Improve token Proof of Work mechanism implementing dynamic complexity

[evilaliv3]

Proposal

This ticket is about to keep track of the discussions and implementation of a dynamic complexity mechanism for the proof of work implementation.

An idea could be to apply the following changes apply thresholds to implement a dynamic complexity of the proof of work for whistleblower sessions varying the complexity from level LOW to level HIGH based on a set of conditions:

• Level should be set to HIGH if more than proof_of_work_reports_per_ip are detected over a certain proof_of_work_observation_period
• Level should be set to HIGH if more than proof_of_work_reports_per_tenant are detected over a certain proof_of_work_observation_period

An example of default configuration could be:

thresholds_proof_of_work_reports_per_ip = 1
thresholds_proof_of_work_reports_per_tenant = 20
thresholds_proof_of_work_requests_per_ip = 100
thresholds_proof_of_work_requests_per_tenant = 4000
thresholds_proof_of_work_complexity_low = 4 (zeroes)
thresholds_proof_of_work_complexity_high = 5 (zeroes)

Motivation and context

In recent times, given maybe the diffusion of the software and a growing interest on the technology we have observed multiple attempts of performing denial of services on setups run by the community.