As the demand for delay constrained network slicing augments, the possibility rate of network slicing resource assignment algorithms descend. To address the issue, this article constructs a resource model with network slicing delay constraints and network features related to delay constraints. The network characteristics of network slicing include slice propagation delay constraints, slice processing delay constraints, and slice node delay constraints. The network characteristics of the physical include the degree of nodes, the average link transmission delay of nodes, the peripheral link resource value of nodes, and the peripheral node resource value of nodes. A slicing delay constrained resource assignment algorithm using TOPSIS was designed. The main steps of this algorithm include sorting the resource assignment order of network slices using their features, assigning physical node resources that meet delay constraints to slice nodes using TOPSIS, and assigning physical path resources that meet delay constraints to slice links using K-shortest path algorithm. The conclusion was drawn through experimental comparison. This algorithm improves the success rate of resource assignment and enhances the utilization of the physical network.