Package Overview

bijux-gnss-receiver owns receiver runtime execution: configuration becomes a running receiver, sample frames enter through ports, stages produce evidence, and runtime artifacts leave the crate before repository persistence takes over.

This crate is large because the receiver boundary is large. Acquisition, tracking, observations, optional navigation handoff, runtime sinks, simulation, and receiver-side validation must agree on one execution story.

Runtime Chain

flowchart LR
    config["receiver config"]
    ports["source, clock,<br/>logger, trace, metrics"]
    acquisition["acquisition"]
    tracking["tracking"]
    observations["observations"]
    navigation["optional navigation"]
    artifacts["run artifacts<br/>validation reports"]

    config --> ports --> acquisition --> tracking --> observations --> navigation --> artifacts
    observations --> artifacts

Owned Families

family	owns	first proof
engine	runtime config, validation, defaults, logging, metrics, support matrix, receiver composition	engine source
ports and I/O	sample sources, artifact sinks, clocks, runtime effect seams	port and I/O source
acquisition	request planning, search windows, candidates, ranking, explainability, handoff evidence	acquisition stage source
tracking	channel lifecycle, lock evidence, loop state, reacquisition, sample-rate diagnostics	tracking stage source
observations	pseudorange, carrier phase, smoothing, residuals, quality, epoch manifests	observation stage source
navigation handoff	receiver-owned calls into navigation solving and filtering when enabled	navigation handoff source
simulation	synthetic sources, truth, stage accuracy, scenario validation	simulation source
validation reports	receiver-side accuracy, consistency, covariance, and reference checks	validation-report source

Reader Rules

• Use this handbook for execution order, state handoff, runtime diagnostics, and in-memory receiver artifacts.
• Leave for bijux-gnss-signal when the question is reusable code generation, signal catalog truth, sample math, front-end filtering, NCO behavior, or DSP discriminators outside one receiver run.
• Leave for bijux-gnss-nav when the question is navigation science rather than receiver-side invocation of that science.
• Leave for bijux-gnss-infra when runtime artifacts are named, persisted, indexed, compared across run directories, or tied to dataset registry state.
• Leave for bijux-gnss-core when the question is shared record meaning, observation fields, diagnostic codes, units, or artifact envelopes.

Failure Reading

When a receiver test fails, first identify which stage emitted the wrong evidence. Do not flatten every receiver failure into “pipeline”:

• acquisition failures usually involve signal model selection, code phase, Doppler search, candidate ranking, ambiguity, or explainability artifacts
• tracking failures usually involve lock state, loop bandwidth, Doppler ramp, reacquisition, fade handling, or channel diagnostics
• observation failures usually involve timestamping, pseudorange, carrier phase, smoothing, residuals, or quality classification
• navigation failures at this boundary usually involve receiver-to-nav handoff or filter configuration, not standalone navigation model law

First Proof Check

Inspect the receiver crate README, pipeline guide, runtime guide, and artifact guide. Then inspect the relevant pipeline stage and the integration tests named in the crate README.