Payment Channel Networks with Resource-constrained Devices

Payment-channel networks (PCN) represent the leading solution to scale blockchain-based payments to the performance levels of centralized payment systems. However, current PCNs require nodes to stay permanently online and have enough resources to execute payment security mechanisms. Such assumptions are difficult to guarantee in battery-powered devices with intermittent connectivity patterns, such as mobile phones, smart objects, and sensors. In this thesis, we address the case of PCNs with resource-constrained devices on several fronts. First, we formalize a hybrid PCN model that considers light nodes and propose a mechanism to protect payment channels with resource-constrained devices. Our experiments show that the proposed mechanism is efficient for devices with high and medium availability in mobile broadband connections. Next, we propose PCNsim, a simulator that replicates the main functionalities of a PCN in the OMNeT++ framework. PCNsim allows researchers to experiment with payments under custom networking conditions representing resource-constrained devices’ connections. PCNsim’s demonstrations show that it correctly reproduces the behavior of a PCN over unreliable communication channels. Finally, we address the problem of routing payments from resource-constrained devices. We present a payment scheme that anticipates payment confirmations for time-sensitive applications and two routing algorithms that route payments considering application-specific constraints. The results show that our routing algorithms are efficient both for single-path and multi-path payments and reach their top performance when the problem’s constraints are tight.