Skip to main content

Omni Security Model

In the Omni network, validators play a pivotal role by managing the Omni EVM and facilitating cross-network message requests among external rollup VMs. To guarantee the integrity of validator actions, the Omni protocol incorporates a dual staking mechanism, securing the network through a combination of re-staked ETH and staked OMNI. A significant advantage of this approach is Omni can immediately access Ethereum’s 30.6M ETH security budget. This provides Omni with an unmatched level of security assurance among interoperability protocols, setting a new standard for network safety and stability.

One of the challenges with introducing a dual security model is ensuring augmentative security. Let us propose an architecture for dual staking that does not achieve this: modular dual staking. For each consensus game in the network, Token A stakers reach quorum on the outcome while Token B stakers independently come to their own quorum on the outcome. If one of these independent networks does not reach quorum, the vote does not pass.

Each network possesses its own security function: sa(A),sb(B)s_a(A), s_b(B). Under this model, the cost to violate safety SS of the network is defined as:

S=23min(sa(totalAStaked),sb(totalBstaked))S = \frac{2}{3} \cdot \min(s_a(\text{totalAStaked}), s_b(\text{totalBstaked}))

Additionally, the cost to violate liveness LL of the network is defined as:

L=13min(sa(totalAStaked),sb(totalBstaked))L = \frac{1}{3} \cdot \min(s_a(\text{totalAStaked}), s_b(\text{totalBstaked}))

As we can see, the total cryptoeconomic security of a network under the modular dual staking model is determined by the minimum security derived from Token A or Token B. Effectively this approach favors implementation simplicity over security.

The Omni protocol implements its security model according to a native dual staking model. Instead of using two independent networks that reach quorum separately, Omni treats both as one set. Thus, the cost to violate safety SS of Omni is defined as:

S=23(SETH(totalETHstaked)+SOMNI(totalOMNIstaked))S = \frac{2}{3} \left( S_{\text{ETH}}(\text{totalETHstaked}) + S_{\text{OMNI}}(\text{totalOMNIstaked}) \right)

Additionally, the cost to violate liveness LL of Omni is defined as:

L=13(SETH(totalETHstaked)+SOMNI(totalOMNIstaked))L = \frac{1}{3} \left( S_{\text{ETH}}(\text{totalETHstaked}) + S_{\text{OMNI}}(\text{totalOMNIstaked}) \right)

Omni’s implementation of this native dual staking model enables greater security but comes with more implementation complexity. Specifically, the protocol will implement separate functions for mapping existing stake to voting power and for mapping existing stake to staking rewards. This will allow the protocol to dynamically incentivize validators to contribute more security from either ETH or OMNI depending on market conditions. Altogether, Omni is able to bootstrap an unparalleled security budget from the outset using re-staked ETH while establishing a mechanism for its total security to grow over time with the addition of staked OMNI.