Ultra light client

[kimurayu45z]

When I consider the competition between IBC and LayerZero, LayerZero is not better than IBC at all as a TAO layer, but I can conclude that even after IBC Eureka, LayerZero has slight competitiveness with the advantage of cost of "Ultra Light Node".

Ultra Light Node can be regarded as a light client model using Oracles to provide the block headers.
The cost of this should be lower than ZK light client.

	ZK Light Client	Ultra Light Client
Trustlessness	〇	△
Cost	△	〇

Proposal

How about supporting "Ultra Light Client of Cosmos Hub in Ethereum" as a function of Cosmos Hub?
The implementation can be like using Secp256k1 ECDSA Threshold Signature Scheme of Cosmos Hub validators.

ZK Light Client still can be like "ZK light client SDK" to establish P2P IBC connection between VM chains and sovereign Cosmos SDK chains, which can keep IBC full trustless protocol.

Users can choose whether

• Connecting direct P2P IBC connection between VM chains and sovereign Cosmos SDK chain
• Using packet forward middleware of CosmosHub by using cheap Ultra Light Client to interact with VM chains

after seeing the trustlessness of ZK light client connection between Cosmos Hub and Ethereum as a show case.

In this case, LayerZero will completely lose the competitiveness against IBC Eureka, and Cosmos Hub will have greater revenue.

[yihuang]

Threshold signature scheme is based on the number of shared, how to adapt that to staking based?