Hierarchical Coordination Protocol

Origin. Developed in parallel by Mesarovic (Serbian-American, Case Western Reserve) and Soviet researchers including Burkov at the Institute of Control Sciences; applied extensively in Soviet economic planning theory by Fedorenko and CEMI (Central Economic-Mathematical Institute).

Mechanism. A complex system is decomposed into subsystems with local autonomy, coordinated by a higher level that specifies goals or constraints but not methods. The coordinator's intervention is minimal: it adjusts parameters (prices, targets, resource allocations) only when subsystem plans conflict. Subsystems optimize locally against those parameters. The protocol converges to a globally feasible solution if the coordinator updates in the direction of the gradient of the coordination error.

Procedure. Decompose the problem into subsystems with minimal interaction. Assign a coordinator. Subsystems propose plans. The coordinator checks feasibility: do the plans satisfy global constraints? If not, identify the violated constraint and adjust the shadow price on that constraint upward. Subsystems revise plans against the new prices. Iterate until feasible. The coordinator never specifies subsystem internals; it specifies only the interface.

Applies to. Budget allocation, project portfolio management, scheduling with shared resources, and any problem with decentralized execution and centralized coordination.

Limitations. Convergence requires convexity or at least unimodality; nonconvex problems can trap the iteration at local optima. The protocol also depends on subsystems optimizing honestly against the coordinator's prices; if subsystems game the pricing mechanism (strategic misrepresentation of costs), the iteration diverges. This was the chronic Limitations of Soviet planning: enterprises inflated cost estimates to secure slack, and no pricing mechanism could correct for it because the coordinator had no ground truth.