Going beyond structure - Can Dynamic Multi-Layer Algebra Support Multi-Level Simulation?

Multi-level simulation (MLS) combines simulation models that differ in abstraction, resolution, timing, or computational cost. This makes MLS attractive for complex systems, but simulation artefacts must be selected, coupled, and exchanged without losing semantic compatibility. At the same time, “level” denotes different things in MLS and in multi-level modeling (MLM), and is already ambiguous within MLS. We therefore first separate the notions of level involved and show that the orderings used in MLS induce no instantiation relation, whereas the instantiation chains that do exist in MLS run orthogonally to them, inside a single simulation stage. Against this background we examine Dynamic Multi-Layer Algebra (DMLA) as a modeling basis for MLS compatibility: not as an encoding of MLS levels, but as a level-blind, refinement-based formalism for representing simulation artefacts, simulator capabilities, compatibility constraints, and validation operations. An electronic control unit (ECU) example and a DMLA model excerpt illustrate how validation operations expose missing inputs, semantic mismatches, and unjustified simulator switches. We condense recurring challenges into six modeling needs, identify which of them any MLM approach with a constraint language can address and which benefit specifically from DMLA, and state where further work is required.