Title :
Optimization of communications traffic systems using learning control
Author_Institution :
NTT Telecommun. Networks Lab., Tokyo, Japan
Abstract :
A learning control method applicable to optimization control of communications traffic systems is proposed. The method consists of a system identification algorithm using the method of potential functions and an extremum search algorithm that selects the appropriate control input at each control step. The method makes it possible to achieve maximization by setting the initial approximation of the system to be greater than that of the real system and by selecting the control input to be the temporal maximum at each control step. An application example of a throughput maximization problem at a switching node in a circuit-switched communications network is shown
Keywords :
circuit switching; identification; learning systems; optimisation; search problems; telecommunication systems; telecommunication traffic; telecommunications control; circuit-switched communications network; communications traffic systems; extremum search algorithm; learning control; optimization; potential functions; system identification; throughput maximization; Communication system control; Communication system traffic control; Control systems; Laboratories; Noise measurement; Nonlinear control systems; Optimization methods; Telecommunication control; Throughput; Traffic control;
Conference_Titel :
Decision and Control, 1990., Proceedings of the 29th IEEE Conference on
Conference_Location :
Honolulu, HI
DOI :
10.1109/CDC.1990.203832
