Router CSI Lab
EN ZH

Methodology and evidence grading: how every claim on this site is verified

The most common failure in technical writing is not a wrong number — it is treating hearsay as a conclusion. Our rule: every external claim carries a source and an evidence grade, and everything "we say" is turned into reproducible first-party evidence.

1. Evidence grades (A-E)

Grade Meaning Typical source How we use it
A Standard / systematic review IEEE standard project pages, highly cited surveys (e.g. the CSI sensing survey [1]) background and consensus, never alone for a contested claim
B Peer-reviewed paper INFOCOM / ICDCS / IMWUT and similar supports method-level claims (CSI respiration, fall detection)
C Official documentation / vendor material platform limits, fees, eligibility pages supports "how the platform is defined" facts
D Preprint / encyclopedia / community arXiv, Wikipedia lead only; must be labelled as not peer-reviewed
E First-party reproducible measurement our captures, scripts, SHA-256 values the primary evidence for engineering claims (below)

The grading is inspired by the evidence hierarchy used in evidence-based medicine: bind the strength of the claim to the strength of the evidence instead of writing everything with the same certainty.

As of the last update of this page, the reference registry holds 34 verified sources (A 2 / B 9 / C 18 / D 2 / E 3), each with the verification method and date recorded. An article may only cite entries from that registry; claims we cannot verify are not published.

2. Verification workflow (before publishing)

  1. every external claim must be traceable (DOI, standard page, official doc);
  2. DOI metadata is checked with the CrossRef API — title, authors, year and citation counts for every paper-level citation;
  3. official docs are quoted with the key sentence and the verification date;
  4. preprints are labelled as such and never written as settled fact;
  5. first-party measurements must ship data + scripts + checksums, otherwise they are downgraded to "unverified";
  6. if strong evidence is missing, we either drop the claim or mark it explicitly as speculation.

3. First-party evidence (grade E)

Evidence Content Verification
10-minute CSI capture 57,293 records / 600.0 s / 256 subcarriers reassembled chunks are byte-identical to the source (SHA-256) [3]
10-second sample 312 records to validate your parser site download, field-by-field format description
Capture/analysis scripts capture, parse, heatmap, watchdog, bandwidth switch MIT, runnable on real hardware [4]
Key performance findings firmware ATF pinning, 40 MHz client in an 80 MHz BSS controlled A/B (one variable, 6 samples each) with raw data published

4. Three things we do not do

  1. Never present a preprint as a conclusion (e.g. "Wi-Fi can reconstruct body pose" must be labelled controlled-setting + preprint);
  2. No medical claims: respiration/heart-rate/fall pages always state "not a medical device, not for diagnosis";
  3. No hidden failures: measurements we cannot reproduce or explain are written as such (e.g. our unresolved 2.4 GHz anomaly).

5. Corrections policy

6. Scope statement

Content here is for technical research and learning. Human-sensing material is not medical, safety or legal advice. Do not use it for non-consensual surveillance.

References and data sources

Every external claim carries a source and an evidence grade (grading scheme). Grade E marks first-party, reproducible measurements from this project.

  1. Grade A · standard/review WiFi Sensing with Channel State Information: A Survey
    W. Ma, W. Zhou, X. Wang · ACM Computing Surveys · 2019
    doi:10.1145/3310194 · citations 857 · verified: CrossRef API, 2026-10-11 · The field's most-cited CSI sensing survey (857 citations, verified via CrossRef).
  2. Grade A · standard/review IEEE 802.11bf — WLAN Sensing (standard project)
    IEEE Standards Association · IEEE 802.11 Working Group · 2020
    standards.ieee.org/ieee/802.11bf/10331/ · verified: standards.ieee.org HTTP 200, 2026-10-11 · Official project page for turning sensing into a WLAN standard capability (approved September 2020).
  3. Grade E · first-party measurement This project's 10-minute raw capture (57,293 records / 600.0 s, SHA-256 00373f83…95e2)
    Router CSI Lab · 2026
    csi.okwujun.pw/zh/datasets/ · verified: Reassembled chunks verified byte-identical (2026-10-11)
  4. Grade E · first-party measurement Router CSI Toolkit — capture/analysis/watchdog scripts (MIT)
    Router CSI Lab · 2026
    csi.okwujun.pw/zh/tools/ · verified: Repository bundle has a SHA-256; scripts run on real hardware