I have just released an in-depth article where I develop, step-by-step, a new secure blockchain model without Proof-of-Work:
https://medium.com/@cv.alkan/decentralized-objective-consensus-without-proof-of-work-a983a0489f0a#.erem149vuHere are some teasers from the article:
Decentralization and objective consensus are two of the main challenges faced by today’s cryptocurrencies. While the lack of objectivity is commonly attributed to Proof-of-Stake blockchains, the actual degree of decentralization is questionable in any existing consensus algorithm whether based on financial stake or computational resources. We will analyse both difficulties and present a novel blockchain model that offers objective consensus along with a strong tendency of decentralization.
This leaves us with the question why stake holders would build blocks if they get their interests anyway. Also, a powerful attacker might try to obtain the majority of the accounts with the aim of launching a double-spending attack. Hence we must curtail the trade with accounts and make it hard to gain majority control over the coins. Fortunately, both problems can be solved if we turn minter accounts into a non transferable, limited resource.
Even if we assume that some minters will keep their child accounts for themselves, they won’t be able to increase their total minting power without buying new accounts since the minting pool grows as well. All they can achieve is retaining their relative impact on the consensus. Eventually, with more and more stakeholders selling their child accounts, the minting power will become increasingly decentralized.
To summarize, our blockchain design is based on a dual token scheme with interacting PoS relationships. Minter accounts act as a second token, which you need to build the blockchain and also to “mint” child accounts, and there’s an opposite PoS relationship in that you need coins (the first token) to “fill” the minter account so that they can produce more coins (interests).
With the presented mechanism, new nodes don’t need to get the canocial blockchain from a trusted source since they can recognize it as such all by themselves. Our blockchain model thus offers objectivity of consensus under the assumption that at least “some” of the accounts are honest, whereby the required percentage of honest accounts depends on the relation between their life expectancy and the heartbeat period. As the heartbeat period can theoretically be set arbitrarily low, the protocol can reach objectivity even in the presence of an extremely powerful attacker.
The article is based upon the model that I briefly outlined here:
https://bitcointalk.org/index.php?topic=1719396.msg17365846#msg17365846.
Any feedback would be highly appreciated.