The entity–control–boundary approach finds its origin in Ivar Jacobson's use-case–driven object-oriented software engineering (OOSE) method published in 1992.
[1][2] It was originally called entity–interface–control (EIC) but very quickly the term "boundary" replaced "interface" in order to avoid the potential confusion with object-oriented programming language terminology.
[4] It uses graphical stereotypes introduced in Jacobson's early work: The following robustness constraints apply: In principle entities should not know about boundaries and controls.
However the role of the ECB-control is very different from MVC-controller, since it encapsulates also use-case business logic whereas the MVC controller processes user input which would be of the responsibility of the boundary in ECB.
The ECB control increases separation of concerns in the architecture by encapsulating business logic that is not directly related to an entity.