Diagnosability of multigraph composition networks

A common approach to scale up a network is by connecting several small-scale networks to form a larger network. A multigraph composition network is a typical network structure obtained by adding an edge set between several connected graphs in the same order. This paper investigates the diagnosability on two categories of multigraph composition networks, that is, multigraph alternating composition networks and multigraph 2-matching composition networks, under the PMC and MM* models. As corollaries, the diagnosability of several known networks, such as alternating group graphs, k-ary n-cubes, and round matching composition networks, can be derived directly.