Pith. sign in
module module moderate

IndisputableMonolith.Information.ChannelCapacity

show as:
view Lean formalization →

Formalizes discrete memoryless channels, mutual information, and Shannon capacity inside the Recognition Science information layer. Defines capacity as the supremum of mutual information over input distributions, records nonnegativity, symmetry, and elementary bounds, and exposes a ledger-to-capacity bridge. Information theorists working in RS units would cite it for capacity statements. The module is mostly definitions plus short inequality proofs from Mathlib.

claimA discrete memoryless channel $W$, input distributions $p$ on a finite alphabet, mutual information $I(X;Y)$, and channel capacity $C(W)=\sup_p I(X;Y)$. Supporting facts: $I\ge 0$, symmetry in the usual sense, a uniform input distribution, and a map from ledger data to a capacity value, together with a Shannon coding bound sketch.

background

Classical discrete memoryless channel (DMC) theory: a channel is a conditional kernel from a finite input alphabet to a finite output alphabet. An input distribution $p$ induces a joint law on $(X,Y)$; mutual information $I(X;Y)$ measures the reduction in uncertainty about $Y$ given $X$ (equivalently the KL gap from the product of the margins). Capacity is the best rate achievable by coding, realized as $\sup_p I(X;Y)$.

The module sits in the RS Information domain and imports Constants (RS time quantum $\tau_0=1$ tick) and Cost (the $J$-cost calculus). Sibling objects include Channel, InputDistribution, mutualInformation, channelCapacity, uniform input, nonnegativity and symmetry of $I$, a mutual-info bound, Shannon's theorem stub, and capacity_from_ledger.

Local setting: discrete, finite alphabets; no continuous channels or multiuser networks. The ledger bridge is the RS-specific hook that ties abstract capacity back to recognition bookkeeping.

proof idea

Definition-first module. Channel, input distribution, mutual information, and capacity are introduced as data and real-valued functionals. Nonnegativity and symmetry of mutual information are short inequality arguments (KL or entropy identities via Mathlib). Uniform distribution is the normalized counting measure on a finite set; its mass sums to one by a direct sum. Capacity nonnegativity follows from $I\ge 0$. Shannon's theorem and capacity-from-ledger are thin wrappers or interfaces that package the classical coding bound and the RS ledger map; they are not deep analytic developments inside this file.

why it matters in Recognition Science

Gives the Information domain a standard Shannon capacity object so later RS arguments can quote $C(W)$ and $I(X;Y)$ without leaving the monolith. The ledger bridge (capacity_from_ledger) is the distinctive RS landing: it connects recognition bookkeeping to an information-theoretic rate. Downstream use list is currently empty, so the module is foundational rather than a leaf lemma. It does not itself force $\phi$, the eight-tick octave, or $D=3$; those live in the UnifiedForcingChain. Anyone proving rate limits, coding bounds, or ledger-derived capacities in RS units would import this file.

scope and limits

depends on (2)

Lean names referenced from this declaration's body.

declarations in this module (20)