Möbby Tokenomics

2 The Möbby Decentralized Ledger Technology (DLT) backbone

At the heart of any decentralized ledger (DL) system is the way in which it achieves consensus on the ledger’s state. The Möbby DLT is analyzed in Sunday Group research team’s technical paper. Here we provide a high-level description that can help our tokenomics analysis. Details can be found in the Möbby technical paper.

The consensus layer of Möbby relies on a novel combination of an iterated Byzantine fault-tolerant consensus mechanism which relies on a paradigm we call Proof-of-Reputation with a Proof-of-Stake secondary/backup blockchain.

  • The PoR blockchain distinguishes parties into reputation nodes (aka Möbby servers) and endorsers (aka Möbby users). Endorsers are regular users of Möbby tokens; in addition they provide to the system their perspective of the trustworthiness of reputation nodes. Reputation nodes are in charge with maintaining and updating the Möbby blockchain—they can be thought of as the Möbby analogue of Bitcoin miners. The primary (PoR) Möbby chain advances in synchronous block-rounds (often referred to as slots in iBFT systems’ specifications). In each round, a new block is added to the PoR chain (with finality) as follows: The user-reported trustworthiness of Möbby servers is used in a novel AI mechanism to extract their global trustworthiness, which is interpreted as their probability of misbehaving from the users’ perspective. We will refer to this as a reputation system.

    The reputation system is used to select, for each slot, a committee CBA that is chosen to propose and validate (vote on) the contents of the next block. (We discuss the size of CBA in different phases of the system below.) The lottery used for this selection is yet another novelty of the Möbby system. It is built in such a way that it guarantees both safety and liveness, while enforcing a fairness property that incentivizes servers (even ones with low reputation) to participate. In particular, it ensures that if the users’ perception of the servers’ trustworthiness is not too far from the ground truth, then the super-majority in CBA is honest which will allow the system to achieve consensus on the state (both safety and liveness). Furthermore, it ensures that every server gets a fair (according to its trustworthiness) chance to participate in the lottery.

    The block-generation protocol proceeds in the following manner: A small sub-committee of CBA (referred to as CBC below) is in charge of choosing the transactions to be included in the current block. Each member of CBC, called a proposer, broadcasts (by means of a byzantine broadcast protocol) their transactions to CBA (only). Each CBA member takes the union of the received transactions into a block and appends his own signature—i.e., validates it—and sends it to the whole Möbby network (servers and users) by means of a standard gossip protocol as in Bitcoin. We will therefore, refer to the nodes in CBA as the validators for the current slot. Note that only one gossip round per slot is needed, which allows the Möbby PoR blockchain to obtain superior communication complexity and speed than competing DLTs (as discussed below). Every node on the Möbby network accepts the block of the current slot as valid if it bears a signature from at least a majority of CBA nodes.

  • The safety and liveness of the PoR chain, as discussed above, depends on the perception of the users on the trustworthiness of the servers being close to the ground truth. The reputation-system extraction AI mechanism includes procedures that aim to correct any misperception by the users’ by observing the server’s behavior. However, naturally, this is susceptible to covert servers (that e.g., operate honestly until they gain the users’ trust and then deviate). To protect the system against such a situation the Möbby DLT employs a backup mechanism: This is a very lightweight proof-of-stake chain, whose role is to polish the primary PoR chain, and promptly detect and correct potential attacks on PoR due to misconception of the users. Importantly, if the perception of the users is accurate, then even if the majority of the stake shifts to an attacker, the system’s liveness and safety are not affected. We refer to the Möbby white paper for details.