What stops an attacker from crippling the Bitcoin network?

Question

An individual (or government) who wants to bring down the Bitcoin network could conceivably broadcast a ton of junk/illegitimate transactions to the network thereby consuming the bandwidth and processing power necessary for the network to deal with legitimate transactions.

I understand Bitcoin has two protocols in place to address this vector of attack:

1) First and foremost it has each node verify all information received before relaying it to its peers. As such, an attacker would barley get past first base (i.e, legitimate node 1) to be able to cause catastrophic damage to the network at large

2) banscore (which effectively computes how bad a node is behaving. Once it reaches a certain critical threshold a neighboring node will stop receiving connections from that node/IP for a period of time)

My main question is: considering the fact that the government (or a government run entity) controls IP allocation isn't this solution completely reliant/dependent on a centralized entity to work effectively?

Answer 1

...could conceivably broadcast a ton of junk/illegitimate transactions to the network thereby consuming the necessary bandwidth and processing power for the network to deal with legitimate ones.

If the transactions are invalid, then they would not be relayed by the network's nodes, making the attack somewhat ineffective. At best, the attacker could consume computational resources of peers they are directly connected to, though if they are spamming invalid transactions then they would probably be banned by their peers soon.

If the transactions are valid, then the attack becomes expensive, which is a disincentive over the long term. It would be extremely expensive to fill every block with 'spam' transactions, and there will still be other users that will compete (in terms of fees) to have their transactions confirmed first.

If some entity attacked bitcoin via IP/DNS blocks, this would not entirely shut down the network, though it would likely cause disruptions to some nodes. I believe that nodes which connect through Tor, or through other methods (such as by satellite) would still function without issue.

Answer 2

Attackers are financially disincentivized to do this.

• They have to produce enough transactions, frequently enough to fill up blocks and keep them full
• The transaction fee has to be large enough so they cause an actual clog instead of other users outbidding them with higher fees

To keep the above two points going for any meaningful amount of time costs a lot of money. It would only be temporary and only costs the attacker money. UTXOs still belong to their rightful owners and users on the network can start transacting again once the attacker loses interest and money.