| 摘要 |
The invention relates to the scheduling of data transfers in a multi-hop packet network. The nodes of the network are adapted to schedule their transmissions according to a common time sequence, recurring in time domain and comprising a control portion for transmission of at least one control packet and a data portion for transmission of data packets. In order to accomplish a simple and controlled way for minimizing delay and delay variation, the network is classified into several levels with respect to a certain node, each level comprising the nodes located at the same distance from said certain node, measured in number of hops along the shortest path in the network. The data portion is further divided into successive reservation periods, each being allocated to transmissions of delay sensitive traffic through the hops between two predetermined neighboring levels so that a data packet can be transferred across the network within a single time sequence. |
