TEA Security

DDoS Attack

Although a DDoS is hard to avoid, TEA can reduce the benefits of DDoS attacks through anonymity and randomness. When the number of nodes in the entire network reaches a certain level (for example, TEA modules are attached to a large number of IPFS mining machines), then the DDoS damage is minimal.

If a great number of nodes are attacked and suddenly lose trust, how does the TEA network handle this attack? How long before the network is able to recover?

First of all, if an attack is initiated by an untrusted node, it will have little impact on the trusted nodes. Because trusted nodes only communicate with other trusted nodes, it will not accept any network communication from an untrusted node. If the attack is through DDoS, that means the router is under attacked and the machines can’t handle it. A DDoS attack is beyond the scope of blockchain protection, just like any server’s design cannot solve unexpected power outages. These kind of external attacks can only be solved through data center facilities and economic rewards/punishments.

For a trusted node to be attacked successfully, there must be more than 2/3 of the nodes becoming bad actors at the same time. Any attack that doesn't reach this threshold will automatically be killed, and all assets of attackers will be confiscated. The remaining 2/3 who didn't participate in the attack would make a fortune as these miners token holdings would capture the forfeited assets of the attacking 1/3.

Because TEA uses a CML licensing system that isn't free, it'll be very expensive to gather a large number of nodes to cause disarray on the network in a short period of time. It would take an attacker several years to lie in wait before they can finally control a majority of the nodes. Time becomes a huge cost for the attacker. In the mean time, they could've attacked many other blockchains several times over and made a good fortune. So it's not clear why they'd go through all the extra hoops to attack the TEA network.

Due to the NFT properties of TEA’s CML, each CML is different. When the technology matures and the scale is large enough, the diversity of the technical stack will grow to the point that it's difficult for a single technical stack to occupy 1/3 of the nodes, nevermind 2/3. When the network is just beginning, it's relatively easy for one popular tech stack to reach or exceed 2/3 of the network nodes. Therefore, we will be very cautious in the initial stages of issuing CML. We will make sure that the interests are positively related to the project in order to obtain CML. This way, the probability of deliberate attacks will be much lower.

Optimizing the mining rewards such that miners receive more revenue from normal work rather than attacking the system is another solution for preventing attacks. Once the benefit of helping the network is greater than doing evil, then there's much less possible incentive to do evil.

Last not the least, because of the decentralization and zero-knowledge of the data, even if the attack succeeds, the benefits obtained are usually 0. You don’t want to take a huge risk to rob a homeless person who may be poorer than you.

This theory is explained in the economic design of the white paper. The security of the TEA network is based 60% on the economic design of the project and 40% on computer technology.