Game Theory Model Solves Risk-Aware Infrastructure Pricing for Edge Computing

A new research paper from Amal Sakr, Andrea Araldo, Tamer Başar, and Tijani Chahed tackles a critical challenge in shared infrastructure deployment, particularly for Mobile Edge Computing (MEC). The paper, titled "Shared Infrastructure Investment and Pricing: Stackelberg Equilibria in Risk-Aware Take-or-Pay Contracts," models a scenario where an Infrastructure Provider (InP) invests upfront in capacity, and multiple firms commit to future usage via take-or-pay contracts under uncertain revenues. The firms exhibit heterogeneous risk aversion, modeled through Conditional Value-at-Risk (CVaR), and face operational costs and resource congestion. The authors introduce a Stackelberg game where the InP acts as leader, setting capacity and prices, while firms are followers that commit to usage levels. They prove the existence of a unique equilibrium among followers (a generalized Nash equilibrium) and develop a polynomial-time algorithm that boundedly approximates the Stackelberg equilibrium.

The key contribution is the integration of risk aversion into the classic infrastructure investment problem, which has real-world implications for how telcos and cloud providers plan edge computing rollouts. Simulations in a realistic MEC scenario reveal that as firms become more risk-averse, the InP must lower capacity and prices, reducing its own profit but increasing the lower bound on firms' Probability of Profit (PoP). This suggests a trade-off: risk-averse firms are more likely to survive but require less aggressive pricing from the infrastructure provider. The work provides a theoretical foundation for designing contracts that balance investment risk and usage commitment, with potential applications beyond MEC to shared fiber networks, data centers, and even energy grid infrastructure.