Title :
Resource allocation adaptive to traffic changes in communication networks
Author :
Onda, Kazuyuki ; Saito, Shigeru
Author_Institution :
ATR Adaptive Commun. Res. Labs., Kyoto, Japan
Abstract :
Next-generation communication networks will require the ability to adaptively maintain their performance against traffic changes and service diversification. To achieve this ability in connectionless packet networks, we propose a method of network resource allocation and topology arrangement adaptive to traffic demand. An optimization problem of such resources as the node service rate and link bandwidth is defined in the network layer, and solved with the Hamiltonian algorithm, which is effective in searching for the optimum solution with a large number of unknown variables. The obtained solution shows the contribution of every link to packet transportation and therefore enables the network topology to be arranged. Network resources are optimized again for the arranged topology. The proposed method is applied to a network that exhibits traffic concentrated to its backbone nodes. Successful resource allocation and topology arrangement are confirmed
Keywords :
adaptive systems; network topology; optimisation; packet switching; queueing theory; telecommunication networks; telecommunication traffic; Hamiltonian algorithm; adaptive resource allocation; backbone nodes; communication networks; connectionless packet networks; link bandwidth; network control; network design; network layer; network resource allocation; network topology; network topology arrangement; node queueing model; node service rate; optimization problem; optimum solution; packet transportation; performance; service diversification; traffic changes; Adaptive control; Adaptive systems; Bandwidth; Communication networks; Intelligent networks; Network topology; Resource management; Spine; Telecommunication network topology; Telecommunication traffic;
Conference_Titel :
Adaptive Systems for Signal Processing, Communications, and Control Symposium 2000. AS-SPCC. The IEEE 2000
Conference_Location :
Lake Louise, Alta.
Print_ISBN :
0-7803-5800-7
DOI :
10.1109/ASSPCC.2000.882512