This is an interesting scheme: a digital currency with no central issuing or verification authority. I first heard of it in relation to silkroad, an anonymous marketplace, apparently commonly used for trading various contraband, where bitcoin is/was used as an anonymous digital currency, a useful thing if you want to trade contraband anonymously…
So how does bitcoin work? Well, the initial paper is written very clearly. And here is my quick summary:
- Bitcoins are merely transaction records.
- Transactions are between wallets.
- A wallet is a record protected by public key cryptography. If you own the private key for the wallet, you own the contents of the wallet, as recorded in the transaction history.
- You send bitcoins by creating a transaction moving a value from your wallet to another wallet, and signing that transaction with your private key to prove the wallet owner intended it. (You receive bitcoins by somebody else doing the same thing to you…) You then submit that signed transaction to the network, to be hardened into the global transaction history at the next opportunity:
- In order to prevent people double-spending their balance (ie, spending the same money twice) the entire transaction history of the entire bitcoin currency is a matter of public record, stored in a distributed database, maintained by the peer-to-peer network. A degree of anonymity is afforded by obfuscating the relationship between wallets and wallet owners.
- In order to solve the problem of issuing currency in an authority-less system, bitcoins are brought into existence at ~10 minute intervals in a lottery-like system (where increased computational power buys you increased likelihood of winning new currency).
There are some nifty implementation features of bitcoin which make it tolerant of dishonest nodes attempting to corrupt the system.
- Proof-of-work system: others will only accept (consistent) modifications to the global transaction history if those modifications are accompanied by proof that the editing node solved a computationally hard problem. The protocol adjusts the difficulty of the problem such that the entire CPU power of participating nodes will only generate 1 solution roughly every 10 minutes. (The reward for expending effort to solve the problem is new currency assigned to the solver, as agreed by participants of the network, and/or transaction fees from those having their transactions appended to the global history.)
- The block-chain: each edit to the transaction history (an edit is known as a ‘block’ in bitcoin parlance) reinforces prior blocks by incorporating a cryptographically secure hash of the previous edit (guaranteeing that the new block is newer than the previous block), resulting in a continuous chain of blocks reaching back to the start of the currency. This guarantees a globally-observed, decentrally generated, clock-tick to peg transactions to in order to prevent double-spending of currency. The proof-of-work system makes it infeasible to rewrite the transaction history because you’d have to rewrite the whole history from your edit to the present, and do all the proof-of-work that would entail (which would entail overpowering the rest of the network in CPU power terms, which is typically infeasible, and in any case, that CPU power could be used to legitimately ‘mine’ new currency by participating honestly…)
Anyway, proper details are through the links. I thought it was interesting. I like the idea of checks and balances being implicit in the protocol and decentrally administered: it seems more secure than having to trust central authorities.
However, bitcoin is not as anonymous as you might think: wherever the bitcoin network interfaces with the wider economy (ie, through currency trading with normal currencies, or through the transfer of goods and services), or through network traffic analysis, there is potential opportunity for identifying data to leak through, so as with all security technologies, bitcoin is not a magic bullet, it’s just an interesting piece of component technology.
I wonder when technology like this will find its way into widespread use.
