Idea for BIP : Deterministic Wallets with Token support

[Forrest96er]

Hello,

The number of new tokens for Ethereum and Ethereum-like coins has increased dramatically. However, the wallet structure for managing multiple coins based on a single seed has not been updated to accommodate this new scenario. Currently, all tokens are managed using the same derivation path, resulting in the creation of identical addresses across different tokens, significantly reducing privacy. To address this issue, the wallet structure for HD wallets needs to be updated.

Simply adding an additional node to the derivation path is not practical for various reasons.

A better solution is to use the address or identifier of the token for creating private and public keys. This can be achieved by adding an additional input to the HMAC function, which is used to generate child private and public keys. It is advisable to apply a collision-free hash function before using HMAC.

m / purpose' / coin_type' / account' / change / index

I recommend applying the modification at the "Change" node. Without modification, the creation of an address for the base coin (no token) is targeted.

With the modification, the token- adress is targeted.

This approach also has the advantage that if hardware wallets are used, only the extended public keys of a coin need to be exported once to the front-end application. After that, the front-end application can generate all public keys needed to scan for transactions on all tokens. Even if a token did not exist at the time of the public key export, it could later be found without any additional export.

  Did I miss something?
If an attacker obtains some public keys used in a transaction for a token, he should not be able to calculate the public keys of other tokens or the base coin.  

[Aneesh Karve]

Not sure this is relevant to a Bitcoin list but I'll answer in a Bitcoin context.

> Simply adding an additional node to the derivation path is not practical for various reasons.

What are those reasons?

> I recommend applying the modification at the "Change" node.

The change node does not use hardened derivation and is therefore unlikely to suit your security needs. From BIP-32:

> One weakness that may not be immediately obvious, is that knowledge of a parent extended public key plus any non-hardened private key descending from it is equivalent to knowing the parent extended private key (and thus every private and public key descending from it). This means that extended public keys must be treated more carefully than regular public keys. It is also the reason for the existence of hardened keys, and why they are used for the account level in the tree. This way, a leak of account-specific (or below) private keys never risks compromising the master or other accounts.

I may be wrong but I'm not sure that proposing different HMAC params helps standardization or Bitcoin in general. I suggest BIP-85 for your purposes, expressly to solve the issue of a single secret that is used, in an irreversible way, to populate multiple wallets for multiple purposes. You could have a different application code for each token, each of which is derived from a single master secret.