A common problem in a packet radio network (PRN) environment is to construct a multicasting network from a single source to a set of remote destinations which minimizes the number of transmissions. This problem is known to be NP-complete, thus computing an optimal solution may be infeasible for sizable networks. This paper provides two alternative solutions to this problem. The first is a heuristic algorithm which iteratively builds a spanning tree from the destinations to the source. A second solution, included for comparative purposes, is based on the Hopfield neural network whose dynamics are governed by a motion equation and a set of constraints. Both solutions are tested on a variety of instances against an optimal algorithm. Results show the approaches form good solutions (the number of transmissions is within about 3% of the optimum) and run in a fraction of the time required to form the optimal solution.
|Original language||English (US)|
|Number of pages||9|
|Journal||International Journal of Smart Engineering System Design|
|State||Published - Jan 1 2002|
All Science Journal Classification (ASJC) codes