The Bitcoin Lightning Network is still an experiment, and the new paper about a flaw by Jona Harris and Aviv Zohar supports this hypothesis.
In a research paper entitled “Flood & Loot: A Systemic Attack On The Lightning Network,” two researchers from the renowned Israel Hebrew University have exposed an inherent flaw in the Lightning Network payment technology.
As a reminder, the Lightning Network is a network built as a second layer, allowing to decongest Bitcoin’s on-chain transactions. The promise of this technology is to promote microtransactions by making transactions faster and cheaper. To do this, the participants in the network are connected to each other, step by step and in the form of a kind of a web of payments. This spider’s web consists of payment channels, which are in fact funds placed in escrow on-chain, in transactions co-signed by the participants (so-called multisig transactions).
To achieve this, the LN relies on HTLC, for Hash Time-locked Contracts. In such a system, payments are made via conditional promises, which are only validated on arrival by the two nodes exchanging payments.
Ultimately, therefore, payments are first validated off-chain after the cooperation of the various participants, and then validated on-chain when the payment channels used are closed and the balances redistributed to their respective owners.
By using the Lightning Network, users can therefore send payments through “intermediate nodes”.
The details of the attack
These are the same nodes that – according to Jona Harris and Aviv Zohar – could try to steal your Bitcoins, but the time frame is often too short to make such an operation possible. However, if a malicious actor were to “flood” the network with requests, it would lengthen the time required and make such an attack possible.
In the attack described by the researchers, “the attacker forces several victims at the same time to flood the blockchain with requests for their funds. It is then possible to take advantage of this congestion to steal the funds that were not claimed before the allotted time.”
To be plausible, the paper describes an action that should be distributed simultaneously among 85 channels in order to be profitable. It is described as “fairly easy” to find potential victims of this attack. Indeed, “all vulnerable nodes must show a willingness to open a channel with an attacker.”
“We have found that a majority of the nodes (95%) are willing to cooperate upon request, and therefore fall victim to the attack.”
This flaw, now theorized, had already been identified by several developers during the year 2019 – even if it had not been publicly announced.
According to the study, it will not be easy to counter this vulnerability.
“We believe it is inherent in the way the Lightning Network operates, and that this type of attack cannot be countered without significant modifications.”
The same goes for René Pickhard, who after proposing 6 different solutions, concludes:
“I’m really unhappy with the proposals that I myself have outlined. I hope you will find more intelligent solutions. I think that the solution is not obvious. But it was important for me to list some ideas, even if they contain obvious disadvantages.”
Even if it’s not widely reported on social networks, this news should raise awareness that Bitcoin’s second layer technology is still in its infancy. The promise of a decentralized system hangs by a thread, and only the genius of the developers will be able to maintain the romance. Let’s hope for the world that a timely solution can be found.