This article explores how resilient Filecoin is to the impact of network computing power.

Original source: Filecoin Network

TLDR

• Due to the anti-fragile mechanism built into the Filecoin protocol design, Filecoin is able to withstand the impact of a large drop in computing power.
• This mechanism provides strong economic incentives for new storage providers (SPs) to join and for existing storage providers to remain in the network after a power outage.
• Specifically, after the massive loss of computing power, the computing power entering the network will receive excess rewards due to the sharp increase in reward concentration.

preface

Filecoin is a distributed data storage network that stores humanity's most important information. SP provides storage capacity to the network during the specified contract period and is compensated by block rewards. To ensure that the contract is fulfilled, FIL tokens need to be used as collateral. The amount of collateral and block rewards received will vary based on the type of computing power added to the network (committed capacity vs. real data).

What does elasticity mean? Merriam-Webster defines resilience as the ability to recover or adjust easily from misfortune or change. Resilience is relevant to many aspects of cryptoeconomics such as Filecoin, such as 51% attacks, Sybil attacks, and Finney attacks. This article will focus on one specific aspect of resilience: the Filecoin network’s resilience to hash power shocks. In Filecoin, computing power refers to the storage and data volume of network services. For various reasons, an SP exiting the network can cause a power shock.

First, let’s ask exactly the resiliency-related question we want answered quantitatively: How will the Filecoin network respond to an outage if the SP experiences the following?

• SP left gracefully and cut off power and off-grid after completing the storage contract.
• SP abruptly terminated the storage contract before the end date, resulting in a shock to the network and a sudden power outage.

modeling

How to model network resilience in a methodical way and ensure predictions follow Filecoin cryptoeconomics? Our approach is to use the Agent Model (ABM) of the Filecoin economy, which consists of an environment and agents that interact with the environment. In this ABM, the environment implements Filecoin's cryptoeconomic mechanisms (such as locking, vesting, minting, and token supply interactions), while the agent is the model of the SP. For more information about Filecoin ABM, please see this article.

To explore resilience, we built agents that model three types of SPs in the Filecoin network:

• Only SPs that load committed capacity (CC) sectors
• Only load the SP of the transaction sector (Fil+)
• Load both the CC sector and the SP of the transaction sector

Each SP can have the following behaviors:

• Dollar-cost averaging behavior - In this mode, agents will plug in and continuously renew their pre-configured hashrate until a termination date, after which the agent will leave the network (either through power expiration or termination).
• Adaptive ROI behavior - In this mode, if the predicted ROI FoFR is greater than the configured threshold, the agent will use the predicted FoFR returns (FIL-on-FIL Returns) to determine the loading computing power.

With these agents and environment specifications, we can configure multiple counterfactual scenarios that represent different network states in which computing power shock scenarios may occur. Network states are defined as the distribution of running SP types in the network, and each network state may lead to different results. Because each state has unique raw bytes and quality-adjusted hashrate, resulting in different minting, staking, and FoFR dynamics. The initial network state we tested was:

• Uniform distribution of CC, FIL+ and hybrid SP
• Uniform distribution of CC and FIL+ storage providers only
• The ratio of FIL+ and CC storage providers is 70/30 (FIL+ skewness)
• The ratio of FIL+ and CC storage providers is 30/70 (CC skew)

Starting from these points, we simulated counterfactual scenarios in which the network's computing power dropped by 30% or 70% from its current value. Consider two situations in which the network's computing power decreases: a) the network's computing power is gradually exhausted, and b) the network's computing power is terminated.

In short, we simulate a 30% or 70% network computing power drop at different network starting points through a gradual sector termination process or an immediate termination event. We simulate each configuration and record network key performance indicators (KPIs).

result

There are some nuances in the case of gradual and sudden power outages, but in both cases of power loss, we observed that network power began to recover after the termination event, due to increased concentration of rewards leading to FoFR Increase.

How it works: After termination, the network computing power drops rapidly, but the standardized rewards for each QA sector have the opposite effect, and the rewards will increase instead. Coupled with the decline in staking, rewards are concentrated, causing FoFR to rise sharply. This combination of network conditions provides higher FIL-on-FIL returns to participants who remain in the network. Our simulation reflects this phenomenon - here, agents (rational actors) observe high FIL-on-FIL returns and exploit the situation to acquire more computing power, allowing the network's computing power to recover.

More details, additional scenarios and discussion can be found in this report.

The above figure shows the KPIs of the Filecoin network in several cases when the SP leaves the network. The vertical dashed line represents the start of the simulation, and the dotted vertical dashed line represents the date when computing power begins to drain from the network. In addition, a baseline case of constant loading using DCAAgent was simulated to provide a basis for comparison (dotted black line).

The figure above shows the KPI of the Filecoin network in several cases where the SP leaves the network. The dashed vertical line indicates the start of the simulation, and the dotted vertical dashed line indicates the date when computing power began to drain from the network. In addition, a baseline situation with DCAAgent simulating constant loading was simulated to provide a basis for comparison (dashed black dotted line).

in conclusion

In summary, we have explored the resilience of the Filecoin network to the impact of network computing power. The conclusion is that the network is resilient due to the centralization of rewards, increasing the revenue of remaining storage providers, and anti-fragile due to the baseline recovery mechanism.

If you would like to learn more about CryptoEconLab and our work, please visit our website or follow us on Twitter.

*This article is for informational purposes only, CryptoEconLab does not provide legal, tax, financial or investment advice, and no party should rely on or expect any related advice. *