Tehnički vjesnik, Vol. 16 No. 4, 2009.
Izvorni znanstveni članak
Better bandwidth utilization of multiple link capacities with mutual traffic correlation
Srećko Krile
orcid.org/0000-0002-0748-1359
; University of Dubrovnik, Department of Electrical Engineering and Computing, Ćira Carića 4, 20000 Dubrovnik, Croatia
Drago Žagar
; J. J. Strossmayer University of Osijek, Faculty of Electrical Engineering, K. Trpimira 2b, 31000 Osijek, Croatia
Goran Martinović
orcid.org/0000-0002-7469-6018
; J. J. Strossmayer University of Osijek, Faculty of Electrical Engineering, K. Trpimira 2b, 31000 Osijek, Croatia
Sažetak
Network providers need very effective optimization tool for good utilization of scarce link capacities during exploitation. In the case of multiple link capacities with mutual traffic correlation such problem could be more demanding. The sizing problem is explained for satellite link, but it could be applicable for other transmission resources. Dimensioning of getaway link (G-link) can be realized only by new constructions (new channel equipment) on the Earth side. Mathematical model for optimal capacity sizing of N different link elements (capacity types) is explained, minimizing the total cost (expansion, conversion and maintenance). Instead of nonlinear convex optimization technique, that could be very exhausting, the network optimization method is applied. With such approach an efficient heuristic algorithm for three different capacity types is being developed. Through numerical test-examples this approach shows the significant complexity savings, but giving us very close to optimal result. However, in real circumstances some adding limitations on capacity state values have to be introduced. In comparison to other options it is obvious that heuristic option M_H (with only one negative value per capacity state) shows the best ratio between complexity reduction and result deterioration.
Ključne riječi
multiple link capacity expansion; capacity expansion/conversion problem; transmission capacities with mutual traffic correlation; satellite link resource management
Hrčak ID:
45717
URI
Datum izdavanja:
29.12.2009.
Posjeta: 1.376 *