8 Conclusion

UNIFY is a new scheme for maintaining consistency between aggregated and disaggregated levels. It does away with explicit switching between aggregated and disaggregated states, and instead, maintains all state information for all levels. Units can comply with interactions at different levels. Perceivers are responsible for demanding information they need. We solve the temporal consistency problem by requiring atomic entity interactions. We resolve chain disaggregations because disaggregation is non-existent. We alleviate network flooding by causing fewer entities to be created and making message aggregation simpler. Also, we provide the mechanisms for aggregations of dissimilar entities and dynamic aggregations. UNIFY requires more memory and CPU cycles, but reduces network traffic. It is not clear how computation of complex Lanchester equations will match speeds at which virtual entities perform interactions, and we recommend research in this area. Existing simulations may not be able to support UNIFY, but this may be an advantageous way to design new aggregate-level simulations. We intend implementing UNIFY to give a proof-of-concept.