Bitcoin Forum

Bitcoin => Bitcoin Discussion => Topic started by: bluemeanie1 on June 17, 2014, 12:45:34 AM



Title: Selfish Mining Reconsidered
Post by: bluemeanie1 on June 17, 2014, 12:45:34 AM
Finally got around to reading this Selfish Mining Paper.  http://arxiv.org/pdf/1311.0243v5.pdf

Quote
We now describe our strategy, called Self sh-Mine. As we show in Section Sel fish-Mine allows a pool of sufficient size to obtain a revenue larger than its ratio of mining power. For simplicity, and without loss of generality, we assume that miners are divided into two groups, a colluding minority pool that follows the self sh mining strategy, and a majority that follows the honest mining strategy (others). It is immaterial whether the hon.est miners operate as a single group, as a collection of groups, or individually.

Here the author makes some very poor assumptions.  It is possible to easily thwart this attack by making a double-agent miner who claims to be part of this subversive pool, but isn't and publishes the selfish chain information to the outside.  This double agent miner need not even have any real mining power, they only need to CLAIM such.  Thus the cost of this defense is virtually negligible.  This would thwart this attack vector entirely.

-bm


Title: Re: Selfish Mining Reconsidered
Post by: bluemeanie1 on June 17, 2014, 01:07:26 AM
the essential problem here, and completely unexplored by the author who is convinced he is in the possession of a economic weapon of mass destruction, is that of secrecy.  How can one maintain the secrecy of blocks in a p2p network?  It's not as easy as the author suggests and he ignores the GREATER incentive(and minimal cost) of maintaining the integrity of the network as a whole.  Basically anyone who is invested in the long term viability need only create a double-agent that participates in the pool.  Then through a back-channel the node publishes the secret blocks.  The selfish pool has no way of knowing who is the snitch.

Sorry but Selfish Mining is not a fatal flaw in Bitcoin.

-bm


Title: Re: Selfish Mining Reconsidered
Post by: TwinWinNerD on June 17, 2014, 05:09:21 PM
the essential problem here, and completely unexplored by the author who is convinced he is in the possession of a economic weapon of mass destruction, is that of secrecy.  How can one maintain the secrecy of blocks in a p2p network?  It's not as easy as the author suggests and he ignores the GREATER incentive(and minimal cost) of maintaining the integrity of the network as a whole.  Basically anyone who is invested in the long term viability need only create a double-agent that participates in the pool.  Then through a back-channel the node publishes the secret blocks.  The selfish pool has no way of knowing who is the snitch.

Sorry but Selfish Mining is not a fatal flaw in Bitcoin.

-bm


pay back the 1 Million NXT you stole from the NXT community.
https://nxtforum.org/nxtventures/nxtautodac-all-of-bluemeanie%27s-automated-profit-making-blockchain-companies/60/


Title: Re: Selfish Mining Reconsidered
Post by: ShakyhandsBTCer on June 17, 2014, 05:46:41 PM
Finally got around to reading this Selfish Mining Paper.  http://arxiv.org/pdf/1311.0243v5.pdf

Quote
We now describe our strategy, called Self sh-Mine. As we show in Section Sel fish-Mine allows a pool of sufficient size to obtain a revenue larger than its ratio of mining power. For simplicity, and without loss of generality, we assume that miners are divided into two groups, a colluding minority pool that follows the self sh mining strategy, and a majority that follows the honest mining strategy (others). It is immaterial whether the hon.est miners operate as a single group, as a collection of groups, or individually.

Here the author makes some very poor assumptions.  It is possible to easily thwart this attack by making a double-agent miner who claims to be part of this subversive pool, but isn't and publishes the selfish chain information to the outside.  This double agent miner need not even have any real mining power, they only need to CLAIM such.  Thus the cost of this defense is virtually negligible.  This would thwart this attack vector entirely.

-bm


In order for the "double" agent miner solution to work the "double" agent needs to be the one who finds the block


Title: Re: Selfish Mining Reconsidered
Post by: jonald_fyookball on June 17, 2014, 08:25:24 PM
just a very quick review of the paper.

1. they are trying to solve a non-existent problem.
By the author's own research (see figure 2),
the revenue would be nearly the same with
honest vs selfish mining.

2. I dont think their proposal would work
as far as propogating chains of the same
length.  At least, it would be risky.
There always has to be a best chain.
One does not mess with the keystone to
distributed consensus.
--

Selfish mining is a real problem if attacker
has 51% but we already knew that.