The similarity is as I wrote in my prior post, that you essentially must use PoW to obtain the randomness of whom to select as the delegate to order the transactions in next block.
That's certainly the case for the model I suggested. But what about the leader election procedure described in
https://eprint.iacr.org/2016/889.pdf (chapter 4.2). Doesn't it ensure true randomness?
Neither of those even with millions of users (and especially if you expect them to be on mobile) will be sufficient to secure the chain against powerful attackers such as botnets with gaming rigs and hijacked Amazon EC2 or other VPS cloud accounts.
Coming back to the problem of unprofitable mining, I just had the following idea:
Instead of using PoW for making direct profit (as in Bitcoin) or decreasing transaction fees (as in the model I described), we could require each stakeholder to perform PoW in order to preserve his stake. Without producing the required number of blocks in a certain timeframe, the stakeholder would be levied a negative interest rate from his stake and thus lose a percentage of his coins (which would be burned). The amount of work would depend on the stake as well, so that small stakeholders with only a few coins could easily do the computation on their computers or even phones while large investors would need access to mining rigs or server farms. (Splitting the stake between several accounts wouldn't help to reduce the resources needed to prevent the sum from decreasing over time.). On the other hand, the amount of stake would serve as an upper limit on block creation for every stakeholder: Once a stakeholder has produced the required number of blocks, his blocks would be rejected by the other nodes and prevented from becoming part of the block chain.
Letting aside the problem of granularity (i.e. block size/block creation rate which also impairs decentralization as it throttles the number of nodes who can participate in block creation), such a scheme would offer some interesting properties.
- In order to attack the system (and make a double-spend), you would need to beat the network in terms of hashrate and stake at the same time.
- Only big investors have an incentive to become high-scale miners with a lot of hashpower. Such investors have little interest to attack the system.
Any thoughts?
By the way, thanks imnotback for uploading the two chapters (which I've read with interest).