feat(release): v0.9.2 - Positioning Upgrade

[imzzaidd]

v0.9.2 -- Positioning Upgrade

Summary

Upgrade the positioning subsystem with maximum likelihood estimation (MLE), Kalman filter tracking, Cramer-Rao lower bound (CRLB) accuracy metrics, zero-knowledge proof of position, and RTT-based distance bounding. This release transforms triangulation from a basic weighted-centroid approach into a statistically rigorous positioning engine suitable for presence verification in adversarial environments.

Planned Changes

Maximum Likelihood Estimation (MLE)

• Replace weighted centroid with MLE-based position estimation
• Gradient descent solver with convergence guarantees
• Configurable iteration limits and tolerance thresholds
• Bounded arithmetic throughout (saturating operations)

Kalman Filter Tracking

• State prediction and measurement update for position tracking
• Configurable process noise and measurement noise matrices
• Innovation-based outlier rejection
• Smooth position trajectory for moving actors

Cramer-Rao Lower Bound (CRLB)

• Theoretical accuracy bound computation from anchor geometry
• Fisher information matrix assembly from RSSI/RTT measurements
• Confidence ellipse parameters for position uncertainty
• Automatic quality assessment (reject estimates below CRLB threshold)

ZK Proof of Position

• Zero-knowledge proof that an actor is within a specified region
• Circuit: prove position within region with consistent measurements
• Integration with pallet-zk for on-chain verification
• Privacy-preserving presence verification (prove location without revealing it)

RTT Distance Bounding

• Round-trip-time based distance measurement
• Speed-of-light upper bound enforcement
• Relay attack detection via timing analysis
• Integration with device-scanner for RTT collection

Deliverables

• MLE position estimator with gradient descent solver
• Kalman filter for position state tracking
• CRLB accuracy metric computation
• ZK proof of position circuit and verifier
• RTT distance bounding protocol
• Triangulation pallet upgrade to MLE backend
• Device-scanner RTT collection integration
• pallet-zk circuit for position proofs
• Statistical accuracy benchmarks vs. weighted centroid
• Adversarial positioning test suite

Invariants Affected

• INV73-75: ZK proof system extended with position proof type
• INV64-65: Device positioning enhanced with RTT bounding
• INV1-13: Presence verification strengthened by ZK position proofs
• INV43: Chain binding extended to position proof statements

Breaking Changes

• pallet-triangulation API overhaul (weighted centroid to MLE)
• PositionEstimate type gains uncertainty/confidence fields
• New PositionProof variant in pallet-zk proof types
• Device-scanner inherent format extended with RTT data

Testing

• cargo test --workspace passes
• cargo clippy --all-targets -- -D warnings clean
• cargo fmt --all -- --check passes
• no_std verification for all crates
• MLE convergence tests with known anchor configurations
• Kalman filter tracking accuracy with synthetic trajectories
• CRLB computation validated against analytical solutions
• ZK position proof soundness and completeness tests
• RTT bounding correctness with simulated relay attacks
• Regression tests: MLE accuracy >= weighted centroid in all cases

Milestone

v0.9.2 - Positioning Upgrade (due: 2026-08-31)

Checklist

• cargo fmt
• cargo clippy
• cargo test
• no_std check
• Documentation updated
• CHANGELOG.md updated
• Triangulation architecture doc updated
• ZK circuit specification documented