Definition 1 (Perfect Indistinguishability):
A protocol achieves Perfect Indistinguishability if, to any external observer (e.g., a censor), the transit data stream is statistically or computationally indistinguishable from true white noise (Uniform Randomness). This implies the complete mathematical absence of easily computable invariants in the observed data, regardless of the observer’s knowledge of the protocol’s public parameters or specification.
Definition 2 (Resource Exhaustion Resistance):
A protocol is resistant to Resource Exhaustion strategies if the receiving entity can algorithmically discard invalid data in $O(1)$ time with negligible computational cost, strictly prior to executing computationally expensive operations.
The Invariant Boundary Theorem:
In a permissionless system, it is mathematically impossible to simultaneously guarantee Perfect Indistinguishability of data and resistance to Resource Exhaustion strategies.
- Direct Statement: If a protocol ensures perfect indistinguishability (zero invariants), the system is mathematically vulnerable to resource exhaustion strategies.
- Converse Statement: Any mechanism for the early dropping of junk data (protection against resource exhaustion) requires the presence of an algorithmic or structural invariant in the data, which automatically breaks perfect indistinguishability and renders the protocol vulnerable to censorship.
Proof (Sketch):
-
Part 1 (Perfect Indistinguishability $\rightarrow$ Success of Resource Exhaustion):
Let the protocol possess perfect indistinguishability. Thus, legitimate data $P$ is computationally indistinguishable from random noise $R$. To verify the data, a receiving entity must apply the full deobfuscation function $F^{-1}(P)$ which may include computationally heavy operations. Because an attacker generates ideal white noise $R$ with zero CPU cost, and the receiving entity lacks any metadata markers for early filtering, the entity is forced to apply the computationally heavy function $F^{-1}(R)$ to the entire incoming stream. A fatal asymmetry in computational costs arises, guaranteeing the success of a resource exhaustion strategy. -
Part 2 (Resource Exhaustion Resistance $\rightarrow$ Vulnerability to Censorship):
Let the protocol be resistant to resource exhaustion. Thus, there exists a lightweight filter function $f(x)$ that distinguishes valid data from white noise in $O(1)$ time prior to core protocol execution. According to Kerckhoffs’s principle, the protocol specification is public; therefore, the function $f(x)$ is known to the censor. The censor systematically applies $f(x)$ to transit data and selectively drops any data for which $f(x) = \text{True}$. The property of perfect indistinguishability is irreversibly broken, and the protocol is successfully censored. $\blacksquare$