Publication Details
Abstract
This paper considers a heterogeneous data transmission network (DTN) comprising several mobile information sources and a single central receiver, jointly using radio and wired communication channels. Each channel is characterized by its throughput, service intensity, and probability of successful packet delivery. For every newly arrived packet, the delivery time is estimated along all admissible routes, and the route with the smallest expected delay is selected, thereby providing adaptive traffic distribution among alternative paths. The simulation-optimization model of the network is built on a discrete-event approach. Its effectiveness was assessed through five experiment series examining the effects of source intensity, packet size, alternative-route usage, delivery probability, and channel capacity on the network's time characteristics. The results show that as the delivery probability increases from 0.1 to 1.0, the maximum delay decreases by approximately 90.6%, while beyond a certain packet-size threshold the transmission time begins to dominate over the waiting time, accompanied by a marked decline in the frequency of alternative-route usage. These results can serve as a basis for practical recommendations on justifying channel parameters and improving adaptive routing mechanisms when designing heterogeneous DTNs.