ControlPay: An Adaptive Payment Controller for Blockchain Economies

Recent advancements in blockchain technology have led to the development of various decentralized service platforms for various tasks, like machine learning and wireless networks for example. Central to the operation of these platforms is a token-based economy, rewarding service providers with cryptocurrency tokens for their contributions to the setup, verification, and maintenance of a platform. However, these platforms often rely on predetermined token supply strategies which render a platform's operation susceptible to market fluctuations. A more flexible approach, one allowing for dynamic response to changes in system demand and market conditions, is essential to mitigate such vulnerabilities. To address these challenges, we introduce a control-theoretic approach to stabilizing a decentralized service platform's token economy. Specifically, we first model these blockchain economies as dynamical systems where token circulation, pricing, and consumer demand evolve based on payments to service providers and service costs. Then, we utilize our model to introduce ControlPay: a novel payment controller based on model predictive control (MPC) designed to enhance the performance of decentralized networks while simultaneously ensuring token price stability. Additionally, we also examine the impact of strategic behavior in the market through a Stackelberg game to further enhance the robustness of our payment controller. Finally, we evaluate our methodology on real and synthetic data. Our findings show that ControlPay significantly outperforms conventional algorithmic stablecoin approaches, yielding improvements of up to 2.4× in simulations based on actual demand data from existing blockchain-driven decentralized wireless networks.