Lightning Network

Diane Reynolds: `Lightning Network Will Maintain Decentralization´

Diane Reynolds just penned an article in defense of the Lightning Network (LN) as an off-chain, decentralized payment solution. This article appeared in response to the piece written by Jonald Fyookball and a Reddit post by jstolfi. Both pieces claimed Lightning Network cannot remain decentralized over the long term for mathematical reasons. 

Also read: Delaware Passes Bill Recognizing Stock Trading via Blockchain Technology

Fyookball’s article specifically used various examples to demonstrate decentralization is unlikely to manifest over the long term. Diane responded more directly to a challenge issued by jstolfi. The user’s challenge said:Diane Reynolds Says Lightning Network Will Maintain Decentralization

“There is a very simple way to shut criticisms like Jonald’s and mine. Just provide a hypothetical scenario for 10 million users with topology and numbers — how many customers, merchants, and hubs, how many channels and payments (per day or per month) per user for each pair of those user classes, and how much bitcoin each user commits to his channels, etc. Then anyone who doubts the viability of the LN can simulate it with those data, and conclude for himself. Any takers for this challenge?”

Reynolds’ Critical Analysis of Jstolfi’s challenge Using a Simulation

Reynolds responded with answers to the challenge. In the first part of Reynold’s analysis she mentions she ran some ocaml code to simulate a Lightning Network environment topology with 10 millions users. She used her model to create a situation that is reminiscent of proper lightning network transaction details. For instance, the payments were of all sizes, big, midsized and micro. She also made her code search for the least expensive channel for payments to be accepted.

One of the primary arguments for decentralization via Lightning Network is the notion that plenty of redundancy on the network de facto creates decentralization. Reynolds argues that so long as nodes can hop between and connect with different nodes the system cannot become centralized. She said,

A first possibility for edges is to use the binary representation of numbers from 0 to 9,999,999 (requiring 24 bits) and include an edge whenever the two bit sequences differ by exactly 1 bit (a Hamming graph). All nodes would be reachable with (many) paths of at most 24 hops, and 12 hops on average.

Diane Reynolds Says Lightning Network Will Maintain Decentralization
A Connected Network with a 100 nodes and many paths

Analysis Continued

In her analysis, Reynolds describes the state of the network after simulating 400,000 transactions. At this point, Reynolds navigated through various transactions on the network in order to answer the original challenge. She determined that there were very few transactions failures across the network. However, she also said that there could be a problem if she continued doing simulations on her program.Diane Reynolds Says Lightning Network Will Maintain Decentralization

“Among the 7 million channels used routing the almost 400,000 successful payments 294508 (4%) have 90% or more of the value on one side of the channel. This does not appear to be a problem yet, since this is still only 0.4% of the total available 70 million channels, but it might become a problem with longer runs of the simulator.”

Regardless according to Reynolds, it was a non-problem. There did not seem to be a significant amount of unbalanced transactions. There could perhaps be problems on the network simulation, but she could not pinpoint any issues.

Conclusion

In her final thoughts, Reynolds suggested that 10 millions users can be sustained on the network in a decentralized fashion. However, she mentions that this is not how Lightning Network is supposed to function, but that it is technically possible. She does mention that if anyone has stronger computing hardware, they can run the simulation with even more users and more channels.

Do you think Segwit will end up being centralized or decentralized if implemented? Let us know in the comments below.


Images via Shutterstock and Medium.com


Bitcoin is cool, and you know everyone wants in – even the ones who say they don’t. Show the world how cutting-edge you are with a bitcoin T-shirt, hoodie, bag, key-ring, even a Trezor hardware wallet. Shipping all over the world, quality merchandise and, of course, a payment system that makes people say “wow!”

  • Khurram Javed

    Excellent to see discussions continue.

  • Ian O’Neill

    I am not an expert on Segwit or Lightening Network, but intuition tells me that for those shepherding both endeavours, the temptation is strongly towards (ever greater) centralization. Like an alcoholic affirming the next drink will be the last, we are more than capable of convincing ourselves we need this last centralizing move to get through the day. I say ‘we’ because users of these networks would be similarly tempted if they had the responsibility. In the global economy competition for money is a race of all against all – against debt – each of us seeks not to be the one without a place to shelter the next time the wild game of debt calls time. There will never be enough fiat currency to go around even in the face of competition from Bitcoin and Lightening Network! The last time I checked, fiat is still practically omnipotent. It should surprise no one that the issue of centralization proves to be more human than technical, and is of course bigger than both Segwit and Lightening Network.

  • She needs star, multi-star aggregation. And, you can find that in an ATM UTOPIA black-plane running four planes of traffic aggregation. Hardware that closely resembles the interconnected Internet is not so easy to code simulate. You will need multi-node mesh network geographically dispersed with functioning exchanges that are cross-connected. Also, real time delays and congestion impairment as well as real time live traffic models base on the geographic time frames of the real world, not just the networks of the transaction involved, but the whole character of the network to be crossed. Going off network is not as easy as it sounds. For example, synchronizing off load in TCP stacks has taken decades and involved many CAM changed and improvements most in standardization bodies, not to mention acceptance between hardware manufactures and Software Developers. A mathematical model is not all that is required for wide acceptance. Good Luck!