 - In ATM switching, cell congestion is caused by the head-of-line (HOL) blocking in cell queuing. In order to alleviate the congestion in cell queuing buffer, HOL should be cleared. In this paper, Hopfield Neural Network (HNN) technology is applied to manage the cells buffers in ATM Switching Fabric (ASF) design. This intelligent algorithm can help to monitor the real-time traffic distribution and efficiently improve the performance of cell switching. Furthermore, with the help of energy function in HNN, the stability of HNN is discussed.
AB - In ATM switching, cell congestion is caused by the head-of-line (HOL) blocking in cell queuing. In order to alleviate the congestion in cell queuing buffer, HOL should be cleared. In this paper, Hopfield Neural Network (HNN) technology is applied to manage the cells buffers in ATM Switching Fabric (ASF) design. This intelligent algorithm can help to monitor the real-time traffic distribution and efficiently improve the performance of cell switching. Furthermore, with the help of energy function in HNN, the stability of HNN is discussed.
KW - Hopfield Neural Network
KW - ATM
KW - buffer management
KW - head-of-line(HOL) blocking
UR - https://openir.whu.edu.cn/AchievementDetail/WOS_e9725562-ace3-432e-b817-a6a049ca9b03
M3 - Conference contribution
SN - 0-9707890-2-5
T3 - The application of neural network algorithm in ATM buffer management
SP - 526
EP - 529
BT - PROCEEDINGS OF THE 7TH JOINT CONFERENCE ON INFORMATION SCIENCES
PB - Assoc Intelligent Machinery
ER -