Lightning network traffic analysis raises costs and privacy issues

A new research report provides the most informative series of snapshots of last year's simulated lightning network traffic.

The paper entitled `` A Cryptoeconomic Traffic Analysis of Bitcoin's Lightning Network '' was written by three Hungarian researchers: Institute for Computer Science and Control (Institute for Computer Science and Control by Ferenc Beres, Eotvos Lorand University, Istvan A. Seres, and Szeni East Andras A. Benczur of Szechenyi Istvan University.

In this study, a lightning network traffic simulator was used to simulate the network transaction process, and it was concluded that the current network transaction rate is not economically long-term feasible. In addition to the immature fee market, this article also explores the trend of Bitcoin senders using the Lightning Network to take advantage of more direct and fewer private payment paths. This paper proposes solutions to these problems, as well as how the Lightning Network should ultimately treat Bitcoin.

Lightning Network Problem 1: The Economic Feasibility of the Toll Market

Compared to institutional giants such as Visa (which processes approximately 40,000 transactions per second), the Bitcoin blockchain processes far fewer transactions (7 to 9) per second. Although Bitcoin is certainly not trying to become Visa, this illustrates the purpose of the Lightning Network: to increase the scale of Bitcoin. Therefore, the core idea of ​​the Lightning Network is to allow people to issue bitcoin payments on the blockchain so that they can perform more seamless operations. This reduces the processing backlog by creating faster and cheaper payments, while also providing an interesting opportunity for better privacy.

Lightning networks are run by nodes based on Lightning Technology (BOLT), which is the basis of Lightning Network protocols. Similar to running a Bitcoin node on a chain, Lightning nodes route payments between senders and receivers, which in turn receive transaction fees from their work.

However, because the Lightning Network is newer and more private than the Bitcoin blockchain itself, the authors of the research paper stated in a briefing that "as far as we know, there is currently no paper on quantifying the financial incentives of LN router nodes." Analysis uses traffic data extracted from 40 consecutive snapshots of Lightning Network from February to March 2019, and subsequent estimates of these data are simulated.

The paper found that "currently, LN provides almost no financial incentives for payment routing. Lower routing costs do not adequately compensate routing nodes, which actually connect the network together." In their report, the authors estimate that in different The main router entity of the level of traffic and payment value, the daily routing revenue is about 100,000 sats (about $ 7).

For clarity, the developers of Lightning Network set default transaction fees for node operators. They can agree to do so, or they can implement independent actions for their own Lightning software. However, node operators can manually adjust their transaction rates to any degree they want. The paper described this tendency to network-wide with low transaction costs as "economically irrational," but in a conversation with Bitcoin Magazine, Budapest's PhD in mathematics, Serez, quickly acknowledged the behavior. Perhaps more similar to an early form of altruism that emerged from the fee market on the Bitcoin chain.

Assuming that the node operator wants to obtain a reasonable return on investment from transaction costs, this article asserts that the financial incentives of Lightning Network are not enough. In order to make the node operator's payment routing economically feasible, this article proposes either increasing traffic or increasing transaction costs.

"In this sense, many larger nodes can increase transaction costs because they are in a monopoly position," the author wrote.

This means that because many larger nodes currently account for the percentage of network routing exchanges, they are indispensable. "Neighboring nodes don't have many options to route payments," Seres said.

Lightning Network Issue 2: Routing Privacy

Thesis also found that, despite the random onion routing, "the sender and receiver of payment can collect strong statistical evidence because a large portion of payments involve only one routing agent."

As a potential solution to this privacy flaw, the authors suggest using "deliberately suboptimal, longer routing paths to restore privacy, which only slightly increases the cost of the average transaction." "Furthermore, on the sender and receiver sides, Establishing a direct payment channel between them can create the best level of privacy.

According to the paper's suggestion, if the default transaction costs increase, taking longer and fewer direct payment methods to protect privacy will cost more. Similar to Concerned about the default transaction fees set by node operators, and in the spirit of the earliest development of Bitcoin, the findings of this article currently place privacy responsibility on the user's choice of payment path.

Reasonably reduce transaction fees

Many developers and nodes in the Lightning Network community Operators are unlikely to agree with the paper's conclusion that low transaction fees are economically unreasonable or even problematic.

For example, many Lightning Network node operators are Bitcoin businesses, such as Bitrefill, which allows customers to pass the Lightning Network Pay using Bitcoin to buy various products and services. Companies like Bitrefill Almost at low transaction costs to use lightning network services to their customers.

Bitrefill company's CCO • John Carvalho (John Carvalho) said in an interview with "bitcoin" (Bitcoin) magazine interview: "If the cost of the route itself Participants have accumulative costs and consider Lightning Routing to be economically unreasonable. This idea is completely short-sighted. "At least two main factors must be considered. First, Lightning provides some features that are not on the Bitcoin transaction chain, such as high frequency and real time. These are features that people are willing to pay to access. Just like Bitrefill, economic participants have an incentive to expand their node activities as a business from the outside. The need for business-friendly bitcoin transactions is real.

Another bitcoin company that uses Lightning Network with low transaction costs is Sparkswap, an application that allows users to buy Bitcoin directly in USD and put it into the Lightning Network wallet.

Sparkswap founder Trey Griffith said: "Looking at our routing fees, you may think that our way of doing business is uneconomical, but considering our overall business, the situation is different."

Griffith said that because the Bitcoin business dominates the routing of the Lightning Network, it is difficult to draw meaningful conclusions from the data in the paper, such as why certain openings form triangular routes.

Griffith agreed that transaction fees would increase over time, and Carvalho acknowledged that "over time they may approach levels on the chain ... predictable rate rankings: on the Bitcoin chain > Lightning Network> Junk Tokens. "Do

you increase Lightning Network transaction costs?

The ultimate hazard of low Lightning Network transaction costs" may be that nodes will exit the network and lose their ability to pay channels. "

In October 2019 Rusty Russell, developer of Blockstream, points to another pair of issues with setting low default transaction fees.

In a message to the Lightning Dev mailing list, Russell noticed that two-thirds of the node operators "sit on the default transaction fees (1000 [millisatoshi] + 1 [ppm (ppm is a millionth)]]) ". This means that most node operators on the Lightning Network will not choose to deviate from the default rates set by developers.

Russell pointed out that accepting the default cost means lower reliability, and routing will be slower when trying to find a responding node. He also pointed out that “there is no clear market signal in terms of fees”, which means that the Lightning Network fee market does not look like a real market.

Russell's colleague, Blockstream researcher, and general Bitcoin and Lightning Network fanatic Christian Decker recently elaborated on Russell's views in more detail and responded to the findings of the paper. By increasing the default cost, we will automatically deviate from the routing of those nodes that have just been set up and left unattended. "
Decker also pointed out that this paper seems to ignore that the c-lightning implementation has chosen a random route for the sender. Decker said: "Because of the random nature of routing, increasing the default rate will not cause those nodes to be excluded from forwarding payments, but we can still turn routing to more actively managed nodes."

Looking ahead to 2020,

Although this article interviewed Most sources disagree with the paper's conclusions, but their concerns do reveal the Lightning Network in 2019, and the likely main focus in 2020.

"Since we started public testing more than a year ago, the Lightning Community has built a lot of infrastructure to guide the network and try what works and what doesn't," Decker said. "This lays the foundation for our development in the next few years, and we expect that many future developments will be more user-oriented." This includes a simpler user experience and further improved reliability, which can drive user adoption. We agree with the authors that privacy may be a major topic of the Lightning Network in 2020. "

Carvalho low transaction costs do not think flash grid have any questions, and plans to major improvements in the coming year.

He said:" Ultimately, P2P networks and voluntary nature means against the rationality of the argument must refute the fact : The network still exists, is still evolving, and is being used by businesses today. "In my opinion, we just touched the surface, and Lightning will show more use cases and participation incentives in 2020." "

In addition, Carvalho pointed out that lightning never committed network address privacy concerns - this is their engineers and more focus on user privacy culture of side effects and, in the context of the data chain, have the ability to see some money move in the lightning. Does not mean that you have useful external pollution applied to it.