Selection of User's Connection in Last-Mile Problem.

This correspondence paper focuses on multicriterion combinatorial models in the “last-mile” problem for selection of user's connections. The following models are considered: 1) multicriterion ranking; 2) knapsack-like problems (e.g., multicriterion multiple-choice problem); 3) clustering; and 4) hierarchical morphological synthesis based on morphological clique problem. The last-mile problem is examined from the following viewpoints: 1) user(s); 2) provider(s) (i.e., communication service provider); and 3) users and providers. The following problems are described: 1) selection of the best connection for user; 2) selection of the best connections from several providers; 3) multistage selection of the best connections for user; 4) clustering of users; and 5) design of hierarchical connection system from provider. Numerical examples illustrate some applications. [ABSTRACT FROM AUTHOR]