Path 07 · Module 07 · Learn synthesis

Coexistence &
Technology Selection

Which technology—or combination—fits the same product once its radios operate together? Follow a weak wanted signal through a crowded gateway, then defend a connectivity portfolio with explicit constraints, uncertainty and one useful experiment.

01 / 10

Certified radios fail when used together

The two radios passed their individual checks. Why does the gateway lose its weakest node report whenever Wi-Fi transmits?

The fictional condition-monitoring gateway puts a local receiver, Wi-Fi backhaul, a cellular option and GNSS timing inside one enclosure. Each radio can work alone while the product fails in a shared state: one transmitter is active, the wanted signal is weak, and the victim’s gain or filtering is configured differently from its individual test. The missing item is an operating condition with a coupling path, not another logo on the comparison sheet.

Think about itWhich condition would you add before replacing the weak receiver?
Answer

Pin the aggressor’s waveform and on-time output, the wanted level, the victim mode and decision criterion, and the concurrent state. Measure the path and compare the same receiver boundary. A failure synchronized with a transmit burst could be RF blocking, receiver recovery or supply droop; timing alone does not identify the mechanism.

A specification defines scoped behavior. A qualification record concerns a particular implementation and program. A concurrent test observes a selected combination under stated conditions. A national requirement has a different authority and applicability. For example, Bluetooth’s receiver blocking text specifies test conditions; it cannot be replaced by an arbitrary blocker number. The −35 dBm limit used later is explicitly synthetic. [BT-62] [BT-QUAL]

Common misconceptionCertification of each radio proves concurrent operation.

Individual scope does not establish the shared antenna, supply, timing, enclosure and host configuration. Preserve qualification evidence, then add the missing concurrent product tests and market questions.

Build on the existing evidence chain

Review 07.1’s source discipline, 05.4’s blocker conventions, and 05.7’s architecture review. Antenna integration in 06.3 and channel distributions in 06.5 remain part of the product case. The prior generic 2.450 GHz QPSK waveform is a teaching baseline, not a Bluetooth, Wi-Fi, LoRa or cellular mode.

Your first review entry is a failure statement with a named shared state. The next question is whether all candidate architectures are even solving the same service problem.

02 / 10

Freeze one common service basis

Would you compare a fast radio event, a low average current and a long-range headline as if they were three answers to one requirement?

Freeze the service boundary before collecting attractive numbers. D0 is the application payload; D1 is framed/coded data. R1 is a component RF port, R2 the antenna-feed plane, S0 a spatial/OTA quantity and R3 the receiver decision boundary. R3 is not an antenna input. A successful R3 decode is only one step toward a timely D0 delivery.

p07-m07-common-basis-v1 · chosen product basis, not a service guarantee
Common itemFrozen requirement or explicit unknown
Traffic / fleet32 D0 bytes (256 bits) per node every 600 s; 1000 nodes; include failed generated reports.
Delivery / endpointAt least .99 unique complete reports by 2 s, from node application release to nominated gateway/service application delivery.
Energy / boundaryAt most 200 mJ per generated report at the node radio supply, including discovery, search, receive, retries and residual sleep. Whole-node and gateway energy require separate budgets.
Gateway roleAt least 20 Mbit/s D0 backhaul service when architecture requires that role. A direct terminal-to-host architecture records why local gateway service is not applicable.
Installation / populationFixed nodes; maintenance repositioning. Metal-enclosure and open-pole strata with declared orientations. Population weights, channel distributions and rare-event exposure unknown.
Infrastructure / lifecycleOwned local gateway versus operator service must be explicit; installation, power, maintenance, provisioning and replacement ownership unknown. No lifetime or battery-capacity promise.
Markets / mobilityCH and EU/EEA candidates; actual deployment countries, class and network/operator availability unresolved. No universal range or roaming claim.
Ereport=0TV(t)Iradio(t)dtq2s=Nunique,D02sNgenerated\begin{aligned}E_{\mathrm{report}}&=\int_0^T V(t)I_{\mathrm{radio}}(t)\,\mathrm dt\\q_{2\mathrm s}&=\frac{N_{\text{unique,D0}\le2\mathrm s}}{N_{\mathrm{generated}}}\end{aligned}T=600 s is the report-generation period. Energy is joules at the declared radio supply, including failed periods. q₂s is the observed or assumed fraction of unique complete reports delivered within the application deadline over a declared population; it is not link-only packet success.
Think about itDoes changing the report period from 60 s to 600 s leave energy per report unchanged?
Answer

Only if the omitted 540 s consumes no energy. At 3 V and 2 µA residual sleep, it adds 3.24 mJ. The unchanged synthetic event then gives 3.8709038808 mJ per generated report. That calculation still says nothing about actual delivery within 2 s.

Reconciliation ledger · changed assumptions create new variants
Origin / preserved IDsOriginal basisCommon-basis transformationEvidence / remaining gap
07.1 / P07-M01-LE-EXT-v1Extended advertising document trace; not a connectionSelect explicit connection variant; preserve original IDBT-62 / BT-QUAL; actual firmware/profile unknown.
07.2 / P07-M02-LE-CONN-v1; THREAD-SED-v160 s event; 32 D0 B; synthetic currents at 3 V600 s variant: 0.6309038808 → 3.8709038808 mJ/generated report; sleep adds 3.24 mJRECON: same event model, no measured current or ≤2 s delivery. Thread SED remains an alternative, not silently relabelled LE.
07.3 / P07-M03-HE-BACKHAUL-v1HE SU MCS 6: 77.426470588 Mbit/s PHY45.294485294 Mbit/s proxy before availability; 25.364911765 at a=.56IEEE-AX / RECON; R1 blocker and noise definitions match; D0 deadline unknown.
07.4 / P07-M04-REMOTE-FLEET-v160 s report / 30 s urgent-command origin600 s period; SF7 airtime .092416 s, offered G=.0513422222, e^(−2G)=.902411690LORA-L2 / RECON: no unique-delivery, capture or queue/deadline proof. SF12 airtime 2.138112 s already exceeds 2 s.
07.5 / P07-M05-CELLULAR-OPTIONS-v1Registered and search states, distinct from outage average600 s: registered 111.55968 mJ; 30 s search 4431.01968 mJ; assumed 10% search mixture 543.50568 mJRECON / 3GPP; unknown actual search fraction. A registered snapshot cannot supply B’s 120 mJ acceptance input.
07.6 / P07-M06-GNSS-TIMING-v1; SBD-FALLBACK-v1±100 ns / 99% / 1 s timing epochs; messaging 900 sTiming retains its own service role; 900 s fallback is incomparable to 2 s reporting until testedGPS-SPS / SBD; DOP and propagation delay exclude bias, terminal queues and host service.
05/06 / p06-evidence-map-v1; M01-A; M01-B-LOSS; M01-INSTALL-UNKNOWNComponent, feed and spatial records; installation unknownRetain named R1↔R2 transformations and measured S0 population; no invented rangeRECON: do not count feeder loss in both total isolation and a second insertion-loss term.
Common misconceptionUnknown means zero.

Missing search energy is not zero energy; an unmeasured channel is not a clear channel. Keep unknown as a separate state. A genuine architectural exemption needs a reason, as with D’s absent local-gateway role.

Record the matched units, planes and statistic before any comparison. The complete source-card snapshots in the final review retain inherited IDs; none imports the later A–D acceptance inputs as product measurements.

03 / 10

Map every victim, aggressor and path

Which transmitter, path and receiver state explain this failure—and which remain unmeasured?

A victim–aggressor matrix is a set of physical hypotheses. Give each path an ID, including parallel paths from the same source. T01 carries a synthetic on-time +20 dBm Wi-Fi aggressor at 5.180 GHz, a 2.450 GHz teaching victim with 200 kHz equivalent noise bandwidth, and a weak −105 dBm wanted stimulus at R1-RX. Its +2.730 GHz carrier offset and integrated factor are disclosed assumptions, not a real spectral mask or a protocol receiver sensitivity.

p07-m07-threats-v1 · every listed path retains its own evidence
Path / victim ← aggressorMechanism / controlRequired state / evidence
T01 · A-WIFI-NODE: Weak-node gateway RX ← Internal gateway Wi-Fi TXBlocking + noise-like leakage; internal coordinatedWi-Fi TX and victim RX; editable 100 ms half-open schedule. Illustrative V82: +20 dBm, total I=55 dB nominal; no measured suppression or receiver threshold.
T02 · EXT-WIFI-NODE: Node/gateway receive ← External Wi-Fi networkCochannel collision / adjacent leakage; externalRequired state to establish; no measured duty or overlap. Unknown; internal PTA has no authority over the external transmitter.
T03 · CELL-GNSS: GNSS timing receiver ← Internal cellular TXBlocking / harmonics; internal uncoordinatedRequired state to establish; no measured duty or overlap. Unknown; no harmonic magnitude follows from a carrier name.
T04 · EXT-GNSS: GNSS receiver ← External interferenceNoise / blocking; externalRequired state to establish; no measured duty or overlap. Unknown; satellite visibility does not establish interference immunity.
T05 · SUPPLY-RX: Receiver supply ← DC converter / radio loadSupply coupling; internal uncoordinatedRequired state to establish; no measured duty or overlap. Unknown; total antenna isolation cannot characterize supply transfer.
T06 · CLOCK-RX: Receiver / antenna ← Digital clock / cableDigital / common-mode coupling; internal uncoordinatedRequired state to establish; no measured duty or overlap. Unknown; no spectral mask generated from the technology name.
T07 · LO-MIX: Victim mixer ← Strong off-channel signal + LO noiseReciprocal mixing; internal uncoordinatedRequired state to establish; no measured duty or overlap. Unknown; V82 excludes reciprocal mixing.
T08 · RECOVERY: Weak-node receiver ← Internal TX turn-offReceiver recovery / filter settling; internal coordinatedRequired state to establish; no measured duty or overlap. Unknown in product; 2 ms is only a chosen guard allocation.
T09 · TWO-TONE: Victim nonlinear stage ← Two simultaneous transmittersIntermodulation; internal uncoordinatedRequired state to establish; no measured duty or overlap. Unknown; no IM3 power inferred.
T10 · FEED-PATH: Victim feed ← Wi-Fi TX feedCoherent alternate path; internal coordinatedRequired state to establish; no measured duty or overlap. Unknown; may already be included in T01 total isolation. Never add its power independently.
T11 · SAT-SUPPLY: Gateway or node rail ← Satellite TX / startupSupply droop / reset; internal uncoordinatedRequired state to establish; no measured duty or overlap. Unknown; RF watts are not DC energy.
T12 · FILTER-GAP: Victim LNA ← Out-of-band networkMissing filter / blocking; externalRequired state to establish; no measured duty or overlap. Unknown; absence of filter data is not absence of threat.
Inspect every path’s configuration, planes and next observation
  1. T01 · A-WIFI-NODE

    5.180 GHz, HE-style SU MCS 6, 20 MHz; synthetic 2.450 GHz victim, 200 kHz ENBW; +2.730 GHz offset; mounting A, 25 °C; on-time total-power / integrated-noise detectors.

    Planes: R1-TX → R1-RX. Wanted: −105 dBm at R1-RX, chosen weak wanted stimulus only; not a sensitivity limit.

    State: Wi-Fi TX and victim RX; editable 100 ms half-open schedule.

    Next evidence: RF owner: coupling, selectivity, blocker and recovery evidence.

  2. T02 · EXT-WIFI-NODE

    2.4 GHz external channel/width/traffic unknown; mounting A; packet detector and observation population unknown.

    Planes: S0 → R2-RX → R1-RX → R3. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: Field owner: time-aligned external occupancy and unique D0 delivery.

  3. T03 · CELL-GNSS

    Selected LTE-M band/resource/offset and GNSS channel unresolved; mounted receiver and integrated bandwidth unknown.

    Planes: R1-TX → R1-RX. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: RF/timing owners: concurrent RF and calibrated PPS logs.

  4. T04 · EXT-GNSS

    GPS L1 C/A candidate; interferer waveform/offset, receiver bandwidth and wanted C/N₀ population unknown.

    Planes: S0 → R1-RX → R3. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: Timing owner: independent reference and receiver-input characterization.

  5. T05 · SUPPLY-RX

    Switching state/harmonic and load step unknown; mounting A; voltage spectrum/settling detector required.

    Planes: Supply → receiver; not R1 RF power. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: Power owner: probe supply with RF path terminated and controlled load.

  6. T06 · CLOCK-RX

    Clock modes, harmonics, cable layout and receiver band unknown.

    Planes: Clock/cable → R2/R1. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: Digital/RF owners: toggle clock mode and compare conducted/near-field signatures.

  7. T07 · LO-MIX

    LO noise at blocker offset, receiver bandwidth and filter state unknown; input-referred integrated noise.

    Planes: R1-RX → mixer → R3. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: Receiver owner: independent LO-noise and blocker-level tests.

  8. T08 · RECOVERY

    Actual TX ramp, AGC/filter state and recovery-to-decode statistic unknown; mounting A.

    Planes: R1-RX → R3 timing. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: Firmware/RF owners: request/grant/RF/decision timestamp correlation.

  9. T09 · TWO-TONE

    f₁/f₂ and 2f₁−f₂ candidate only; amplitudes, IIP3, LO/filter and load evidence missing.

    Planes: Named nonlinear stage → R3. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: RF owner: two-tone or actual modulated test with source filtering.

  10. T10 · FEED-PATH

    Same transmitter as T01; phase and complex transfer unknown.

    Planes: R1-TX → R2 → R1-RX. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: Antenna owner: complex transfer with a nonoverlapping loss ledger.

  11. T11 · SAT-SUPPLY

    Terminal mode, peak current, supply impedance and startup state unknown.

    Planes: Radio supply → digital state. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: Power/service owners: total terminal voltage/current/state trace.

  12. T12 · FILTER-GAP

    External band and waveform, filter stopband, wanted insertion loss, mounting and detector unknown.

    Planes: S0 → R2-RX → LNA R1. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: Receiver/field owners: pin external conditions and front-end transfer.

Think about itIf you cannot see an external interferer during a short bench run, can you delete T02?
Answer

No. State the observation interval and detector, then retain the missing field population. An absent threat needs defensible scope evidence. An unobserved or unmeasured threat remains unknown.

For an RF path, transform between R2 and R1 using the actual feed/filter network. If total isolation already includes an antenna route, adding that route’s power again double-counts it. For supply and digital paths, record voltage, current or complex transfer where appropriate; an antenna isolation number cannot stand in for those measurements.

Go deeperWhy two paths from one transmitter need more than two dBm numbers

With the portfolio’s e^(+jωt) convention, complex path transfers add before taking magnitude squared. Two unknown phases can reinforce or cancel. T10 may already be part of T01’s total measured isolation. Only genuinely uncorrelated noise-like contributors with compatible bandwidth and plane can use the independent-power sum below.

The artifact now has named paths and missing observations. Diagnose each path before choosing a mitigation.

04 / 10

Identify the interference mechanism

Does a packet collision have the same signature as a strong off-channel blocker?

Collision concerns overlapping useful transmissions and their decode conditions. Leakage adds unwanted energy in the victim’s admitted band. Blocking describes degraded receiver operation under a strong signal that may lie outside that band. Different center frequencies alone exclude none of these mechanisms. A receiver mixer can also translate oscillator noise around a blocker into the wanted band: reciprocal mixing needs LO-noise and filter evidence that this simple leakage factor does not contain. [ADI]

Mechanism → discriminating evidence → bounded consequence
MechanismWhat distinguishes itAvoid this inference
Cochannel collisionTime-aligned packets, wanted/interferer levels, capture and unique decode statisticsOverlap fraction is not packet error probability.
Integrated leakage / selectivityCompatible PSD/integrated power, filter state, equivalent noise bandwidth and input-referred noiseDo not apply selectivity twice.
Out-of-band blockingPinned on-time blocker, wanted stimulus, gain state and receiver decision criterionAverage duty does not reduce on-time coupled level.
Nonlinear mixing / harmonicsTwo-tone or actual waveform tests, source purity, nonlinear stage and filter/LO evidence2f₁−f₂ locates a possible product, not its power.
Reciprocal mixingLO noise at the blocker offset, admitted bandwidth and strong-signal levelOne total isolation number cannot predict the added noise.
Digital / supply / recoveryControlled clock/load/RF states; supply trace and recovery-to-decode timestampsAn RF attenuator may not fix a supply or timing path.
Pcoupled=PTXIH=PlimitPcoupledP_{\mathrm{coupled}} = P_{\mathrm{TX}} - I\qquad H = P_{\mathrm{limit}} - P_{\mathrm{coupled}}P_TX is total on-time dBm at R1-TX. I is total dB isolation to R1-RX, including each declared loss once. P_limit=−35 dBm is an illustrative maximum blocker. H in dB is positive when coupled power is below that maximum.
Checked 07.3-compatible on-time anchor · V82 / mounting A / 25 °C
Quantity / planeI=45 dBI=55 dB
Total P_TX / R1-TX+20 dBm+20 dBm
Coupled blocker / R1-RX−25 dBm−35 dBm
Blocker headroom H−10 dB: fails0 dB: inclusive boundary passes
Required blocker isolation20−(−35)=55 dB55 dB
Wanted stimulus / victim detector−105 dBm; 200 kHz ENBW; R3 criterion must be pinned for a real testSame synthetic receiver mode
Think about itWould a 20% transmitter duty cycle repair the −10 dB headroom?
Answer

No. It reduces period-average power, but the receiver still sees −25 dBm during transmission. Do not subtract 10log₁₀(.2) from the blocker. Scheduling can avoid a required overlap only when the internal control and recovery contract are supported.

L=PTXIKΔN=10log10(1+10LN10)L = P_{\mathrm{TX}} - I - K\qquad \Delta N = 10\log _{10}(1 + 10^{\frac{L-N}{10}})L is admitted noise-like leakage in dBm at R1-RX. K=82 dB is a chosen combined spectral/selectivity factor integrated over the victim’s 200 kHz ENBW. N=−110 dBm is input-referred receiver noise in that bandwidth. ΔN is noise rise in dB; maximum 1 dB is a separate illustrative criterion.

At I=45 dB, L=−107 dBm and ΔN=4.764348624365 dB. Inverting the 1 dB limit gives L_allowed=N+10log₁₀(10^(.1)−1)=−115.868253243801 dBm, hence required isolation 53.868253243801 dB for leakage. The blocker needs 55 dB, so it binds first for V82. V72, a deliberately leakier synthetic receiver, can make the noise screen bind instead. [RECON]

Psum,dBm=10log10(i10(Pi,dBm30)/10)+30P_{\mathrm{sum,dBm}}=10\log_{10}\left(\sum_i10^{(P_{i,\mathrm{dBm}}-30)/10}\right)+30Convert each compatible, uncorrelated noise-like contribution to watts before summing. Two −117 dBm contributors produce −113.98970004336 dBm at the same input plane and bandwidth. Coherent paths with unknown phase remain unresolved.
Common misconceptionDifferent channels do not interfere.

Channel separation may reduce admitted energy, but blockers, LO noise, nonlinear products and circuit paths remain. A filter can reduce an aggressor while its wanted insertion loss worsens sensitivity. Record both effects in the same plane ledger.

Add the RF calculation, its mode and its limitation to the review. Then choose a mitigation within the part of the system you actually control.

05 / 10

Mitigate within the controllable boundary

Can moving a transmit burst fix the problem without moving it into a service deadline?

Frequency planning, filtering, antenna placement, isolation, linearity, reduced power and time coordination change different parts of the failure. Each spends something: spectrum or deployment flexibility, wanted insertion loss, antenna volume, DC energy, coverage reserve or service time. Internal request/grant coordination also depends on real radio and recovery timing. A legacy manufacturer guide explains that mechanism but is explicitly deprecated; it supplies no current API or proof for this lesson’s guard. [SILABS]

Overlap = measure(union(TX) ∩ union(RX))Use a half-open 100 ms cycle: include each start, exclude each end. Union overlapping entries before intersecting the roles. Split an explicit wrap-around interval at the cycle boundary. Equal endpoints mean empty, never a full cycle.
Checked schedules · deterministic endpoints in ms
CaseTX / RXIntersection / interpretation
Original[0,20) / [10,50)[10,20): 10 ms; .10 of cycle; .25 of victim listening
Adjacent[0,20) / [20,50)Empty; adjacent half-open boundaries contribute no duration
Wrapping TX[90,10):wrap / [0,20)TX splits [90,100)+[0,10); 10 ms intersection
Protected[0,20) / [22,62)No TX in [20,64): 2 ms guard each side. Zero declared overlap; recovery/control still needs evidence.
Remaining allocation100−44=56 msAssume all remaining time allocated to Wi-Fi: a=.56, applied once.

A separate independent random-duty orientation gives joint wall-time overlap .2×.4=.08, but P(TX | RX)=.2. Those are probabilities under an independence assumption, not the deterministic .10/.25 above, and none is a packet error probability. Correlated traffic or retransmission invalidates that independence assumption.

Think about itWill the protected schedule silence an external Wi-Fi network?
Answer

No. The gateway cannot grant or deny another network’s transmissions. T02 remains required and unknown when the product threat scope includes it. T03–T12 likewise retain separate evidence; clearing one internal overlap does not erase supply, clock, recovery, nonlinear or external paths.

Common misconceptionTime coordination controls external networks.

Record internal coordinated, internal uncoordinated and external authority separately. A zero-overlap picture can justify a proposed internal schedule; it cannot become a measured coexistence pass.

For the artifact, write the permitted, prohibited and unknown states and the cost of each mitigation. If a quiet window removes capacity or delays a report, carry that loss into the next service ledger.

06 / 10

Compare complete service and energy scenarios

What did the radio do during the other 599 seconds?

Energy belongs to a complete state timeline. 07.2’s local-radio model has expected attempts 1.05, event charge .0000903384 C and active duration .0185532 s. At 3 V, adding sleep over the residual 600 s gives 3.8709038808 mJ. Its eventual link-success proxy .9975 is not a deadline result. Discovery, application queues and failed generated reports still need matched evidence.

Elocal=3[0.0000903384+2×106(6000.0185532)]JE_{\mathrm{local}} = 3[0.0000903384 + 2\times 10^{-6}(600-0.0185532)] JSynthetic event and residual sleep at one radio-supply boundary. Moving from 60 to 600 s adds 3.24 mJ. A different retry, discovery or temperature state requires a named new variant.

07.3’s gateway uses 234 data tones × 6 bits × 3/4 coding per 13.6 µs symbol: 77.4264705882353 Mbit/s PHY. With chosen efficiency .65 and bounded independent retry model p=.1, at most three attempts, delivered packets per attempt is .9. The proxy is 45.29448529411765 Mbit/s before availability and 25.36491176470588 Mbit/s at a=.56. That clears the chosen 20 Mbit/s capacity screen, but neither a mean service rate nor spare capacity proves each report’s two-second deadline. [RECON]

07.4’s LPWAN exposes a different bottleneck. For 45 framed bytes at SF7, 125 kHz and coding 4/5, airtime is .092416 s. A 1000-node, 600 s, three-channel pure-ALOHA orientation gives G=.0513422222222 and e^(−2G)=.9024116895793. It assumes independent Poisson attempts, equal durations and no capture; it omits real receive windows, queues and interference populations. At SF12 the 2.138112 s airtime alone is longer than the 2 s application deadline. Class A downlink opportunities also depend on uplinks. [LORA-L2]

07.5’s cellular state ledger gives 111.55968 mJ in its illustrative registered 600 s case, but 4431.01968 mJ with 30 s search. A deliberately assumed search fraction s=.1 yields (1−s)E_registered+sE_search=543.50568 mJ. This fails 200 mJ; the actual fraction is unknown. An abort case, 4337.45928 mJ, belongs in the failed-report population. RF output watts are not DC supply energy, and a modem version is not an operator service commitment.

07.6’s GNSS timing and satellite messaging serve separate roles. DOP maps specified ranging uncertainty through geometry; user biases and receiver conditions remain. A short propagation delay omits acquisition, queues, retries and host delivery. The inherited 900 s SBD fallback has no demonstrated two-second service. It may be useful in a hybrid portfolio under its own requirement. [GPS-SPS] [SBD]

Go deeperWhy a coverage distribution belongs beside the current trace

Weak installed channels change airtime, retries, adaptation, searching and time spent awake. A single typical link margin cannot populate those states. Keep placement/orientation strata, wanted signal distribution, infrastructure availability and uncertain correlations explicit. Wider uncertainty cannot be repaired by copying a nominal antenna gain or subtracting the same feed loss twice.

The reconciled physics explains what is still missing. The next acceptance exercise deliberately supplies hypothetical complete values so that hard screening can be learned exactly, without claiming these earlier calculations proved them.

07 / 10

Keep ecosystem and market gates visible

If a candidate passes the RF fixture, what still prevents a real release decision?

Source identity, claim evidence, applicability and freshness are separate. An adopted specification is not proof that the selected product implements it. A superseded amendment can still describe a selected legacy mode. A review-overdue record calls for rechecking, not deleting the physics. Use the frozen 2026-09-08 UTC evaluation and 2026-12-07 due date for reproducible examples; browser time never changes these results. [BT-INDEX] [IEEE-INDEX] [3GPP]

Independent release questions · all remain open for the fictional product
Gate / ownerEvidence to obtainWhat it cannot replace
Specification / firmware ownerSelected Core 6.2 and update 28108; HE amendment identity; LoRa L2/RP versions and errata; exact cellular tuple; GPS IRN-004 impactA product capability list is not a normative requirement or a measured service result.
Qualification / certification ownerExact Bluetooth product record, Wi-Fi SKU/firmware certificate, cellular GCF/PTCRB/module/host scopeRF margin and weights do not close program obligations or establish test coverage.
Operator / service ownerDeployed band/features, SIM/provisioning, coverage, roaming, lifecycle and host endpointA release’s feature catalogue does not establish operator deployment.
Market / compliance ownerCountry, band, device class, antenna, installation, host modifications and applicable legal requirementsA qualification program is not national law; no market approval is inferred here.
Product evidence / systems ownerInstalled antenna/channel population, concurrent RF and supply states, deadline and energy observationsA source document cannot substitute for an actual product observation.

Bluetooth’s qualification guidance, an actual Wi-Fi certificate and PTCRB/GCF process pages illustrate distinct evidence scopes. PTCRB explicitly separates additional operator and governmental obligations; GCF’s public feature table is labelled January 2024 and cannot establish current device capability. Detailed member criteria were not accessed. LoRa regional parameters were checked at catalogue level only in this review; no inaccessible national or normative table is reconstructed. [BT-QUAL] [WFA] [PTCRB] [GCF] [LORA-RP]

Common misconceptionOne technology must win every role and market.

A local sensor link, high-capacity gateway backhaul, GNSS timing and delayed satellite fallback can have different owners and service boundaries. A hybrid is defensible only when each role has its own evidence and the shared product passes its concurrent conditions.

Add source/version/market status beside the technical screen. Keep it visible when preferences produce a leader.

08 / 10

Apply hard constraints before preference

C is easiest and quickest to integrate. Can it win while missing the delivery requirement?

Think about itPick the naive score winner before evaluating the hard constraints.
Answer

C has the best illustrative effort and readiness values, so its naive score is 1 at every weight. Its .90 delivery assumption fails .99. Reject it before computing any eligible utility or Pareto set; otherwise the score rewards an architecture that does not provide the required service.

The kernel below uses A=6 mJ/.995/25 Mbit/s, B=120 mJ/.999/30 Mbit/s, C=16 mJ/.90/35 Mbit/s and D’s missing required evidence. These are chosen acceptance inputs, not measured, predicted or certified technology performance. A has 0 dB nominal blocker headroom at 55 dB isolation and a separately required protected-state plan; B has +10 dB. C’s illustrative internal headroom is +5 dB. D’s local backhaul is architecturally inapplicable; its energy, deadline and RF evidence remain unknown.

Think about itD’s required energy is unknown. Can an attractive readiness estimate make it eligible?
Answer

No. Required energy and delivery evidence must be established before preferences are evaluated. D remains inspect, with no eligible score. Its reasoned local-backhaul exemption only resolves that one cell; it does not supply the missing radio energy or deadline evidence.

state = fail if any required predicate fails; else inspect if any is unknown; else passPass means feasible in this synthetic technical fixture only. Preserve every failure/missing reason. A known failure takes precedence over unknowns in the summary without erasing them. No fail or inspect candidate enters preference scoring.
Inclusive threshold boundary checks · comparisons use full precision
CriterionJust belowExactly atJust above
Delivery ≥.99.989999: fail.99: pass.990001: pass
Energy ≤200 mJ199.999: pass200: pass200.001: fail
Gateway ≥20 Mbit/s19.999999: fail20: pass20.000001: pass
Headroom ≥0 dB−.000001: fail0: pass+.000001: pass
Required unknowninspectNot a boundary valueNever silently zero
veffort=clip(50effort40,0,1)vtime=clip(10weeks8,0,1)v_{\mathrm{effort}} = \operatorname{clip}(\frac{50-\mathrm{effort}}{40},0,1)\qquad v_{\mathrm{time}} = \operatorname{clip}(\frac{10-\mathrm{weeks}}{8},0,1)Effort is a synthetic integration unit with fixed best/worst anchors 10/50. Readiness uses 2/10 weeks. Clip explicitly beyond those anchors; negative effort or weeks is invalid. Inserting a candidate never changes the anchors.
Score=wveffort+(1w)vtime0w1\text{Score} = w v_{\mathrm{effort}} + (1-w) v_{\mathrm{time}}\qquad 0 \le w \le 1w is the owner’s preference for integration effort relative to readiness, default .6. It is not empirical confidence, a legal status or a concealed evidence multiplier. Eligible Pareto comparisons use these same two utilities.

A’s utilities are (.75,.25); B’s are (.50,.75). At w=.6, A=.55 and B=.60. At w=.8, A=.65 and B=.55. D has no eligible score. If every candidate fails or needs inspection, the result is “no supported choice,” never a least-bad winner. This constraint-first workflow follows the decision-record discipline in NASA’s guidance; the particular utility functions and priorities are original teaching rules. [NASA]

Common misconceptionA weighted score compensates for a failed hard constraint.

A score is meaningful only among options that satisfy the declared requirements. Waiving a failure is a proposed requirement change with a new scenario and owner review. It is not an adjustment to a preference weight.

Technology & Coexistence Decision Studio

Follow one decision from assumptions to a discriminating observation. Every number here is illustrative. The complete baseline, equations, source cards and answers are server rendered.

Predict C first, reveal its delivery failure, compare weights .6 and .8, then load the exact tie. Examine the simultaneous isolation interval, protect the internal radio, and finally require the external threat. A preset restores its complete inputs and metadata atomically.

Scenario identity
Only the matched common basis is valid for these cards. Repair the reconciliation before comparing.
Candidate set · select 2–6; default A–D

E demonstrates fixed anchors and dominance; F demonstrates unknown evidence and beyond-anchor effort. Adding a candidate does not rescale anyone else.

Hard limits and RF inputs · finite numbers; arbitrary decimal step
01; default 0.6. Step: any finite decimal.
-1030; default 20. Step: any finite decimal.
20100; default 55. Step: any finite decimal. With B selected, A≤90 dB; interval mode requires A=23–87 dB.
01000; default 200. Step: any finite decimal.
01; default 0.99. Step: any finite decimal.
0100; default 20. Step: any finite decimal.
020; default 0. Step: any finite decimal.
Closed assumptions · default fixed, protected, coordinated, reviewed
Default: Fixed hypothetical inputs. Choose one listed state.
Default: Declared synthetic technical fixture. Choose one listed state.
Default: Protected RX + 2 ms guards. Choose one listed state.
Default: Complete synthetic guard assumption. Choose one listed state.
Default: Internal coordinated. Choose one listed state.
Default: V82: 82 dB integrated factor. Choose one listed state.
Default: Matched R1-TX → R1-RX. Choose one listed state.
Default: v1 reviewed; due 2026-12-07. Choose one listed state.
Default: Internal coupling / blocking. Choose one listed state.
Deterministic schedule · 100 ms cycle, at most 8 intervals per role

Enter start:end in ms, separated by semicolons. Use 90:10:wrap for a boundary crossing; equal endpoints are empty. Enter none for no intervals. Endpoints 0–100, any finite decimal; no reversed interval without :wrap.

Default 0:20 ms. 2 ms guards are allocated around the RX union.
Default 22:62 ms. 2 ms guards are allocated around the RX union.
Named threat set · 1–12 rows; T01 required; default T01–T08

Canonical A/B/C/D baseline · preserved alongside the committed scenario
HypothesisEnergy mJ / generatedDelivery ≤2 sGateway Mbit/sEffort / weeks
A · Local node + coordinated gateway60.9952520 / 8
B · Cellular node + separated reception1200.9993030 / 4
C · LPWAN + shared gateway hypothesis160.93510 / 2
D · Satellite fallback hypothesisUnknownUnknownN/A: Direct terminal-to-host architecture has no local gateway backhaul role.25 / 6

Committed results

technology-coexistence-decision-studio/2.0 · p07-m07-decision-v1 · p07-m07-constraint-rules-v1. Evaluation 2026-09-08 UTC; synthetic review due 2026-12-07. Version review: current.

Conditional fixture choice: B. Real release evidence remains open.

Scenario: p07-m07-decision-v1; fixed; fixture; A plan protected; w=0.6. Basis p07-m07-common-basis-v1. Only the declared technical fixture is screened. Real external, supply, clock, recovery and market evidence remains open.

Committed candidate assumptions · hypothetical acceptance inputs
HypothesisEnergy mJ / generatedDelivery ≤2 sGateway Mbit/sEffort / weeks
A · Local node + coordinated gateway60.9952520 / 8
B · Cellular node + separated reception1200.9993030 / 4
C · LPWAN + shared gateway hypothesis160.93510 / 2
D · Satellite fallback hypothesisUnknownUnknownN/A: Direct terminal-to-host architecture has no local gateway backhaul role.25 / 6
Common hard limits · inclusive, unrounded comparisons
Quantity / common boundaryCommitted limit
Radio-supply energy, failures included≤200 mJ/generated report
Unique complete D0 reports by 2 s≥0.99
Gateway D0 service, where required≥20 Mbit/s
On-time blocker headroom≥0 dB; noise rise ≤1 dB separately
Scenario integrityp07-m07-common-basis-v1; 32 bytes / 600 s / 1000 nodes. Changed limits are a custom proposal requiring requirement-owner review.

RF and timing ledgers

T01 V82/V72: −35 dBm synthetic blocker limit; −110 dBm input-referred noise over 200 kHz; 82/72 dB integrated factor. On-time power at matched R1-TX → R1-RX; B shares A’s isolation variable plus 10 dB. Wanted −105 dBm, mounting A, 25 °C; exact configurations remain visible in the path matrix.

Required-mode RF ledger · no duty correction
Candidate / I dBCoupled / leakage dBmH / noise rise dBUnscheduled simultaneous state
A / 55-35 / -1170 / 0.790097497Permitted by these synthetic RF screens only
B / 65-45 / -12710 / 0.085799992Permitted by these synthetic RF screens only
Declared deterministic 100 ms scheduleHalf-open interval unions. Solid dark bars show internal TX, outlined cyan bars victim RX, dashed hatched bars their overlap. Guard recovery remains an independent assumption. The following table lists all endpoints and durations.020406080100TXRXOverlapms
Illustrative schedule, not a measured RF capture. A bar includes its start and excludes its end. A blank overlap lane means no declared overlap; it says nothing about unlisted external traffic or receiver settling.
Timeline alternative · half-open union, 100 ms cycle
QuantityCommitted value
TX union (ms)0:20
RX union (ms)22:62
Intersection (ms)none
Overlap duration / cycle fraction0 ms / 0
Overlap / victim listening0
TX / RX duty fractions0.2 / 0.4
RX plus 2 ms guards (ms)20:64
Protected planGeometric guards clear; internal coordinated; recovery assumed. Geometry alone is not evidence.
Remaining allocation / HE goodput0.56 / 25.364911765 Mbit/s if every remaining interval is allocated. Availability applied once; not a deadline prediction.

Screen before scoring

First binding or missing issue: C / delivery: 0.9 ≥ 0.99 (unrounded SI comparison).

Complete technical screen and eligible utilities
Candidate / stateAll predicate statesEffort / time utilitiesEligible score
A · passenergy: pass; delivery: pass; backhaul: pass; blocker: pass; leakage: pass; protected-plan: pass; scheduled-capacity: pass0.75 / 0.250.55
B · passenergy: pass; delivery: pass; backhaul: pass; blocker: pass; leakage: pass0.5 / 0.750.6
C · failenergy: pass; delivery: fail; backhaul: pass; blocker: passExcluded before utility evaluationNo eligible score
D · inspectenergy: inspect; delivery: inspect; backhaul: pass; blocker: inspectExcluded before utility evaluationNo eligible score
Every hard predicate and reason, including failed and missing evidence

A · pass

  • energy: pass. 0.006 ≤ 0.2 (unrounded SI comparison).
  • delivery: pass. 0.995 ≥ 0.99 (unrounded SI comparison).
  • backhaul: pass. 25000000 ≥ 20000000 (unrounded SI comparison).
  • blocker: pass. Unscheduled simultaneous pass: 0 ≥ 0 (unrounded SI comparison). Protected operation is evaluated separately.
  • leakage: pass. Unscheduled simultaneous pass: 0.790097497 dB noise rise; maximum 1 dB; compare inverse isolation threshold at full precision. Protected operation is evaluated separately.
  • protected-plan: pass. Two-ms guard clear under an explicitly assumed synthetic recovery contract.
  • scheduled-capacity: pass. 25364911.76470589 ≥ 20000000 (unrounded SI comparison).

B · pass

  • energy: pass. 0.12 ≤ 0.2 (unrounded SI comparison).
  • delivery: pass. 0.999 ≥ 0.99 (unrounded SI comparison).
  • backhaul: pass. 30000000 ≥ 20000000 (unrounded SI comparison).
  • blocker: pass. 10 ≥ 0 (unrounded SI comparison).
  • leakage: pass. 0.085799992 dB noise rise; maximum 1 dB; compare inverse isolation threshold at full precision.

C · fail

  • energy: pass. 0.016 ≤ 0.2 (unrounded SI comparison).
  • delivery: fail. 0.9 ≥ 0.99 (unrounded SI comparison).
  • backhaul: pass. 35000000 ≥ 20000000 (unrounded SI comparison).
  • blocker: pass. 5 ≥ 0 (unrounded SI comparison).

D · inspect

  • energy: inspect. Required evidence unknown.
  • delivery: inspect. Required evidence unknown.
  • backhaul: pass. Not applicable: Direct terminal-to-host architecture has no local gateway backhaul role.
  • blocker: inspect. Required evidence unknown.

Feasible in fixture: A, B. Rejected: C. Inspect: D. Eligible Pareto set: A, B. Pareto compares only effort/time utility; it does not close release gates.

Fixed anchors: effort 10 best / 50 worst, readiness 2 best / 10 worst. Utilities clip to [0,1]. Score = w·veffort + (1−w)·vtime. C’s naive score is 1, but C never enters a score set while delivery fails. Exact A/B crossover is w=2/3, both 7/12; retain tied IDs in alphabetical navigation order.

p07-m07-threats-v1 · every listed path retains its own evidence
Path / victim ← aggressorMechanism / controlRequired state / evidence
T01 · A-WIFI-NODE: Weak-node gateway RX ← Internal gateway Wi-Fi TXBlocking + noise-like leakage; internal coordinatedprotected Illustrative V82: +20 dBm, total I=55 dB nominal; no measured suppression or receiver threshold.
T02 · EXT-WIFI-NODE: Node/gateway receive ← External Wi-Fi networkCochannel collision / adjacent leakage; externalRequired state to establish; no measured duty or overlap. Unknown; internal PTA has no authority over the external transmitter.
T03 · CELL-GNSS: GNSS timing receiver ← Internal cellular TXBlocking / harmonics; internal uncoordinatedRequired state to establish; no measured duty or overlap. Unknown; no harmonic magnitude follows from a carrier name.
T04 · EXT-GNSS: GNSS receiver ← External interferenceNoise / blocking; externalRequired state to establish; no measured duty or overlap. Unknown; satellite visibility does not establish interference immunity.
T05 · SUPPLY-RX: Receiver supply ← DC converter / radio loadSupply coupling; internal uncoordinatedRequired state to establish; no measured duty or overlap. Unknown; total antenna isolation cannot characterize supply transfer.
T06 · CLOCK-RX: Receiver / antenna ← Digital clock / cableDigital / common-mode coupling; internal uncoordinatedRequired state to establish; no measured duty or overlap. Unknown; no spectral mask generated from the technology name.
T07 · LO-MIX: Victim mixer ← Strong off-channel signal + LO noiseReciprocal mixing; internal uncoordinatedRequired state to establish; no measured duty or overlap. Unknown; V82 excludes reciprocal mixing.
T08 · RECOVERY: Weak-node receiver ← Internal TX turn-offReceiver recovery / filter settling; internal coordinatedRequired state to establish; no measured duty or overlap. Unknown in product; 2 ms is only a chosen guard allocation.
Inspect every path’s configuration, planes and next observation
  1. T01 · A-WIFI-NODE

    5.180 GHz, HE-style SU MCS 6, 20 MHz; synthetic 2.450 GHz victim, 200 kHz ENBW; +2.730 GHz offset; mounting A, 25 °C; on-time total-power / integrated-noise detectors.

    Planes: R1-TX → R1-RX. Wanted: −105 dBm at R1-RX, chosen weak wanted stimulus only; not a sensitivity limit.

    State: protected; 20 dBm on-time; I=55 dB for A; V82; source plane R1. TX 0:20 ms; RX 22:62 ms.

    Next evidence: RF owner: coupling, selectivity, blocker and recovery evidence.

  2. T02 · EXT-WIFI-NODE

    2.4 GHz external channel/width/traffic unknown; mounting A; packet detector and observation population unknown.

    Planes: S0 → R2-RX → R1-RX → R3. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: Field owner: time-aligned external occupancy and unique D0 delivery.

  3. T03 · CELL-GNSS

    Selected LTE-M band/resource/offset and GNSS channel unresolved; mounted receiver and integrated bandwidth unknown.

    Planes: R1-TX → R1-RX. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: RF/timing owners: concurrent RF and calibrated PPS logs.

  4. T04 · EXT-GNSS

    GPS L1 C/A candidate; interferer waveform/offset, receiver bandwidth and wanted C/N₀ population unknown.

    Planes: S0 → R1-RX → R3. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: Timing owner: independent reference and receiver-input characterization.

  5. T05 · SUPPLY-RX

    Switching state/harmonic and load step unknown; mounting A; voltage spectrum/settling detector required.

    Planes: Supply → receiver; not R1 RF power. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: Power owner: probe supply with RF path terminated and controlled load.

  6. T06 · CLOCK-RX

    Clock modes, harmonics, cable layout and receiver band unknown.

    Planes: Clock/cable → R2/R1. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: Digital/RF owners: toggle clock mode and compare conducted/near-field signatures.

  7. T07 · LO-MIX

    LO noise at blocker offset, receiver bandwidth and filter state unknown; input-referred integrated noise.

    Planes: R1-RX → mixer → R3. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: Receiver owner: independent LO-noise and blocker-level tests.

  8. T08 · RECOVERY

    Actual TX ramp, AGC/filter state and recovery-to-decode statistic unknown; mounting A.

    Planes: R1-RX → R3 timing. Wanted: Unknown R1-RX wanted level; must be pinned for a real receiver test.

    State: Required state to establish; no measured duty or overlap.

    Next evidence: Firmware/RF owners: request/grant/RF/decision timestamp correlation.

Independent evidence gates and experiment

Real qualification, operator, host, market and lifecycle status: unknown / review required, independently of current synthetic source review. Review overdue means recheck; it does not erase a published physical definition.

Experiment ordering · lexicographic teaching rule
Experiment / relative effortHard relevance / potential changeWhy here
cellular / 3 unitsRelevant / Can change eligible set or selectionReplace B’s assumed energy/deadline evidence; an outage/search failure can remove B from the eligible set.
lpwan / 4 unitsRelevant / Can change eligible set or selectionA matched unique-delivery observation can replace C’s .90 assumption; only a result at or above the chosen deadline threshold can repair this hard screen.
coupling / 2 unitsNo current hard reversal / Not establishedCharacterization remains useful, but other predicates or the selected plan prevent an established current eligibility reversal.

Order: relevance to an unresolved/failed requirement for a plausible choice, then a possible eligibility/selection change, then lower stated effort. Retain exact ties. Effort is neither a price nor a promised lab duration.

E01 · Is the internal coupling path below the decision boundary?

Owner / configuration: RF + receiver owners. A-WIFI-NODE, required simultaneous scenario; 5.180 GHz HE-style MCS 6 / 20 MHz aggressor, synthetic 2.450 GHz weak victim, 200 kHz ENBW, mounting A, 25 °C. Pin actual product waveform and detector before execution.

Observe: Calibrated on-time +20 dBm at R1-TX; transfer/coupled level at R1-RX in dBm; isolation and input-referred noise in dB; time-aligned R3 decode and D0 delivery.

Control / competing explanations: Vary a documented feed/antenna isolation configuration across 55 dB while holding wanted stimulus −105 dBm, gain/filter, supply and traffic fixed. Compare RF-terminated and clock/load controls.

Calibration / uncertainty: De-embed fixture loss once; source purity, leakage floor, bandwidth, receiver calibration, placement repeats and expanded uncertainty must resolve which side of 55 dB the result occupies. A bound touching the boundary remains inconclusive for release.

Contrasting outcomes: 52 dB → H=−3 dB and leakage screen fails: reject A in simultaneous mode. 58 dB → H=+3 dB and leakage clears: admit A only if all other predicates hold. At w=.8 selection can reverse; at .6 B still leads.

Decision and review: Disentangle blocking from reciprocal mixing, nonlinear products, supply coupling and recovery; do not fit an RF isolation number to a supply failure. Owner records configuration and next review by 2026-12-07.

Committed decision record · illustrative, ready for review

Decision: Conditional fixture choice: B. Real release evidence remains open. A additionally depends on internal control and verified recovery.

Alternatives: C is rejected when .90 delivery is below the common limit; D is deferred while node energy, delivery and RF evidence are missing. A and B may be retained as conditional alternatives or a hybrid with separately budgeted roles.

Dissent: preference anchors and w=0.6 reflect the exercise owner, not consensus. A different weight can reverse A/B; missing external/installation evidence can invalidate either. Replacing a failed requirement needs a named new scenario.

Evidence and action: p07-m07-decision-v1; p07-m07-common-basis-v1; p07-m07-threats-v1; p07-m07-constraint-rules-v1; selected experiment coupling. Systems owner reviews by 2026-12-07, earlier on mode, firmware, antenna, market or requirement change. No product release approval or lab request is generated.

The committed results become the middle of the review artifact. Use the next section to explain how fragile the conditional choice is and which new observation would matter.

09 / 10

Find reversals and choose an experiment

Would a different preference or one coupling measurement reverse the decision?

Think about itB leads at w=.6. What happens when effort weight rises to .8, with the same hard evidence?
Answer

A’s stronger effort utility now dominates the preference difference: A=.65 and B=.55, so A leads if it remains eligible. This is a preference reversal, not new RF evidence. At the exact crossover retain both candidates instead of rounding a displayed weight into a sole winner.

A=.25+.50wB=.75.25wA=Bw=23,Score=712A = .25 + .50w\qquad B = .75 - .25w\qquad A=B \Rightarrow w=\frac{2}{3}, \text{Score}=\frac{7}{12}The exact algebra gives the tie. Below 2/3, B leads; above it, A leads, provided each remains eligible. At equality retain both. Stable alphabetical ID order supports navigation and is not a tiebreaking recommendation.

Now change the requirement explicitly: both radios must operate simultaneously. This is a different scenario from A’s protected plan. Let one named shared variable z∈[−1,+1] set A’s total isolation to 55+3z dB and B’s to 65+3z dB. Do not independently choose an optimistic A and pessimistic B from the same shared source. Bounds describe possibilities, not confidence intervals or an unlabelled probability distribution.

Required simultaneous alternative · all other fixture limits unchanged
Shared cornerA / B isolationA / B headroomLeakage and choice
z=−152 / 62 dB−3 / +7 dBA leakage rise 1.455404631 dB also fails. B clears both RF screens. B selected at .6 and .8.
z=0 nominal55 / 65 dB0 / +10 dBA/B clear; A nominally leads at .8, B at .6. A’s interval status still inspect.
z=+158 / 68 dB+3 / +13 dBA leakage rise .413926852 dB clears. A selected at .8 if all other predicates pass; B at .6.

The studio recomputes leakage and every applicable hard predicate at each corner. A is inspect across the interval because its outcomes differ; B passes this interval fixture. A measured result whose uncertainty still straddles 55 dB does not resolve the decision. Nor does typing a larger isolation number retire the engineering risk.

Think about itIs the internal coupling experiment equally decisive at w=.6 and w=.8?
Answer

No. At .8 it can switch the selected portfolio between B and A. At .6 it can change A’s eligibility, but B already leads on the specified preferences. It is still informative, with a different immediate decision consequence.

Choose an experiment with contrasting outcomes before entering a chamber. The studio’s explicit order is: relevance to an unresolved or failed hard requirement for a plausible candidate; then possible eligible-set or selection change; then lower synthetic effort. Exact ties are retained. This is an explanatory rule, not an expected financial value calculation.

E01 · Is the internal coupling path below the decision boundary?

Owner / configuration: RF + receiver owners. A-WIFI-NODE, required simultaneous scenario; 5.180 GHz HE-style MCS 6 / 20 MHz aggressor, synthetic 2.450 GHz weak victim, 200 kHz ENBW, mounting A, 25 °C. Pin actual product waveform and detector before execution.

Observe: Calibrated on-time +20 dBm at R1-TX; transfer/coupled level at R1-RX in dBm; isolation and input-referred noise in dB; time-aligned R3 decode and D0 delivery.

Control / competing explanations: Vary a documented feed/antenna isolation configuration across 55 dB while holding wanted stimulus −105 dBm, gain/filter, supply and traffic fixed. Compare RF-terminated and clock/load controls.

Calibration / uncertainty: De-embed fixture loss once; source purity, leakage floor, bandwidth, receiver calibration, placement repeats and expanded uncertainty must resolve which side of 55 dB the result occupies. A bound touching the boundary remains inconclusive for release.

Contrasting outcomes: 52 dB → H=−3 dB and leakage screen fails: reject A in simultaneous mode. 58 dB → H=+3 dB and leakage clears: admit A only if all other predicates hold. At w=.8 selection can reverse; at .6 B still leads.

Decision and review: Disentangle blocking from reciprocal mixing, nonlinear products, supply coupling and recovery; do not fit an RF isolation number to a supply failure. Owner records configuration and next review by 2026-12-07.

E02 · Does search consume the energy reserve?

Owner / configuration: Power + cellular/service owners. P07-M07-CELLULAR-v1, one pinned SKU/firmware/band/operator/SIM; 32 D0 bytes / 600 s; registered, weak-coverage, search, outage and abort strata.

Observe: Integrate radio-supply V(t)I(t) in joules over generated-report periods, including failures; paired application release/delivery timestamps yield unique reports by 2 s. Separate radio-supply boundary from whole-node load.

Control / competing explanations: Vary attenuated network availability/registration state; hold traffic, supply, firmware and temperature. Distinguish long RF airtime, network grant delay, host queueing and modem search.

Calibration / uncertainty: Calibrate voltage/current and time alignment; capture peaks and sleep, disclose censoring, strata weights and uncertainty for both energy and delivery. A 10% search fraction is a scenario, not a measured distribution.

Contrasting outcomes: Registered illustrative 111.55968 mJ is below 200 mJ; 10% search mixture 543.50568 mJ fails. Delivery must independently meet .99 by 2 s; neither energy result proves it.

Decision and review: Passing evidence may replace B’s assumption; energy or deadline failure removes B and triggers a supply/service architecture review. Owner/review 2026-12-07.

E03 · Can the service deliver the urgent message?

Owner / configuration: Field + network-service owners. P07-M07-LPWAN-v1: pinned Class A, SF/BW/coding, regional channel plan, gateway placement and backhaul. Keep 1000-node-equivalent offered traffic and deployment strata explicit.

Observe: Unique complete 32-byte D0 reports by 2 s / generated reports, plus a separately declared urgent-downlink population; node/gateway/server timestamps in seconds. Record retries, airtime and receive-window opportunities.

Control / competing explanations: Vary gateway placement and load separately; hold node generation and commanded opportunities. Compare RF decode, collision, duty scheduling, missing receive windows and server delay.

Calibration / uncertainty: Calibrate timestamp offsets; count duplicates once and failed/censored trials in the denominator. Declare sample size, confidence procedure and population; never substitute the e^(−2G) proxy for an observed deadline fraction.

Contrasting outcomes: .90 matched delivery → reject C; evidence supporting ≥.99 with the chosen uncertainty rule may admit it if energy, RF and infrastructure predicates also clear. Missing command windows require redesign or a new deadline scenario.

Decision and review: Preserve evidence and dissent; field owner proposes the next constrained experiment, review 2026-12-07. No automatic lab request.

Record the selected observation, owner, uncertainty rule and action after each outcome. Laboratory execution belongs to the planned Path 08; no lab request or schedule is sent by this lesson.

10 / 10

Record the connectivity portfolio review

Could another engineer reconstruct your choice—and identify the evidence that would overturn it?

Produce a Connectivity portfolio review, an ungraded Learn synthesis. Use the source cards below as complete starting snapshots, then replace assumptions with traceable observations. Separate a conditional technical selection from release readiness and preserve dissent about requirements, utility anchors and uncertain populations.

P07-M07-LOCAL-v1 · complete source evidence card
p07-technology-evidence-card-v1 · P07-M07-LOCAL-v1 · own static snapshot
FieldRecorded evidence
Identity / ownerP07-M07-LOCAL-v1; Local-radio + gateway RF owners
Inherited immutable IDsP07-M06-GNSS-TIMING-v1 · P07-M06-SBD-FALLBACK-v1 · P07-M05-CELLULAR-OPTIONS-v1 · P07-M04-REMOTE-FLEET-v1 · P07-M03-HE-BACKHAUL-v1 · P07-M02-LE-CONN-v1 · P07-M02-THREAD-SED-v1 · P07-M01-LE-EXT-v1 · p06-evidence-map-v1 · M01-A · M01-B-LOSS · M01-INSTALL-UNKNOWN
Scenario / requirementp07-m07-common-basis-v1: fictional 1000-node condition-monitoring fleet; 32 D0 bytes every 600 s. Fixed nodes, occasional maintenance repositioning; installed metal enclosure/open-pole strata and orientations must be sampled. Population weights, channel distributions and mobility transitions unknown. ≥.99 unique complete reports by 2 s; ≤.200 J radio-supply energy per generated report, including failures; ≥20 Mbit/s D0 gateway service where applicable.
Family / selected versionBluetooth LE node + Wi-Fi gateway backhaul. Selected Bluetooth Core 6.2, LE 1M connection candidate; update 28108 impact open. Selected IEEE 802.11ax-2021 HE SU, 20 MHz, MCS 6, one stream, 0.8 µs GI; exact product capability unknown.
Region / classCandidate LE 2.4 GHz and gateway 5.180 GHz test configuration; CH/EU device class and permitted operating conditions unresolved.
D0 / deadline256 bits; 600 s period; 2 s deadline. Node application release → nominated gateway/service application D0; include discovery, queues, retries, network and host delivery. A PHY decode at R3 is not D0.
PHY / MAC resources256 D0 bits; 384 illustrative D1 bits for node calculation. Gateway: 234 data tones × 6 bits × 3/4 per 13.6 µs symbol; not a MAC deadline model.
State duration / currentSynthetic wake: 0.005 s, 0.003 A; Discovery: 0.01 s, 0.006 A; TX per attempt: 0.000384 s, 0.012 A; RX per attempt: 0.001 s, 0.006 A; Wait per attempt: 0.002 s, 0.002 A; Residual sleep; after expected active duration: 599.9814468 s, 0.000002 A
Antenna / RF interfacesR1 component port; R2 antenna feed; S0 installed spatial population; R3 receiver decision. Record R2↔R1 feed/filter loss once. Antenna pattern, detuning, volume, ground and mutual coupling remain unknown.
Infrastructure / ownershipOwned gateway, Wi-Fi access point, installation, upstream capacity and maintenance ownership required; gateway energy is outside the stated node boundary and needs a separate budget.
QualificationExact product/SKU/firmware/profile listing, host configuration and concurrent-operation test coverage unknown; individual qualification is not product coexistence evidence.
Regulatory questionsCH and EU/EEA proposed, deployment countries not frozen. Band, device class, antenna, installation, national applicability and evidence owner unresolved; no authorization inferred.
Claim-level provenance
  • LOCAL-IDENTITY: Selected documentary identity; actual implementation unknown. [BT-62] [BT-QUAL] [IEEE-AX] [WFA]
  • LOCAL-STATE: Derived synthetic 600 s variant; 3 V, p=.05, at most two attempts. Not delivery by 2 s. [RECON]
Uncertainty / decisionThese source cards do not supply the synthetic A–D acceptance numbers. Delivery, installation distribution and concurrent state evidence remain unknown. Conditional local + backhaul portfolio if owned infrastructure and control/recovery evidence hold. Technical fixture A is a separate hypothetical card.
Next evidence / reviewRF owner: T01 coupling and recovery; firmware owner: supply/state and D0 timestamp trace. Due 2026-12-07. Frozen review basis 2026-09-08 UTC; recheck at due date or mode, firmware, antenna, operator, market or requirement change.
P07-M07-CELLULAR-v1 · complete source evidence card
p07-technology-evidence-card-v1 · P07-M07-CELLULAR-v1 · own static snapshot
FieldRecorded evidence
Identity / ownerP07-M07-CELLULAR-v1; Cellular + service owners
Inherited immutable IDsP07-M06-GNSS-TIMING-v1 · P07-M06-SBD-FALLBACK-v1 · P07-M05-CELLULAR-OPTIONS-v1 · P07-M04-REMOTE-FLEET-v1 · P07-M03-HE-BACKHAUL-v1 · P07-M02-LE-CONN-v1 · P07-M02-THREAD-SED-v1 · P07-M01-LE-EXT-v1 · p06-evidence-map-v1 · M01-A · M01-B-LOSS · M01-INSTALL-UNKNOWN
Scenario / requirementp07-m07-common-basis-v1: fictional 1000-node condition-monitoring fleet; 32 D0 bytes every 600 s. Fixed nodes, occasional maintenance repositioning; installed metal enclosure/open-pole strata and orientations must be sampled. Population weights, channel distributions and mobility transitions unknown. ≥.99 unique complete reports by 2 s; ≤.200 J radio-supply energy per generated report, including failures; ≥20 Mbit/s D0 gateway service where applicable.
Family / selected versionLTE-M node and separately specified gateway reception. Inherited real trace: LTE-M Cat-M1 candidate, TS 36.101 V13.25.0; nRF9160 revision 2 (B1), modem 1.3.7. Exact deployed band/resource/operator tuple remains unverified here. Synthetic T1 is not that normative product.
Region / classCH/EU candidates; actual operator, SIM, band/duplex/resource, power class, antenna and certification combination unknown.
D0 / deadline256 bits; 600 s period; 2 s deadline. Node application release → nominated gateway/service application D0; include discovery, queues, retries, network and host delivery. A PHY decode at R3 is not D0.
PHY / MAC resources32 D0 bytes each 600 s; framed/coded bits, repetitions, grants and deployed feature availability unknown.
State duration / currentRegistered / transmit / receive / sleep: unknown, unknown; Search / attach / retry / abort / outage: unknown, unknown
Antenna / RF interfacesR1 component port; R2 antenna feed; S0 installed spatial population; R3 receiver decision. Record R2↔R1 feed/filter loss once. Antenna pattern, detuning, volume, ground and mutual coupling remain unknown.
Infrastructure / ownershipOperator coverage, provisioning, roaming, service lifecycle and host endpoint required. The B gateway service number is an independent architecture hypothesis, not LTE-M node throughput.
QualificationExact product/SKU/firmware/profile listing, host configuration and concurrent-operation test coverage unknown; individual qualification is not product coexistence evidence.
Regulatory questionsCH and EU/EEA proposed, deployment countries not frozen. Band, device class, antenna, installation, national applicability and evidence owner unresolved; no authorization inferred.
Claim-level provenance
  • CELL-VERSION: Process/version scope read; current deployed compatibility unknown. [3GPP] [PTCRB] [GCF]
  • CELL-ENERGY: Inherited illustrative registered/search state arithmetic; real state distribution unknown. [RECON]
Uncertainty / decisionThese source cards do not supply the synthetic A–D acceptance numbers. Delivery, installation distribution and concurrent state evidence remain unknown. Defer real selection until search/outage energy and D0 service are observed; B=120 mJ is chosen acceptance input only.
Next evidence / reviewPower/service owners: E02 supply integration and unique deadline delivery, including failed generated reports. Due 2026-12-07. Frozen review basis 2026-09-08 UTC; recheck at due date or mode, firmware, antenna, operator, market or requirement change.
P07-M07-LPWAN-v1 · complete source evidence card
p07-technology-evidence-card-v1 · P07-M07-LPWAN-v1 · own static snapshot
FieldRecorded evidence
Identity / ownerP07-M07-LPWAN-v1; LPWAN + field-service owners
Inherited immutable IDsP07-M06-GNSS-TIMING-v1 · P07-M06-SBD-FALLBACK-v1 · P07-M05-CELLULAR-OPTIONS-v1 · P07-M04-REMOTE-FLEET-v1 · P07-M03-HE-BACKHAUL-v1 · P07-M02-LE-CONN-v1 · P07-M02-THREAD-SED-v1 · P07-M01-LE-EXT-v1 · p06-evidence-map-v1 · M01-A · M01-B-LOSS · M01-INSTALL-UNKNOWN
Scenario / requirementp07-m07-common-basis-v1: fictional 1000-node condition-monitoring fleet; 32 D0 bytes every 600 s. Fixed nodes, occasional maintenance repositioning; installed metal enclosure/open-pole strata and orientations must be sampled. Population weights, channel distributions and mobility transitions unknown. ≥.99 unique complete reports by 2 s; ≤.200 J radio-supply energy per generated report, including failures; ≥20 Mbit/s D0 gateway service where applicable.
Family / selected versionLoRa modulation / LoRaWAN Class A candidate. Selected TS001-1.0.4 L2; RP002-1.0.5 identity with errata review open. SX1276-style LoRa, 125 kHz, SF7–12, coding 4/5, explicit header and uplink CRC; actual regional mapping unknown.
Region / classCH/EU proposed; regional channel/data-rate/class, national conditions and gateway installation unresolved. No generic ISM permission inferred.
D0 / deadline256 bits; 600 s period; 2 s deadline. Node application release → nominated gateway/service application D0; include discovery, queues, retries, network and host delivery. A PHY decode at R3 is not D0.
PHY / MAC resources45 illustrative PHY payload bytes for 32 D0 bytes, no FOpts, application FPort; shared three-channel access model does not imply a deployed channel plan.
State duration / currentSF7 TX arithmetic: 0.092416 s, unknown; Receive windows / startup / retries / sleep: unknown, unknown
Antenna / RF interfacesR1 component port; R2 antenna feed; S0 installed spatial population; R3 receiver decision. Record R2↔R1 feed/filter loss once. Antenna pattern, detuning, volume, ground and mutual coupling remain unknown.
Infrastructure / ownershipPrivate gateway ownership, placement, backhaul, network server and maintenance required; receive-window and downlink availability must be demonstrated.
QualificationExact product/SKU/firmware/profile listing, host configuration and concurrent-operation test coverage unknown; individual qualification is not product coexistence evidence.
Regulatory questionsCH and EU/EEA proposed, deployment countries not frozen. Band, device class, antenna, installation, national applicability and evidence owner unresolved; no authorization inferred.
Claim-level provenance
  • LPWAN-MODE: L2 text read; regional identity only, manufacturer calculator interface only. [LORA-L2] [LORA-RP] [SEMTECH]
  • LPWAN-SERVICE: Airtime and ALOHA proxy derived; no measured D0 deadline distribution. [RECON]
Uncertainty / decisionThese source cards do not supply the synthetic A–D acceptance numbers. Delivery, installation distribution and concurrent state evidence remain unknown. Reject C’s .90 delivery hypothesis for the current .99 requirement; a different delay requirement needs a named owner-reviewed scenario.
Next evidence / reviewField/service owner: E03 unique-report and urgent-command timestamp experiment. Due 2026-12-07. Frozen review basis 2026-09-08 UTC; recheck at due date or mode, firmware, antenna, operator, market or requirement change.
P07-M07-SBD-v1 · complete source evidence card
p07-technology-evidence-card-v1 · P07-M07-SBD-v1 · own static snapshot
FieldRecorded evidence
Identity / ownerP07-M07-SBD-v1; Satellite-service + timing owners
Inherited immutable IDsP07-M06-GNSS-TIMING-v1 · P07-M06-SBD-FALLBACK-v1 · P07-M05-CELLULAR-OPTIONS-v1 · P07-M04-REMOTE-FLEET-v1 · P07-M03-HE-BACKHAUL-v1 · P07-M02-LE-CONN-v1 · P07-M02-THREAD-SED-v1 · P07-M01-LE-EXT-v1 · p06-evidence-map-v1 · M01-A · M01-B-LOSS · M01-INSTALL-UNKNOWN
Scenario / requirementp07-m07-common-basis-v1: fictional 1000-node condition-monitoring fleet; 32 D0 bytes every 600 s. Fixed nodes, occasional maintenance repositioning; installed metal enclosure/open-pole strata and orientations must be sampled. Population weights, channel distributions and mobility transitions unknown. ≥.99 unique complete reports by 2 s; ≤.200 J radio-supply energy per generated report, including failures; ≥20 Mbit/s D0 gateway service where applicable.
Family / selected versionIridium SBD fallback; separate receive-only GNSS timing role. Iridium Short Burst Data service identity; terminal, firmware, SBD interface and acknowledgement mode unknown. GPS L1 C/A timing uses IS-GPS-200N + IRNs, including unresolved IRN-004 impact; timing is not messaging.
Region / classCH and remote-country candidates; transmit band, antenna, device class, service provisioning and authorization unknown.
D0 / deadline256 bits; 600 s period; 2 s deadline. Node application release → nominated gateway/service application D0; include discovery, queues, retries, network and host delivery. A PHY decode at R3 is not D0.
PHY / MAC resources32 D0 bytes; satellite framing, queues, service windows and terminal resources unknown. No local gateway backhaul role in direct terminal-to-host architecture.
State duration / currentAcquisition / search / TX / RX / retries / sleep: unknown, unknown
Antenna / RF interfacesR1 component port; R2 antenna feed; S0 installed spatial population; R3 receiver decision. Record R2↔R1 feed/filter loss once. Antenna pattern, detuning, volume, ground and mutual coupling remain unknown.
Infrastructure / ownershipOperator service, sky visibility, terminal supply/antenna and host delivery needed; GNSS additionally needs calibrated timing reference and bias treatment.
QualificationExact product/SKU/firmware/profile listing, host configuration and concurrent-operation test coverage unknown; individual qualification is not product coexistence evidence.
Regulatory questionsCH and EU/EEA proposed, deployment countries not frozen. Band, device class, antenna, installation, national applicability and evidence owner unresolved; no authorization inferred.
Claim-level provenance
Uncertainty / decisionThese source cards do not supply the synthetic A–D acceptance numbers. Delivery, installation distribution and concurrent state evidence remain unknown. D remains inspect; possible hybrid fallback under a separately approved 900 s service requirement. GNSS may serve gateway timing independently.
Next evidence / reviewService owner: terminal state/host timestamps; timing owner: calibrated PPS and placement strata. Due 2026-12-07. Frozen review basis 2026-09-08 UTC; recheck at due date or mode, firmware, antenna, operator, market or requirement change.
Review artifact · write the evidence, not a vendor winner
RecordWorked starting point / required learner addition
Identity / common basistechnology-coexistence-decision-studio/2.0; p07-m07-common-basis-v1; scenario, owner, frozen 2026-09-08 UTC and review 2026-12-07. Preserve inherited IDs and create variants for changed requirements.
State / RF / service reconciliationInclude at least three source cards, complete radio energy and D0 deadline boundaries, state/channel populations, infrastructure and unresolved distributions.
Threats / concurrencyRetain victim–aggressor–path IDs, on-time RF and leakage screens, explicit permitted/prohibited/unknown states, guard and external limitations.
Hard screen / preferencesAt default .6: A/B feasible in fixture; C rejected for delivery; D inspect. Record all reasons, fixed value functions, weight rationale, exact tie and interval reversals.
Conditional / hybrid / deferred portfolioWorked response: retain B conditionally at .6 and A as an integration-effort alternative above 2/3. GNSS timing may remain an independent gateway role. Delayed SBD can be a separate fallback only under its own service requirement. No real release is ready.
Two rejected alternativesReject C’s .90 hypothesis for the .99 deadline requirement. Reject unscheduled A at I=45 dB for −10 dB blocker headroom and excessive noise rise; distinguish that operating variant from protected A. Defer D because evidence is missing.
Dissent / reversal evidenceDocument a colleague’s different preference or infrastructure assumption. A 52/58 dB isolation observation, search-state energy or a matched LPWAN deadline distribution can change the supported set.
Experiment / next reviewChoose one E01/E02/E03 with owner, mode, planes, statistic, calibration/uncertainty and action for both outcomes. Systems owner reviews by 2026-12-07 or earlier on relevant change.
Explanatory curriculum review emphasis · no learner score or certification
EmphasisWhat a reviewer should be able to follow
Source/version discipline · 20%Identity, accessed text, selected mode, current status, unknown applicability and review trigger.
Shared-scenario RF reasoning · 25%Compatible planes, complete common basis, conditioned installation/range claims and correct RF arithmetic.
State/energy/capacity · 20%Failures, search and sleep included; PHY/service/deadline distinctions retained.
Coexistence diagnosis · 20%Distinct mechanisms and paths; honest internal/external authority and mitigation costs.
Uncertainty/experiment quality · 15%Reversals, correlated scenarios, calibration limits, dissent and discriminating outcomes.

A valid alternative review can choose protected local operation, an operator-dependent portfolio, or defer all options. Its quality lies in the traceable conditions and evidence. No rubric scoring, saved progress, learner pass/fail state or certification is implemented.

Handoff by evidence owner

Path 08 · RF Measurement — planned: execute the calibrated coupling/current/deadline experiment. Path 09 · Compliance & Certification — planned: resolve actual market, host and legal applicability. Path 10 · Practical RF Design — planned: implement the selected modes, antenna/supply arrangement and control contract. These are future learning paths, not available route links.

Ungraded review

Check your understanding

Answer each question in your own words, then reveal the model answer.

  1. 01Can the 60 s LE energy and a 900 s satellite deadline be compared directly?
    Model answer

    No. Recompute LE residual sleep at 600 s: add 3 V × 2 µA × 540 s = 3.24 mJ, giving 3.8709038808 mJ/generated report. Keep the same D0 boundary and count failed reports. The 900 s satellite fixture requires new ≤2 s evidence, not a unit conversion. Actual installation, coverage and state distributions remain unknown.

  2. 02What are coupled level and blocker headroom at +20 dBm and 45 dB isolation?
    Model answer

    At R1-RX, Pcoupled=20−45=−25 dBm. With the illustrative maximum −35 dBm, H=−35−(−25)=−10 dB. Isolation must reach 55 dB for inclusive H≥0. The 82 dB integrated factor belongs only to the separate leakage screen.

  3. 03How do deterministic overlap and independent random duty differ?
    Model answer

    TX [0,20) and RX [10,50) ms intersect for 10 ms: .10 of the cycle and .25 of listening. Independent duties .2 and .4 instead give .08 joint wall time, and .2 conditional on listening. None is a packet error probability; duty averaging cannot repair the on-time −10 dB headroom.

  4. 04Why are C and D both excluded from scoring, for different reasons?
    Model answer

    C has known .90 delivery below the chosen .99 threshold: fail, even though its naive utility score is 1. D lacks required energy/delivery/RF evidence: inspect, with no score. Its direct terminal-to-host architecture justifies not-applicable local gateway service; that is neither zero nor missing evidence. All fail/inspect means no supported choice.

  5. 05At exactly which weight do A and B exchange preference?
    Model answer

    A=.25+.50w; B=.75−.25w. Equality gives .75w=.50, so w=2/3 and both score 7/12. B leads below, A above; the exact tie retains both, ordered by stable ID for navigation only. Preferences apply after every hard screen.

  6. 06Which observation distinguishes the simultaneous choices at w=.8, and what changes at .6?
    Model answer

    A calibrated A-WIFI-NODE measurement resolving 52 versus 58 dB isolation tests the 55 dB boundary. At w=.8, low isolation rejects A and selects B; high isolation can admit A and select it. At .6 B already leads, so the measurement changes eligibility but not the immediate selected portfolio. Search-state and LPWAN deadline experiments may still retire separate hard assumptions.

References and further study

Primary sources actually accessed 2026-09-08; review due 2026-12-07, sooner on a relevant change. Identity-only and unavailable text are labelled. Synthetic calculations are Derived/Illustrative, never Measured. Values use SI internally; displayed rounding does not decide status. Verification tolerances: scores/overlap fractions 10⁻¹⁰, RF anchors 10⁻⁸ dB and schedule durations 10⁻⁹ ms; IDs, states and ordering exact.

  1. NASA · NASA. Systems Engineering Handbook · Crosscutting Technical Management. Online chapter, §6.8.1 and §6.8.1.2.1–7. Page publication date not established. Informative decision process; text read. Mandatory criteria, alternatives, uncertainty and decision records. Does not prescribe this lesson’s utilities, weights or experiment priority.
  2. BT-INDEX · Bluetooth SIG. Specifications catalogue. Live index observed 2026-09-08. Core 6.3 and 6.2 listed as adopted at access. Identity/status index read. Selected Core 6.2 is not a claim to newest version or actual product support.
  3. BT-62 · Bluetooth SIG. Core Specification 6.2. Selected 6.2; Vol 6 Part A §4.3 blocking text. Selected release 2025-11-03; update 28108 notice present. Normative text read for test scope only. Blocking uses a specified interferer, wanted signal and receiver criterion. The lesson’s −35 dBm proxy and 82 dB factor are unrelated synthetic values. Update 28108 impact not read/resolved.
  4. BT-UPDATE · Bluetooth SIG. Core 6.2 adoption and mandatory update notice. Core 6.2 landing page; update 28108 notice. Status checked at access. Status notice read; update body unavailable in this review. Flags a dependency review; does not prove conformity to the update.
  5. BT-QUAL · Bluetooth SIG. Qualify your product. Public qualification process page. Undated page, accessed 2026-09-08. Informative process text read. Qualification and product identity need their own evidence. No actual product listing or concurrency coverage verified.
  6. IEEE-INDEX · IEEE 802.11. Official project timelines. 802.11 working-group timeline. Observed 2026-09-08. Project/status register read. Base/amendment publication and project status, not deployed features or a normative PHY lookup.
  7. IEEE-AX · IEEE. IEEE 802.11ax-2021. Selected legacy HE amendment; catalogue marks superseded. Published 2021-05-19. Catalogue identity/abstract read; full standard not accessed here. HE scope only. The disclosed inherited resource arithmetic is a teaching calculation; no inaccessible normative table is reproduced.
  8. WFA · Wi-Fi Alliance. Public certification certificate · variant 129909. Broadcom BCM94916R6726W7; firmware 17.10.373.9. Certificate dated 2024-01-05. Four-page public product certificate read. Example of product/firmware/feature scope. Not evidence that the fictional HE gateway, concurrent combination or country is approved.
  9. LORA-L2 · LoRa Alliance. TS001-1.0.4 LoRaWAN L2. 1.0.4; downloaded PDF §2.1 and §3.3, pp.9–13. PDF ©2020; catalogue entry 2023-09-28. Normative class/receive-window text read. Class A receive opportunities follow uplinks; this does not establish urgent-command or unique-report deadline delivery.
  10. LORA-RP · LoRa Alliance. RP002-1.0.5 Regional Parameters. 1.0.5 entry; catalogue also exposes D2D errata. 2025-10-08. Catalogue identity read; PDF retrieval failed in this review. Regional-mode identity only; detailed mappings, errata impact and national applicability remain unresolved.
  11. SEMTECH · Semtech. LoRa calculator. Public calculator interface. Undated interface accessed 2026-09-08. Manufacturer interface read; not independently run. No claim of calculator agreement. SX1276 data-sheet URL retrieval failed here; airtime assumptions and independent algebra are disclosed in RECON.
  12. 3GPP · 3GPP / ETSI. TR 21.900 · Technical Specification Group working methods. V19.1.0, Release 19; §6.4.1–6.4.3. 2026-01. Informative working-method text read. Stages and specification change/freeze distinctions. No LTE-M RF thresholds, current module listing or operator deployment inferred.
  13. PTCRB · PTCRB. Get certified. Public certification process page. Undated page accessed 2026-09-08. Process text read. Operator-specific testing and government obligations can be additional. No member criteria or actual host/module approval checked.
  14. GCF · Global Certification Forum. 3GPP wireless devices. Public overview; displayed feature table explicitly January 2024. Accessed 2026-09-08; feature snapshot is older. Public overview read; member criteria unavailable. Certification scope orientation only. Table freshness cannot establish current device support.
  15. GPS-INDEX · GPS.gov. Interface Control Documents and Interface Specifications. IS-GPS-200N register; IRN-003 and IRN-004. IRN-003 2024-01-19; IRN-004 2026-06-16. Official register read; IRN-004 body not read. Version/status identity and open change review; no inference that an existing receiver implements the update.
  16. GPS-SPS · U.S. Department of Defense / GPS.gov. GPS Standard Positioning Service Performance Standard. 5th edition; §3.4 and Appendix C scope. 2020-04. Authoritative performance-scope text read. Signal-in-space commitments exclude important user-side effects; DOP and a strong signal do not prove installed timing accuracy.
  17. SBD · Iridium. Iridium Short Burst Data. Public service description. Undated page accessed 2026-09-08. Informative service description read. Messaging service identity only; terminal energy, two-second delivery and deployment authorization unknown.
  18. SILABS · Silicon Labs. UG103.17 · Wi-Fi Coexistence Fundamentals. Rev 0.4; §4.1–4.2.2 and revision history. June 2025; explicitly deprecated in revision history. Legacy manufacturer explanation read. Request/grant timing and separation illustrate internal control. Deprecated source is not a current implementation API or evidence for the chosen 2 ms recovery guard.
  19. ADI · Analog Devices. Wideband LO Noise in Passive Transmit/Receive Mixer ICs. Technical article; reciprocal-mixing discussion. Page publication date not established. Primary receiver explanation read. A blocker can mix with LO noise into the wanted band. Magnitude requires LO spectrum, bandwidth, offset and front-end evidence absent from V82.
  20. RECON · RF Journey. Module 07.7 source/model ledger and independent oracle. technology-coexistence-decision-studio/2.0; common-basis/threats/decision/rules v1. Authored and checked 2026-09-08. Definition, Derived and Illustrative; never Measured. Chosen values, compatible prior mathematics, reconciliation, analytic expectations and tolerances. No certification, legal approval or guaranteed product performance.

The local source/model ledger records the chosen-number origin, independent equations and actual access limits. Browser rendering makes no live standards or network queries and cannot silently update the evidence date.