Onion Message Jamming in the Lightning Network

The discussion revolves around the limitations and potential strategies in dealing with onion message jamming and quick channel jamming within a network. The proposed scheme focuses on deploying unconditional fees alongside other mitigations to limit the resources attackers can consume, while still preserving the ability for peers with historically good behavior to make payments. This approach is contrasted against dynamic fee adjustments, which would ideally start low and increase as traffic builds up to prevent honest parties from bearing high costs but would require rapid responsiveness to be effective.

There is skepticism about whether rapid reactions are necessary or beneficial, considering that if an attacker is determined to spend resources to jam the system, those resources could instead be potentially redirected or absorbed without immediate harm to the system's operation. Furthermore, the effectiveness of using onion messages (OMs) as input into pathfinding remains questionable. There's a consensus that OMs may not provide significant pathfinding advantages, suggesting that their failure does not necessarily imply the inability of a path to carry a payment.

The email highlights the importance of simulation to better understand and refine the proposed solutions under adversarial conditions. It references specific efforts and tools such as adapting changes from a GitHub pull request (PR 1280) and utilizing a simulator (jam-ln simulator) which has been used to study channel jamming scenarios. These simulations have not yet incorporated unconditional fees directly; they have been manually calculated post-simulation, hinting at an area ripe for further contribution and refinement.