Atmospheric Link Stabilization Companion
Commercial referenceS20
When contacting BivectorAI, quote S20 — Atmospheric Link Stabilization Companion for this engagement.
A local, air-gapped decision layer that tells your atmospheric HF stabilization program which test and campaign windows are worth running — before you spend the range time, the aircraft hours, or the coordination budget finding out the hard way.
Who this is for
Atmospheric-HF test leads, communications engineering teams, range and campaign planners, infrastructure operators, and program managers who need to decide which redacted-log test windows are worth committing scarce time and budget to before the campaign starts.
The problem you already have
Some windows are expensive because the atmosphere makes the effect hard to observe — not because the system failed
A campaign can burn range time, aircraft hours, transmitter hours and coordination budget on a window whose propagation state is already unfavorable to a clean before/after result. Part 47 treats atmospheric link condition as a state that can degrade or disrupt independently of whether the underlying stabilization hardware is functioning.
“The link was bad” is not specific enough for a defensible campaign decision
The documented atmospheric-assurance model separates normal, degraded and severely impaired propagation states instead of collapsing them into one quality score. That distinction matters when the question is not merely whether a packet arrived, but whether the campaign window can produce evidence strong enough to justify the test cost.
You still need to make the decision without handing an outside vendor your sensitive RF system
The civilian companion keeps the existing redacted-telemetry contract: timing, route, range, frequency, bandwidth, SNR, jitter, packet loss, fade depth, completion time and decode success. Exact site coordinates, raw waveform/IQ, payload content and system-control parameters remain outside the contract.
Real-World Operational Scenario: The Window That Looked Fine Until the Propagation State Turned Against It
Scenario: Two planned campaign windows use similar route and frequency settings, but one repeatedly shows deeper fades, worsening jitter and weaker decode completion.
The engineering team can afford to run only one of the two windows at full campaign cost. Both look plausible on the schedule, and neither contains an obvious hardware fault.
WITHOUT this check: The decision is made from a spreadsheet, calendar availability and engineering intuition. The program can spend its scarce test slot on the window whose propagation evidence is already behaving like a degraded or disrupted link state.
WITH this check: The companion ranks the redacted telemetry windows before resources are committed and preserves the evidence behind the ranking. Part 47's documented atmospheric-assurance states provide the physical interpretation boundary — LINK_ASSURED, LINK_DEGRADED or LINK_DISRUPTED — while the commercial runtime exposes only the customer-safe ranking and evidence needed for campaign planning.
What BivectorAI Provides
Priority Windows Before the Campaign Spend
Ranks the time/route/frequency combinations in the submitted redacted telemetry so the team can concentrate scarce test resources on windows with the strongest measurable evidence opportunity.
Hopeless-Window Flags
Identifies windows whose observed link behavior makes a clean campaign result structurally weak, allowing the team to skip low-value runs instead of discovering the problem after the cost is sunk.
Atmospheric-State Interpretation Without Exposing the Core
The underlying evidence model distinguishes assured, degraded and disrupted atmospheric-link conditions. The customer sees the ranking, evidence and boundary result; protected internal mathematics remain inside the runtime.
Hash-Sealed Campaign Evidence
Every ranked shortlist is accompanied by a reproducible evidence record so program review can trace what telemetry was evaluated and why one window was prioritized over another.
Prove the Gate Before Broader Integration
Do not buy the ranking claim. Back-test it against campaign history.
Start with historical redacted logs whose campaign outcome is already known. Freeze the input contract and ranking rule before revealing the outcome labels. The pilot is useful only if higher-ranked windows consistently correspond to cleaner, more reviewable campaign evidence than lower-ranked or hopeless-window cases.
Commercial expansion should follow measured lift on the customer's own historical campaigns, not a generic promise that every atmospheric link can be predicted from the same conditions.
What you provide
Redacted operational telemetry already described by this service: timing, route, range, frequency, bandwidth, SNR, jitter, packet loss, fade depth, completion time, and decode success.
What BivectorAI actually hands you
Priority-window ranking
The specific time/route/frequency combinations ranked for the current campaign decision.
Hopeless-window evidence
Windows judged structurally weak for producing a measurable result, preserved separately instead of silently disappearing from the shortlist.
Campaign-tier shortlist
A ranked top tier for multi-window campaign planning, allowing planners to compare the best available opportunities rather than make isolated one-window calls.
Hash-sealed review report
A reproducible evidence artifact binding the submitted redacted telemetry to the ranking result for later engineering and program review.
What we will never ask you for
This is the part most vendors get backwards. We built the input contract to make the conversation with your security office short:
Never requested, and structurally rejected if submitted: exact site coordinates, transmitter/pump control parameters, raw waveform or IQ samples, payload or message content, classified or CUI material, or your stabilization system's core design data. The runtime's input validator checks every submitted field against this rejection list before any ranking logic runs — a malformed or over-broad submission is blocked and reported back to you, not silently processed.
How it actually runs
Engagement path
Start with a fixed-scope redacted-log pilot against your own historical campaign data, then expand only if the result improves the decision workflow.
Current commercial tier: S-TECHNICAL. Exclusive field-of-use rights, sublicensing, source-level technical data, IP assignment, unusually broad bulk extraction and dedicated infrastructure are negotiated separately.
Scope a redacted-log pilot
Request Pilot Scope Call →Email fallback: contact@bivectorai.com