Delay tables are utilized in computer networks to assist in routing-protocol decision-making. A solution for the problem of determining the structure of these delay tables for the gateway nodes of hierarchical networks such as the DARPA Internetwork, i. e. , the gateway dynamic congestion table (GDCT), is derived and its performance evaluated. The dynamic congestion Table algorithm serves as a theoretical basis for the GDCT. This conceptually compressible and expandable three-dimensional data structure enables storage and time complexity reduction in the transmission of computer node and channel path delay table matrices. The compressed path delay for subnet paths is obtained from the respective input and output port dimensioned subnetwork DCTs for the gateway table structures.