TorCoin

https://drive.google.com/file/d/0B7r4osQgWVqKTHdxTlowUVpsVmJRcjF3Y3dtcTVscFhEaW5F

by chatmasta • 12 years ago

266 64 12 years ago

drive.google.com

chatmasta4• 12 years ago

I'm Miles Richardson, one of the authors. This research was my Yale senior project, in collaboration with fellow student Mainak Ghosh, Professor Bryan Ford, and Tor community leader Rob Jansen. I thought now was a good time to post, with the EFF initiative on the front page of HN.

We will be presenting our research at the HotPETS 2014 privacy conference in Amsterdam, in July. Please come by if you can. :)

At a glance, the point of this research is to introduce an alternative cryptocurrency, called TorCoin, that rewards relays with coins for transferring bandwidth. Relay operators can then sell those coins on any altcoin exchange. TorCoin is like Bitcoin, but it's bandwidth intensive, rather than CPU intensive. So in effect, relay operators "mine" TorCoins and make money selling them on exchanges.

Let me know if you have any questions or want to discuss. The next step is developing a prototype, and/or a network simulation to run experiments. Anyone interested in helping, let me or any of the authors know. Email is miles.richardson@yale.edu

rgbrgb• 12 years ago

Thanks for posting. It's fantastic to see a cryptocoin where the mining isn't just wasting cycles.

isaacwaller• 12 years ago

What would give "TorCoins" value? It is obvious relays would sell their coins on some altcoin exchange, but who would be purchasing these coins? It seems that the only people purchasing TorCoins would be people looking to donate to Tor relays (this is a totally valid use case, but I was wondering if you have something else in mind.)

Looks really cool by the way.

chatmasta4• 12 years ago

Yes -- This is one of the open questions for discussion, and one of the most important. (HotPETS is a "workshop", which in Academia is the precursor to a conference... papers are meant to be early stage drafts ready for discussion. So we are curious how people would be interested in the currency.)

You're right, one of the early reasons for purchasing TorCoins would likely be donating to Tor. But they are a totally usable altcoin, just like Dogecoin for example. What reason do people have to buy Dogecoin? Not much, beyond using it as a currency or speculating with it as an investment. I could see TorCoin working the same way.

sathishmanohar• 12 years ago

I thought miners get coins for transmitting bandwidth which they can sell at exchanges, and people who do not participate in transmitting but want to trasmit buy these coins to consume bandwidth in the network. It would make sense.

Essentially miners would be bandwidth providers and consumers buy torcoin to pay miners. The monetary value of torcoin will be decided by free market.

nullc2• 12 years ago

It also appears to rely on trusted authorities who could simply lie to their own advantage, and appears to suggest no strong mechanism to make this process robust against sybil or collusion. E.g. I can just very cheaply claim to be running millions of nodes— 99.99% of the whole network, unless you pin the existence of trusted participants who could at least force my fake nodes to be reachable (and even then botnets make forgery inexpensive), but if you go the route of pinning certain trusted participants, why not just have them hand out the reward?

While this might be an interesting resource allocation technology, and tor itself has strong assumptions about non-collusion of directory authorities (and their services like bandwidth authorities), it seems disingenuous to compare it to Bitcoin.

chatmasta4• 12 years ago

This is our primary concern. Our goal for this paper is to maintain at least the status quo of anonymity in the Tor network. Currently, every Tor client uses a trusted group of ~7 "directory servers" to assign them to a circuit. Our implementation relies analogously on trusted "assignment servers" to assign clients to circuits. That's the purpose of the TorPath assignment protocol. In fact, it provides better anonymity than the status-quo, because each node on the circuit only knows the IP address of its neighbors, rather than the current situation, where the client knows the IP of every relay on the circuit. So yes, this assignment protocol relies on trusted servers. But so does Tor in its current state.

Your point about botnets is a very good one, and one we haven't considered fully. We will certainly address this further in future versions of the paper.

However, we do consider sybil attacks. Addressing them is the primary purpose of the TorPath assignment protocol. It ensures that, even if an adversary controls over 50% of the relays, there is a very low chance that (1/16 if I remember correctly) a new circuit will consist of three evil relays. Since we limit the number of coins that can be mined on a given circuit, we call this security a "good enough" margin of error for bad coins. Since it's predictable it can be baked into economic models.

But you're right, these are important considerations that we are taking very seriously.

GrinningFool3• 12 years ago

Proof of bandwidth is an innovate idea in the cryptocurrency space, but it seems a weak point is "make money selling them on the exchange".

If that is their only value - if there is no economy built around them and they have no purchasing power on their own - then they'd only see a brief pump-and-dump rise/fall like any other new altcoin.

Certainly it's an experiment worth trying, but if its sole value is in its potential for exchange, I don't see anything preventing it from becoming yet another short-lived altcoin.

Then again, the same could be said of many coins out there now that have been lingering whole weeks* without dropping to 0 value

1. some have even been surviving for months...

notastartup2• 12 years ago

Assuming that there will be more nodes on the Tor network as a result of a financial incentive for running such nodes, how will the TorCoin deal with over supply of nodes, therefore the drop in payout per node?

Can TorCoin be used to subsequently purchase more bandwidth? How would this bandwidth differ from the free bandwidth? Lower latency?

aburan28• 12 years ago

Maybe this somehow be a away of rewarding bitcoin full-nodes

facepalm2• 12 years ago

TL;DR: why don't you call your paper "TorCoin.pdf" instead of "paper.pdf"?

Minor nitpick: what is up with PDF naming? When I "save" the PDF to my computer I don't even get the choice to rename it (I guess I could configure that but I am too lazy). So now I have a PDF called "paper.pdf" on my HD. If I ever want to find the paper on TorCoin again, it will fail. This problem seems very common in the world of academic papers.

chatmasta4• 12 years ago

Sorry about that, I changed the name for you. :)

facepalm2• 12 years ago

Thanks! Well I think it's not just for me, it's just basic "Search Optimization" :-) In fact your paper was downloaded as "(paper (2).pdf" because there was already at least one other paper.pdf in my Downloads folder.

etiam• 12 years ago

I would like to comment that while the intentions seem very good, I think it may be a severe mistake to use monetary compensation as a means to increase the number of TOR relays.

Framing a situation as an economic transaction has been shown to interfere with other motivations. The classic research paper would be http://papers.ssrn.com/sol3/papers.cfm?abstract_id=180117 where introducing a fine to make parents pick up their children at daycare considerably worsened the situation. The interpretation is that behaviour that was evaluated in a context of prosocial motivations such as being decent towards the staff, keeping a good relation with them, etc, were supplanted by a view that 'it's worth the cost of the fine'. It is worthy of note that going to an economic reference frame is often easy whereas going from an economic reference frame to one based on social values is frequently difficult - it certainly can't be expected to recover spontaneously just because the money is taken out of the picture again.

Now, TOR is very much about prosocial values - liberty, free speech, the right to communicate without being unjustly surveilled, and so on, and because of that, I think it is potentially quite vulnerable to these kinds of effects.

I think there is a substantial chance that running a TOR node is not going to be worth the economic profit in the long run, but that shouldn't matter, because running a TOR node is about something other and something more important than making an economic profit. If we start framing it as a business transaction there is a risk we hurt the (in this context) much more important motivations to operate a node. I think the TOR initiative should be developed on motivation based in the human rights values it naturally supports. And in this case I believe the option of monetary compensation is genuinely at odds with that.

I'm somewhat sad to be saying this, because as I mentioned, it seems to me that the project is based in the best of intentions. Some components of the TorCoin system may be valuable for other service functions in the network.

I summary, I think the psychological effects of introducing the compensation scheme really should be carefully evaluated before any attempts to introduce it in practice. My spontaneous impression is that it is the wrong method of motivation for growing and maintaining the TOR network and should not be introduced at such. If it is after all introduced, that should be done in a manner that is informed of the psychological consequences.

As a preferred alternative we should emphasize the social and ideological benefits of TOR for core motivation and lower the threshold for contributing by making it technically trivial for the contributor and by making the situation as clear as possible about risks and benefits of operating a node.

VMG4• 12 years ago

It's much better when running a Tor node is a decision that is evaluated in an economic reference frame.

There are very few tor nodes currently because there societal incentive is very low. If you can actually start making money off it, many more people will start doing it. It's a concept so basic I'm wondering how one can miss it.

And the paper you linked to not applicable to the situation at all.

vidarh• 12 years ago

It's much better when running a Tor node is a decision that is evaluated in an economic reference frame.

That may be true if the economic incentives available are sufficiently strong, but that is not a given.

cornholio2• 12 years ago

Specifically, while there is a strong social pressure involved when dealing with your children's tutors, there is almost none when dealing with complete anonymous people on the internet, many of which are likely to be criminals.

I would like to see an attempt to quantify the two effects, but I would venture to say that switching from an altruistic motivation to a financial motivation will increase the bandwidth available by orders of magnitude. For example, compare private trackers with public trackers.

gnarbarian6• 12 years ago

Using market forces to incentivise this is a great idea. Running a node incurs a quantifiable cost in terms of bandwidth, causes authorities to watch you more closely, and exposes you to a risk of litigation over the content you are relaying. Ideals regarding privacy, the freedom of information, and anonymity aren't enough for most people to surmount the risk and cost of operating one.

As it stands I wouldn't be surprised if the majority of nodes were operated by various national intelligence agencies simply because they have vastly more incentive to do so than anyone else.

Potentially being able to profit from this changes things dramatically and gives people who wouldn't consider themselves stakeholders actual reason to 'sell' their bandwidth to the tor network.

dllthomas4• 12 years ago

The parent explicitly assumed that running a tor node would not wind up profitable. If that's true, then no one is going to do it for purely economic reasons (even if this makes the economic situation better) and this might be harmful. If that's not true, then certainly we'll see more people start doing it.

shazow• 12 years ago

+1.

In fact, it would probably be better if the whole idea of Tor was further removed from prosocial values. This will help disassociate the idea that contributing to and using Tor is for people who have something to hide, and turn it into an economic decision.

At the very least, if the economics of using Tor were closer to neutral. Right now it costs money to run a Tor relay, would be nice if it was self-funded.

exo762• 12 years ago

I think that situation is very different. Tor nodes operators are running nodes not for Tor users. They are running nodes for idea of free, open Internet. They don't ever socialize with Tor users. Social as in "pro my individual view of society" and social as in "pro my relations with people around me" are two totally different meanings.

Parents interact with daycare staff. Tor relay operators don't interact with Tor users.

wodencafe• 12 years ago

This just isn't true.

I run a relay, and I interact with all sorts of Tor users on irc.oftc.net #tor

In fact, there are lots of relay operators and regular tor users there on IRC, and some of them get pretty chatty.

It's good to interact with and get to know the community you're involved with.

DennisP• 12 years ago

Early adopters of Bitcoin had similar idealistic motivations, but I don't think the financial rewards hurt.

belorn• 12 years ago

On the positive side, such system could help producing symbiotic relationship with new projects. P2P has for example been discourage because of the damage it could cause to the performance in the network. However, with an enforceable share ratio system in place, you could make sure P2P contributed as much as it used.

And with more users, the privacy is increased. That said, incentive schemes are famous topics for sociology and economics research, so there is validity in being careful.

3pt14159• 12 years ago

It is spelled Tor, not TOR. Also I partially agree with you, but more so because mining leads to centralization than anything else.

naturalethic• 12 years ago

Someone call the ghost of McCarthy.

notastartup2• 12 years ago

I would argue that the underlying activity that leads to a cryptocoin being awarded has been the single point of flaw that prevents it from being reliable. If running a tor node and you are able to contribute to the greater good, I think that is a far more valuable proposition than calculating increasingly complex problems across multiple graphic cards that produces absolutely no value other than consume electricity and generate pollution.

I can definitely see the problem if an incentive was offered say on a stackoverflow type of website, because it would directly hinder the quality of the questions but in the case of running a Tor node, it's only a matter of turning it on or off. If someone decides to run it on a better hardware, they would be compensated more.

I'm really excited for TorCoin, it is a serious alternative to Bitcoin.

furyg3• 12 years ago

Slightly OT...

Assuming I'd like to 'donate' to TOR by just running a node myself, what would be a cost-effective hosting provider to use for this?

This would be a relay node, not an exit node (I can imagine that hosting providers wouldn't be so happy about exit nodes).

tech-no-logical• 12 years ago

I've been running a small exit node for over 5 years, with three providers, none of which have complained (I'm in the netherlands, that might matter).

there's one disadvantage though : a lot of sites refuse my ip because it's an exit node. which is no problem because all my own traffic runs through a vpn anyway...

wcoenen• 12 years ago

Couldn't a tor client establish a bitcoin micropayment channel[1] with each relay node instead? This seems much simpler than introducing a separate currency based on proof-of-bandwidth.

[1] https://bitcoinj.github.io/working-with-micropayments

jabgrabdthrow• 12 years ago

Ok, 90% of these "useful mining" projects get the economics backwards. Paying coins to miners does not generate demand for the coin! The miners (in this case, tor nodes) will sell and push the price to the ground. Who will buy the coins? Other altcoin enthusiasts who also don't get the economics?

(And yes, this argument theoretically applies to Bitcoin, but you have to use bitcoins to pay the transaction fee to use the bitcoin network, which has value because of its network effect for use in value transfer).

oleganza• 12 years ago

Your last statement is circular. "Network effect in value transfer" exists only when there is "network effect in seeing value in BTC units". You can't transfer value before something has value. Also, act of transferring or fees does not generate value either.

Bitcoin units are rare collectibles. Like gold coins or USD paper bills. They are scarce, hard to counterfeit and they are portable and fungible. They are easy to collect. Once you have a club of people who collect them and speculatively value them, this club can grow (or shrink). If such token is superior than others, then some people will likely sell their inferior tokens in order to acquire superior ones. This way the club grows.

Plus, there is value in liquidity. People generally want to store wealth in something "secure" which means not only secure in terms of storage (personal vault, swiss bank account etc.), but also in terms of purchasing power. The more and quicker I can buy myself something I suddenly need, the more secure my wealth is. E.g. if my computer breaks and I need a new one, my cash will help me buy a new one better than a collection of postal stamps. So people will generally prefer to keep savings in money that is more marketable, i.e. valued by more people. This means that once a superior collectible starts to grow, it will drive the value of inferior collectibles down to zero. (However, if there are currency controls and banking regulation in your country, then you might be stuck with a less marketable local currency.)

Nick Szabo explains how money evolved: http://szabo.best.vwh.net/shell.html

I explained why we all want one universal money: http://blog.oleganza.com/post/54121516413/the-universe-wants...

jaekwon• 12 years ago

Does this protocol use bandwidth to achieve consensus, or is there another consensus mechanism such as proof of work assumed to be available?

chris123• 12 years ago

Who is working on a "BittorrentCoin?

TorCoin

Related Stories