I think the title and abstract of this are not very clear.<p>I gave it a fast skim to figure out what general class of thing it actually is.<p>This should be compared with "proof of idle" (<a href="https://www.cs.virginia.edu/~shelat/14s-pet/2014/02/11/proof-of-idle.html" rel="nofollow">https://www.cs.virginia.edu/~shelat/14s-pet/2014/02/11/proof...</a>).<p>It is an online scheme for resisting sybil attacks in a P2P network where nodes have cryptographic identities which works by periodically forcing all users to do proof of work within a limited time window. Peers that don't respond fast enough are banned from the system (have to create a new identity to join, which is computationally expensive).<p>The idea is that this get some of the benefits of POW for sybil resistance without spending as much energy.<p>It doesn't, however, produce a large amount of cumulative work building up over a history. So it's not the sort of thing you'd want to use to protect the history of a ledger directly.
If new coins are distributed as block rewards, then collectively, miners will spend (nearly) the monetary value of those coins in competition to solve as many blocks as possible. In other words, any gains in marketcap will translate to increased mining efforts. Paul Sztorc has written a lot more about this in <a href="http://www.truthcoin.info/blog/pow-cheapest/" rel="nofollow">http://www.truthcoin.info/blog/pow-cheapest/</a>
I'm in the process of auditing new low-power Po* algos for a crypto system I'm researching. This paper needs a way better abstract. The math is incomprehensible without knowing what it's end goal is. The paper uses acronyms (ex "Good ID") before explaining what they are. Overall a poorly written paper. The work maybe revolutionary, but what good is that if it's so difficult to understand. People forget that the purpose of a paper is to communicate your idea to other people. If it doesn't do that effectively then no matter how awesome the work is, you've failed.
Here's some "lore" around proof-of-work and why it must burn external resources:<p><a href="https://download.wpsoftware.net/bitcoin/asic-faq.pdf" rel="nofollow">https://download.wpsoftware.net/bitcoin/asic-faq.pdf</a><p><a href="https://download.wpsoftware.net/bitcoin/pos.pdf" rel="nofollow">https://download.wpsoftware.net/bitcoin/pos.pdf</a><p><a href="https://download.wpsoftware.net/bitcoin/alts.pdf" rel="nofollow">https://download.wpsoftware.net/bitcoin/alts.pdf</a>
For those who interested in the topic. Nem uses something called proof of importance <a href="https://www.nem.io/NEM_techRef.pdf" rel="nofollow">https://www.nem.io/NEM_techRef.pdf</a>
There is something thoroughly logical about this approach to the unfortunate waste implicit in Proof of Work (PoW) schemes.<p>In summary of the motivations of the paper: PoW is currently limited to cryptocurrencies as a security system because of the implicit financial incentive in mining coins. This approach could allow PoW schemes to be widely adopted to secure systems as the overhead is lowered dramatically. Additionally, battery powered devices (ie phones) could make use of PoW without incurring large battery drains. This last bit is particularly interesting and could allow some interesting, distributed P2P systems on cellphones to arise.<p>The general principle of the paper revolves around asking network members to prove computational power only as much as necessary as the network scales. Because an attacker could easily spoof their MAC / IP address when joining the network, computational tests are periodically distributed to network members. If the test is unsolved in an allotted time period, their network membership is revoked (and the node is blacklisted). The attack referred to in the paper is an attacker adding bad (fake or otherwise) nodes to a system rapidly.
“Consequently, if the network is attacked, our scheme guarantees security, with algorithmic costs that are commensurate with the cost of the attacker. Conversely, in the absence of attack, algorithmic costs are small.”<p>Hope it works.
So now we have click-baity academic papers? Can someone summarizes to me the genius behind the idea? Because as far as my math goes, it is not theoretically possible.
When I read this: "This shortcoming is highlighted by recent studies showing that PoW is highly inefficient with respect to operating cost and ecological footprint." The moment we try and do away with this inefficiency we are going against the entire solution of what bitcoin was going for: how to use all of the inefficient computer parts laying around the globe for something useful?<p>It makes me think people just don't get it. High inefficiency is the _ENTIRE_ point! It gives rarity to the coin. That 10 minute block time is the same as compressing millions of years of geology into 600 seconds. It is fundamentally sound mathematics, and ultimately, <i>that</i> is why it holds its value. It is not meaningless mathematics, people who say that don't understand fully as to what they are talking about, imho.