We can see Bitcoin as a peer-to-peer network of cooperating nodes. These nodes listen for transactions, order them into subsequent blocks, and publish these blocks on the network. Users use digital signatures to prove ownership of funds when creating transactions, and nodes use a proof of work system based on hashing to rule out double-spending. These techniques bring trust to the whole history of transactions without the need for a central authority, and this trust allows users to exchange value. Let us now unveil the fundamentals of this system.
Brief History
The initial era of Bitcoin is somewhat mysterious. The domain bitcoin.org was registered on the 18th of August 2008. On the 31st of October, Satoshi Nakamoto, whose identity is unknown, published a paper describing a peer-to-peer electronic version of cash named Bitcoin and subsequently released an open-source implementation. On the 3rd of January 2009, the genesis block was mined, which brought the Bitcoin network into existence. The first transaction was made on the 12th of January 2009. Since then, the network and code base have kept growing, and Satoshi Nakamoto disappeared from the Internet. To this day, there has been no successful attack that would deprive the owner of the funds they possess.
Brief Overview
If we wink at the Bitcoin Protocol as much as we can, we can reduce it to the following steps:
- Clients broadcast new transactions.
- Some nodes gather incoming transactions into a block and start finding a proof of work for this block.
- Once a node finds the proof of work for the block, the node broadcasts the block in the network.
- All other nodes verify the block and accept it only if all transactions in it are valid and not already spent.
- Nodes express their acceptance of the received block by starting to create another block on top of the received one. They do so by inserting the hash of the received block into the one currently being created.
Wallets and Addresses
Before we start our discussion on transactions, let us discuss Bitcoin wallets and their addresses. We will regard a wallet as a public/private key pair. Note that this definition differs from the usual notion where wallets refer to whole software suites that handle Bitcoin for their users.
A user generates a wallet on their own, and they can generate and handle any number of them. The public key of a wallet represents its address. The user keeps the private key secret, and they can use it to manipulate the funds belonging to the wallet. The user then shares the address so other users who possess Bitcoin can send their funds to this address by broadcasting a transaction. We describe this process in sections Chain of Transactions and Chain of Blocks. Besides mining, there is no other way to obtain Bitcoin. Note that there is no need to register anywhere. Also, note that the wallet is all the user needs to access their funds. This means the user is not bound to any device while using Bitcoin. If the user wishes, they can generate a new wallet, even with no internet connection, share the wallet address, retain the key pair, and then dispose of the device they used. The user can subsequently travel to the other side of the world only with the key pair and then use it to access their funds on any device with a Bitcoin client installed. This is possible due to the public nature of the blockchain that we describe later on. However, the user loses the funds forever if they lose the private key. Also, once the private key gets exposed to an adversary, this adversary instantly gains complete control over the funds.
Another property of Bitcoin is that transactions are irreversible. This means that once a user sends their funds to the wrong address (even an address that has no associated private key), and this transaction gets written to the blockchain, there is no way to revert the transaction, given the blockchain operates as expected.
A Bitcoin address is a string of length between 26 and 35 of alphanumeric
characters, beginning with the number 1, 3 or the string bc1. The latter
type is case insensitive, while the two preceding ones are case sensitive. No
address contains the uppercase letter O, uppercase letter I, lowercase
letter l and the number 0 to prevent visual ambiguity. An example of a
Bitcoin address is
1BvBMSEYstWetqTFn5Au4m4GFg7xJaNVN2
or
bc1qar0srrr7xfkvy5l643lydnw9re59gtzzwf5mdq.
Modern Bitcoin clients (usually called wallets) implement so-called hierarchical deterministic (HD) wallets for deriving the key pairs from a seed according to BIP-0032. This approach makes Bitcoin wallets more user-friendly and has no impact on the Bitcoin protocol. The seed is a random 128-bit number that can be encoded into a human-readable format. The seed is then called a seed phrase and consists of 12 English words that the user has to keep secret. The wallet can then deterministically create a virtually unlimited number of addresses. The same seed can also be used on multiple devices so the user can control the exact set of addresses on any of these devices. The user doesn’t have to retain any of the private keys. It is crucial to generate the seed at random to prevent the possibility of a collision with another seed. In the case of a collision, the consequences are that both users end up with the same set of key pairs.
Privacy
We can’t consider Bitcoin private or untraceable since all of its transactions (containing addresses and Bitcoin amounts) are public. If we focus on a particular address, we can see all incoming and outgoing transactions that are already in the blockchain and observe any future transactions. Conversely, we can see all source and destination addresses if we focus on a particular transaction. Combining these two approaches allows us to track and backtrack all transactions associated with any address. It is also possible to see how much funds each address holds—there are online blockchain trackers that offer all the details at a glance to anyone. If we manage to associate someone’s identity with a specific address, that user’s privacy is lost. The information about the complete transaction history is immutably and publicly stored in the blockchain forever.
It is possible to prevent potential network surveillance or censorship by using the Tor network to connect to the Bitcoin network. Some clients have this functionality inbuilt.
Chain of Transactions
All transactions in the blockchain are public, and anyone can verify their validity. Transactions~\cite{transactions} in Bitcoin have the following content (simplified):
\begin{description} \item [hash of itself] Every transaction is identified by its hash. The implementation uses SHA-256. \item [input] Contains the following: \begin{description} \item [previous transaction] A hash referencing the previous transaction. \item [signature] The hash of the current transaction is signed with the private key of the owner. \end{description}
\item [output] Contains the following:
\begin{description} \item [value] The amount of bitcoins that the sender is willing to send. \item [receiver’s address] The public key of the receiver’s Bitcoin address. \end{description}
\end{description}
This transaction structure creates a chain of transactions since the input references the previous transaction. We can notice that the transaction also contains the receiver’s address and that the whole transaction is signed (containing the hash of the previous transactions). This in fact creates a chain of trust~—~the receiver (and anyone else) can verify that the sender owned the funds that are being sent in the following way:
The receiver looks at the transaction referenced in the input, and gets the sender’s public key from it. Subsequently, they verify that the signature of the current transaction is valid. If this is true, we can be sure that the funds are transferred (to the owner of the public key specified in the output) only by the owner of the private key that corresponds to the public key stored in the output of the referenced previous transaction.
A simplified chain of transactions is depicted in Figure~\ref{fig:transactions}.
\begin{figure}[!htb] \begin{center} \includegraphics[width=1\textwidth]{./images/blocks} \caption{Outline of transactions and their ordering.~\cite{Nakamoto}}\label{fig:transactions} \end{center} \end{figure}
Transactions can have multiple inputs and multiple outputs. This allows the value to be split and combined. All inputs are summed up, and this sum has to be spent in the outputs.~\footnote{Unused funds are used as an incentive for miners, see Section~\ref{sec:Chain-of-Blocks}.} If we don’t want to use all the funds specified in the input, we can simply send it back to the same wallet that we currently use for the transfer.
The approach described above doesn’t rule out double-spending. The owner could transfer their funds multiple times. To prevent this, we could introduce a global authority that would check every transaction. Bitcoin addresses this issue without the need for such an authority.
Chain of Blocks