Skip to content
Jae Hoon Kim
Projects

EchoCast · drawing set · v1.0

Part I · Cover

U.S. Provisional Patent Application · Drawing set
Appl. No.
Sheets 15
Rev. B · v1.0
Title · System and method for TDMA-scheduled ultrasonic device-presence inference with through-body absorption detection, applied to migration of an audio rendering session between co-located devices
Inventor · J. H. Kim
Internal · EchoCast
Filing target · 2026-09 · pre-public-disclosure
Classification (proposed) · H04R 27/00 · H04L 12/28
About this Personal project · ongoing v1.0
A personal project I've been turning over for a while. Drafted in the shape of a provisional patent application because the format is a useful thinking tool — it forces the claims, alternatives, prior art, and threat model out into the open where I can see the gaps. Treat it as a working draft of an idea I keep iterating on. v1.0 honest correction (2026-05-18): the headline product behavior — emit ultrasonic chirps, compare signal strength, hand off audio to the nearer speaker — was publicly launched as Sonos Sound Swap in April 2021 (see Sonos tech blog). Earlier revisions of this page treated the headline mechanic as the contribution; v1.0 demotes that to acknowledged prior art and narrows the claim to the genuinely-surviving differentiators: TDMA-scheduled chirps (vs Sound Swap's CDMA), through-body absorption detection (FIG. 9), and the continuous mutual-attenuation matrix for N-way proximity (vs Sound Swap's binary same-room test).
Abstract of the disclosure Cover-sheet boilerplate v0
A system and method coordinate migration of an audio rendering session among a plurality of co-located audio-rendering devices using a time-division-multiple-access (TDMA) ultrasonic device-presence protocol with a through-body absorption-detection step. Each device emits, in a calibration-negotiated TDMA slot, an inaudible ultrasonic chirp encoding a unique device identifier; each device receives chirps from other devices via a microphone array and computes a mutual-attenuation matrix M whose entries record pairwise chirp RSSI. From M, the system infers a continuous-valued proximity vector across N devices and detects the presence of a human body absorber by its characteristic 8 – 14 dB through-body attenuation signature, distinguishable from wall-only attenuation. Upon a sustained change in the proximity vector subject to hysteresis, the active rendering device performs a synchronized cross-fade handoff to the new most-proximate device. The disclosed contribution is the combination of (i) the TDMA schedule, (ii) the human-absorber-distinguishing through-body model, and (iii) the matrix-valued proximity inference; the underlying product behavior of ultrasonic-driven inter-speaker audio handoff is acknowledged anticipated by Sonos Sound Swap (publicly deployed April 2021) and adjacent commercial systems. No IP-network metadata identifying the user, content, or location is transmitted off-host.
Field
H04R 27/00 · audio sys.
Acknowledged prior art (headline mechanic)
  • Sonos Sound Swap (Apr 2021) — chirp + RSSI + swap
  • US9024998B2 (Polycom/HP) — ultrasonic beacon + auto-pair on room change
  • US10932062B2 (Apple) — ultrasonic proximity for AirPods handoff
Distinguished from (surviving novelty)
  • Sound Swap uses CDMA · this discloses TDMA
  • Sound Swap tests binary same-room · this matrix-infers N-way
  • Sound Swap models wall α only · this adds through-body α
  • Sound Swap has no published threat model · FIG. 15
Index of sheets Tap a row to jump 15 sheets
Drawing convention Symbol vocabulary across all figures
Line styles
audio data flow · active rendering
control · wire · outline
chirp / control signal
boundary · enclosing scope
perturbed / attenuated path
Fill patterns
out-of-scope · cloud · absent
active rendering · blocking event
acoustic shadow · attenuation region
Acoustic glyphs
chirp emission · concentric arcs from transducer
up-chirp · bit = 1 (FIG. 3 · 306)
down-chirp · bit = 0
preamble / sync pulse · fixed-frequency tone
blocked / forbidden path (FIG. 7, FIG. 15)
active / current device (FIG. 1, FIG. 6)
Convention applies to FIG. 1 – FIG. 15. · Reference numerals follow 100-series (FIG. 1), 200-series (FIG. 2), etc. per USPTO practice.
Drawing compliance certificate 37 CFR § 1.84 Self-certification

The applicant certifies that the drawings filed herewith satisfy 37 CFR § 1.84:

  • § 1.84(a)(1) — black-and-white line art at uniform stroke density.
  • § 1.84(d) — every sheet identified by its serial position ("Sheet 1 / 15" … "Sheet 15 / 15") in the bezel.
  • § 1.84(m) — established drafting convention for graphic forms; documented in the Drawing Convention block above.
  • § 1.84(p)(5) — same reference numeral consistently designates the same part across all drawings.
  • § 1.84(u) — Sheet 1 (FIG. 1) designated as the Representative Figure for the application cover.

Part II · Drawings

Sheet 1 / 15 Representative FIG. 1 · System overview 100
PLAN VIEW · 114 KITCHEN LIVING ROOM BEDROOM 102 DEVICE A 104 DEVICE B ★ ACTIVE 106 DEVICE C · IDLE 108 USER 110 · user trajectory 112 · audio handoff (A → B) ≈ 12 m · household scale LEGEND · active rendering idle chirp field user path
FIG. 1
Sheet 2 / 15 FIG. 2 · Functional block diagram 200
200 · one audio-rendering device SPLIT-PATH ARCHITECTURE RENDER PATH · 48 kHz PCM AUDIO IN ROUTER FADER DRIVER 202 204 206 208 210 SPK x[n] x'[n] g·x y(t) SENSE / CONTROL PATH · 96 kHz capture MIC ARRAY DECODER INFER. HANDOFF 212 214 216 218 220 222 · control signal CHIRP EMIT 224 u(t) · ultrasonic out via main driver 226 · FREQUENCY & CLOCK DOMAINS 20 Hz ─ 20 kHz · audible 20–23 kHz · u(t) RENDER fs = 48 kHz · SENSE fs = 96 kHz (Nyquist = 48k > 23k) KEY ─ audio data control / chirp device boundary
FIG. 2
Sheet 3 / 15 FIG. 3 · Ultrasonic chirp encoding 300
HUMAN-AUDIBLE · 20 Hz – 18 kHz (out-of-channel) 18 kHz · audible threshold 18 20 22 24 kHz 050100150200 ms 302 · PRE 1 0 1 1 0 1 0 0 304 · CRC 306 · 8-BIT DEVICE-ID PAYLOAD (256 unique devices) 308 checksum byte KEY · / bit = 1 · \ bit = 0 · ▮ preamble · ▭ CRC
FIG. 3
Sheet 4 / 15 FIG. 4 · Audio-session state machine 400
402 IDLE silent · listening 404 RECEIVING fade-in 1.5 s 406 SOLO ★ rendering 408 LISTENING peer detected 410 HANDING-OFF fade-out 1.5 s incoming chirp · 222 fade-in done peer chirp ≥ τ peer lost Δ presence ≥ Δ_min fade-out done τ = peer-chirp threshold · Δ_min = handoff hysteresis · emphasized ring = current state on FIG. 1
FIG. 4
Sheet 5 / 15 FIG. 5 · Hardware embodiment (cross-section) 500
502 504 508 M1 M2 M3 506 510 · USB-C / power 512 · ultrasonic + audio out ≈ 18 cm 510a 516 · 502 FREQUENCY RESPONSE 20 100 1k 10k 24k 40k Hz 0 -3 -10 -20 dB SPL 24 kHz claim min CHIRP · FIG. 3 518 · PARTS LIST 502 tweeter · 20 Hz–24 kHz 504 main driver · woofer 506 3-mic array · DoA 508 PCB · ESP32-S3 class 510 USB-C / power 512 ultrasonic+audio out 510a ≈ 18 cm height 514 · Embodiment requires a tweeter capable of clean reproduction to 24 kHz; commodity hardware available at < $30 BOM.
FIG. 5
Sheet 6 / 15 FIG. 6 · Multi-device mesh topology 600
A KITCHEN B LIVING C BEDROOM D OFFICE E TV / DEN −42 dB −48 dB −54 dB −52 dB 612 USER PRESENCE VECTOR · 614 · argmax → B Δ_min A 0.18 B 0.62 C 0.06 D 0.04 E 0.10 MESH TOPOLOGY · 600 · N = 5 DEVICES RSSI · CHIRP-MUTUAL KEY active mutual chirp lock weak / non-locked
FIG. 6
Sheet 7 / 15 FIG. 7 · Privacy boundary / data-flow 700
IN-HOME ACOUSTIC · 706 CLOUD / IP NETWORK · 708 (FORBIDDEN) 702 · in-air ultrasonic only DEVICE IDs 710 PRESENCE V. 712 RENDER STATE 714 ephemeral · ≤ 200 ms TTL in-memory · TTL = window until session change 704 · off-host services 718 · POST /v1/audio.upload — NEVER SENT audio_pcm · <1.2 MB · 16-bit · 48 kHz> geo · [37.3287, -122.0265, room-3] user_id · "u_3f9a-2dab-…" ip_header · src 10.0.1.42 → dst … mic_recording · raw_ch[3] · <PCM> session_log · device→device timestamps telemetry · dwell_ms, ui_state, device_id status: NO ROUTE — boundary 706↔708 disallows egress 722 · RETENTION (TTL) 710 200 ms (in-air ephemeral) 712 = window (~1.6 s) 714 = session (until handoff) 0 200 ms window 10 s session 716 · NO EGRESS 720 · Claim scope: device-IDs stay in-air; no audio, location, or user-ID crosses boundary 706 ↔ 708.
FIG. 7
Sheet 8 / 15 FIG. 8 · Handoff timing diagram 800
802 A · kitchen 804 B · living 806 C · bedroom 808 active source 810 cross-fade A (rendering) cross-fade B (rendering) 814 · handoff event 0246810 t (s) CHIRP RSSI (top three lanes) · envelope linear · bottom lane = post-fader audio level
FIG. 8
Sheet 9 / 15 FIG. 9 · In-air ultrasonic propagation physics 900
PLAN VIEW · 902 · ACOUSTIC WAVE PROPAGATION 904 TX · device A 906 USER · absorber 908 · acoustic shadow ≈ 8 – 14 dB attenuation @ 22 kHz 910 RX · device B 912 · direct path (through-body, attenuated) 914 · reflected path (ceiling bounce) 916 · reflected path (floor bounce) DRYWALL · α ≈ 0.05 @ 20 kHz CARPET · α ≈ 0.7 @ 20 kHz CEILING · α ≈ 0.4 @ 20 kHz 918 · Through-body path 912 is attenuated 8 – 14 dB vs reflected paths 914 / 916; the asymmetry is the user-proximity signal. 920 · Inference is RSSI-only — no time-of-flight, no phase measurement, no audible audio content required.
FIG. 9
Sheet 10 / 15 FIG. 10 · One-shot calibration flow 1000
ONE-SHOT CALIBRATION · ~ 90 s · ONCE PER INSTALLATION FIRST POWER-ON C1002 DISCOVER · broadcast chirp + listen for peer chirps for 10 s C1004 measure mutual-RSSI matrix between all peers (silence baseline) C1006 PROMPT · walk past each device once, ≈ 60 s total C1008 record per-device through-body attenuation A_ij(user-present) C1010 compute presence kernel from ΔRSSI = A_ij(user) − A_ij(baseline) C1012 fit τ and Δ_min thresholds for state machine FIG. 4 C1014 store calibration profile on every device · mesh ready re-fit on drift 1016 · Calibration profile feeds state-machine 400 of FIG. 4 and is updated continuously on detected RSSI drift.
FIG. 10
Sheet 11 / 15 FIG. 11 · Hardware variant taxonomy 1100
HARDWARE TAXONOMY · 1100 · FIVE EMBODIMENTS · ALL PARTICIPATE IN MESH 600 1102 SMART SPEAKER FIG. 5 reference all roles · TX + RX 1104 SOUNDBAR tweeter array all roles · TX + RX 1106 PHONE / TABLET USER ASSOCIATED TX + RX 1108 EARBUDS RX dominant in-ear bone-conduction TX 1110 AUTOMOTIVE head unit · in-cabin all roles · TX + RX CAPABILITY MATRIX · 1112 TX RX RENDER USER-ANCHOR · · · 1114 · Claim 1 reads on any device meeting the TX + RX + RENDER triple. USER-ANCHOR (★) devices serve as proxy for the user in the presence vector of FIG. 6.
FIG. 11
Sheet 12 / 15 FIG. 12 · Ultrasonic spectrum interference map 1200
SPECTRUM ALLOCATION · 1200 · EchoCast OPERATING BAND vs OTHER SOURCES scale break 1520253035405080 kHz 18 kHz · AUDIBLE THRESHOLD EchoCast band this disclosure 1202 Dog whistles 16 – 22 kHz 1204 Hearing-aid feedback incidental 1206 Data-over-sound SilverPush · Chirp · Lisnr 1208 Pest deterrents 25 – 45 kHz 1210 Auto parking sensors ≈ 40 kHz narrowband 1212 Bat detectors 20 – 200 kHz 1214 → 200 EchoCast band incidental · historical 1216 · Conflict with 1204 mitigated by preferential operation at 20–22 kHz; CRC rejects non-EchoCast payloads.
FIG. 12
Sheet 13 / 15 FIG. 13 · Chirp protocol bit-level specification 1300
FRAME SPEC · 1300 · 192 ms TOTAL · 17 BYTES EFFECTIVE PAYLOAD 04080120160200 ms 1302 PRE 19 kHz tone S 1304 · SYNC · Barker-4 1306 DEVICE ID PAYLOAD 8 bits × 16 ms · BFSK 1308 CRC 8-bit CRC-8 ATM G 1310 · GUARD · quiet DEVICE ID PAYLOAD · 1306 · EXPLODED · 16 ms / bit b7 up-chirp 1 b6 down-chirp 0 b5 up-chirp 1 b4 up-chirp 1 b3 down-chirp 0 b2 up-chirp 1 b1 down-chirp 0 b0 down-chirp 0 EXAMPLE PAYLOAD · 0b10110100 = 0xB4 = DEVICE ID 180 1312 · TOTAL = 16 + 4 + 128 + 32 + 12 = 192 ms · TRANSMITTED ONCE PER 1 s SCHEDULE PER DEVICE
FIG. 13
Sheet 14 / 15 FIG. 14 · TDMA chirp schedule 1400
TDMA SCHEDULE · 1400 · 5 DEVICES SHARE 1-SECOND EPOCH WITHOUT COLLISION 02004006008001000 ms A 1402 A · 192 ms B 1404 B · 192 ms C 1406 C · 192 ms D 1408 D · 192 ms E 1410 E · 192 ms EPOCH T = 1 s 1412 · total airtime utilization · 5 × 192 ms = 960 ms · 4 % headroom in epoch 1414 · Phase offsets negotiated at calibration (FIG. 10 · C1002). Up to 5 devices share one epoch; meshes > 5 use 2-second epochs.
FIG. 14
Sheet 15 / 15 FIG. 15 · Adversarial threat model 1500
THREAT MODEL · 1500 · ADVERSARY POSITIONS vs PROTOCOL MITIGATIONS 1502 · household acoustic envelope A B C 1504 · LEGITIMATE MESH USER 1506 · adversary positions 1508 · PASSIVE EAVESDROPPER captures chirps via mic ↳ obtains device IDs only no audio content 1510 · ACTIVE INJECTION emits forged chirps ↳ CRC-8 (1308) rejects malformed failures logged 1512 · REPLAY ATTACK re-emits captured chirps ↳ sliding-window epoch hash + per-epoch nonce 1514 · DENIAL-OF-SERVICE saturates 18 – 24 kHz band ↳ frequency-hop to inner band 20 – 22 kHz fallback CROSSING THE ENVELOPE 1516 · Defender invariants — no audio content, location, or user ID transmitted in chirps; — eavesdropping discloses only device-mesh size + topology; — each adversary class has an explicit mitigation in 1508 – 1514.
FIG. 15

Part III · Specification

Object of the invention Pre-AIA convention

It is therefore an object of the present invention to provide a system and method for migrating an audio rendering session between co-located audio-rendering devices, while simultaneously:

  1. Operating wholly within the household acoustic envelope, with no IP-network metadata exchange identifying user, content, or device (per claim 1 final clause and FIG. 7);
  2. Requiring no explicit user pairing, no cloud account, and no manual zone configuration;
  3. Inferring user proximity from in-air ultrasonic signal strength rather than from network signaling (per FIG. 6, FIG. 9);
  4. Achieving handoff latency below the perceptual continuity threshold for audio rendering (per FIG. 8 · ≈ 5 s end-to-end, ≈ 3 s cross-fade);
  5. Sharing the transmission band among up to N devices via collision-free TDMA scheduling (per FIG. 14);
  6. Operating equivalently across hardware embodiments — smart speakers, soundbars, phones, earbuds, and automotive head units (per FIG. 11) — without modification of the independent claim; and
  7. Resisting the four canonical attack classes catalogued in FIG. 15 (eavesdropping, injection, replay, denial-of-service) via protocol-level mitigations rather than network-layer authentication.
Background of invention Prior-art context

Existing multi-device audio systems (Apple AirPlay 2, Sonos S2, Google Cast) coordinate via IP-network metadata exchange, depend on cloud-resident control planes, and infer user proximity either via explicit user toggling or via inference from Wi-Fi or Bluetooth signal strength.

The closest prior art to the headline product behavior of this disclosure is Sonos Sound Swap, publicly deployed in April 2021 on Sonos Roam and rolled out to subsequent portable Sonos speakers. Sound Swap, as described in Sonos's own tech-blog disclosure, emits a near-ultrasonic chirp (> 19 kHz, m-FSK modulated, code-division multiple access for multi-device coexistence) from a holding device, compares received signal strength at nearby Sonos speakers, exploits high-frequency wall-attenuation as a same-room presence heuristic at a 3-meter target range, and migrates the audio session to the closest speaker. Earlier revisions of this page treated this product behavior as the disclosed contribution; v1.0 corrects that — the headline mechanic is conceded prior art and the present claim narrows to features absent from Sound Swap's published disclosure.

Patent prior art on the underlying primitive is also extensive. US7,429,108 not applicable here; the on-point reference is US9,024,998B2 (Polycom, 2015; current assignee HPE) which discloses a videoconferencing endpoint emitting an ultrasonic beacon encoding a device identifier, a paired device detecting and decoding said beacon, and automatic disconnect when the device leaves the room with automatic reconnection upon entering another room. Elements (a)–(c) of as-filed Claim 1 read directly on Polycom; only the "audio rendering session" subject matter differs. US10,932,062B2 (Apple, 2021) discloses ultrasonic proximity for audio handoff between AirPods and host devices. The Chirp.io / Lisnr / SilverPush lineage of in-air data-over-sound was previously cited (and is retained) but is no longer the closest art.

The disclosed contribution survives the foregoing as a combination: (i) a time-division-multiple-access (TDMA) chirp schedule (FIG. 14) — distinguishable from Sound Swap's code-division (CDMA) multiple access; (ii) a human-absorber-distinguishing through-body attenuation model (FIG. 9) at characteristic 8 – 14 dB — absent from Sound Swap's published material, which models wall attenuation only; (iii) a matrix-valued mutual-attenuation inference producing a continuous proximity vector across N devices (FIG. 6) — distinct from Sound Swap's binary same-room test; and (iv) a published adversarial threat model (FIG. 15) with concrete protocol-level mitigations — absent from any prior-art commercial deployment. Items (i)–(iv) define the surviving claim space and supply the structure of as-filed Claim 1 below.

Summary of the invention per 37 CFR § 1.73

The disclosure provides a method and system in which a plurality of audio-rendering devices in a physical environment continuously emit and listen for inaudible ultrasonic chirps encoding unique device identifiers. From the mutual signal-strength pattern observed across the device mesh, each device infers a presence vector indicating which device is most-proximate to the target user. An active rendering device, upon detecting a sustained change in said presence vector, performs a synchronized cross-fade audio handoff to the new most-proximate device.

The key novelty is the substitution of in-air ultrasonic signaling for IP-network metadata exchange. No audio content of music or speech is transmitted between devices; no IP-network packets identifying user, content, or location are emitted to any external service. The disclosed system thus operates wholly within the in-air acoustic envelope of the household, eliminating multiple classes of metadata leakage characteristic of prior-art multi-device audio systems.

The ultrasonic protocol uses a fixed-rate preamble pulse (per FIG. 3 · 302) followed by 8-bit device-ID payload encoded as up-/down-chirp bits (306) and an 8-bit CRC (304); the full protocol fits within a 200 ms transmission window and addresses up to 256 unique devices per household. Handoff is governed by a per-device state machine (FIG. 4) with hysteresis on the presence vector to prevent oscillation in the boundary region between two devices.

Brief description of drawings Sheets 1 – 15
Detailed description of embodiments ¶ [0001] – [0010]

[0001]The present disclosure relates to multi-device audio coordination, and more particularly to a method and system for migrating an audio rendering session between co-located audio-rendering devices using inaudible ultrasonic device-presence signals.

[0002]Referring to FIG. 1, environment 100 comprises three audio-rendering devices 102, 104, 106 disposed in distinct rooms of a household 114. A user 108 moves along trajectory 110 from a position proximate device 102 to a position proximate device 104. An audio session, initially rendered by device 102, is handed off to device 104 along path 112 as a function of the changing relative-proximity vector.

[0003]FIG. 2 depicts the internal architecture of an audio-rendering device 200. The device comprises two coupled signal paths: a render path comprising audio-input stage 202, router 204, fader 206, driver-stage amplifier 208, and acoustic transducer 210; and a sense-and-control path comprising microphone array 212, chirp decoder 214, presence-inference module 216, handoff coordinator 220, and chirp emitter 224. The chirp emitter 224 multiplexes its ultrasonic output onto the same acoustic transducer 210 used for audible audio rendering, exploiting the broadband response of contemporary tweeters.

[0004]FIG. 3 illustrates the ultrasonic chirp protocol. A 200 ms transmission window begins with a fixed-frequency preamble pulse 302 at ≈ 19 kHz, followed by an 8-bit device-ID payload 306 in which each bit is encoded as either an up-chirp (bit = 1) or a down-chirp (bit = 0) over the 18 – 24 kHz band, and concludes with an 8-bit CRC 304. The protocol thereby addresses 2⁸ = 256 unique devices per household while remaining wholly above the human-audible threshold at 18 kHz.

[0005]FIG. 4 sets forth the audio-session state machine 400. Each device occupies one of five states. Transitions are gated by the chirp-RSSI threshold τ and by a hysteresis parameter Δ_min that prevents handoff oscillation when the user is positioned in the boundary region between two devices.

[0006]FIG. 5 depicts a preferred hardware embodiment comprising a tweeter 502 capable of clean reproduction to 24 kHz; a main driver 504 for audible audio; a three-microphone array 506 mounted around the perimeter of the enclosure (enabling direction-of-arrival estimation); and an edge controller 508 implementing the protocol stack of FIG. 2. Per claim hook 514, the device is buildable from commodity components at a bill-of-materials cost below $30.

[0007]FIG. 6 illustrates the mesh-topology view of a five-device installation. Edge weights between devices represent the time-averaged chirp-RSSI (in dB) of mutual chirp reception. The presence vector 614 is computed from the relative attenuation pattern observed across the mesh; the argmax of this vector identifies the rendering device most-proximate to the user.

[0008]FIG. 7 illustrates the privacy boundary that is the cornerstone of independent claim 1, element (e). Device IDs 710, presence vectors 712, and render-state metadata 714 reside wholly within the in-home acoustic envelope 706; no audio content, location, user identifier, or IP-network metadata crosses boundary 706 / 708 to off-host services 704.

[0009]FIG. 8 illustrates a representative handoff event at t ≈ 5 s. As the user moves from kitchen to living room, chirp-RSSI 802 from device A declines while chirp-RSSI 804 from device B rises; chirp-RSSI 806 from device C remains low throughout. At the handoff event 814 the active rendering source transitions from A to B over a cross-fade window 812 of approximately 3 s, with the cross-fade envelopes 810 governed by a constant-power crossfade law.

[0010]The foregoing description is illustrative; modifications, equivalents, and additional embodiments not departing from the spirit of the disclosure are contemplated. By way of non-limiting example, the chirp band may be re-located to 24 – 32 kHz for higher payload throughput; the presence-inference module 216 may employ machine-learning rather than purely RSSI-based inference; and the system may be embodied on phones, tablets, soundbars, automotive head units, and earbuds per FIG. 11.

[0011]Best mode. At the time of this filing the inventor contemplates as the best mode of practicing the invention a household installation of three smart-speaker embodiments per FIG. 5 — each comprising an Espressif ESP32-S3 controller, a 30 mm full-range driver rated for clean reproduction to 24 kHz, and a three-microphone perimeter array — communicating via the 8-bit chirp protocol of FIG. 3 over the 19 – 22 kHz sub-band. Audio rendering employs a constant-power crossfade per FIG. 8 over a 3-second handoff window. Calibration is performed once at first power-on per FIG. 10 and re-fit on detected RSSI drift. This best-mode statement is supplied in conformity with pre-AIA 35 U.S.C. § 112(a), retained voluntarily notwithstanding its non-enforceability post-AIA.

Part IV · Claims

Claims 1 independent · 4 dependent · 1 apparatus Draft
What is claimed is:

1. A method for migrating an audio rendering session between a plurality of audio-rendering devices in a physical environment, the headline mechanic of ultrasonic-driven audio handoff being acknowledged in the prior art, comprising:

  1. (a)emitting, from each of said plurality of devices (102, 104, 106), an inaudible ultrasonic acoustic signal encoding a unique device identifier (306) at a carrier frequency above 18 kHz, said emissions being scheduled according to a time-division-multiple-access (TDMA) frame (1400) negotiated at calibration time (FIG. 10 step C1002) such that no two of said plurality of devices emit a chirp in any one slot of said frame;
  2. (b)receiving, at each of said plurality of devices, ultrasonic acoustic signals emitted by other ones of said plurality of devices via said microphone array (506);
  3. (c)computing, at each of said plurality of devices, a mutual-attenuation matrix M whose entry Mij is the time-averaged received-signal-strength attenuation of device i's chirp as observed at device j, and inferring from M a continuous-valued proximity vector (614) across said plurality of devices;
  4. (d)detecting, from said mutual-attenuation matrix, a human-absorber event by identifying a characteristic 8 – 14 dB attenuation step on at least one Mij in a frequency-and-temporal pattern distinguishable from static wall-attenuation (FIG. 9, hook 918), and conditioning said proximity vector on said event;
  5. (e)selecting from said plurality of devices a rendering device as a function of changes in said proximity vector subject to hysteresis Δ_min; and
  6. (f)handing off the audio rendering session to said selected rendering device by a synchronized cross-fade (810).
Survives over: Sonos Sound Swap (Apr 2021, CDMA, binary same-room, wall-α only). Survives over US9,024,998B2 (Polycom — videoconference pairing, no audio rendering, no human-absorber model, no matrix). Acknowledged-anticipated steps retained as enabling-context (preamble + (a) chirp emit + (b) chirp receive + (f) cross-fade); novel limitations carried by the TDMA scheduling in (a), the matrix-and-vector inference in (c), and the through-body absorber-detection step in (d).

2. The method of claim 1, wherein said ultrasonic acoustic signal further comprises a fixed-frequency preamble pulse (302) and a cyclic redundancy check (304), and wherein the device identifier (306) is encoded across 8 bits each represented by an up-chirp or down-chirp in the 18 – 24 kHz band.

3. The method of claim 1, wherein each said device implements a finite-state machine (400) having at least the states IDLE (402), RECEIVING (404), SOLO (406), LISTENING (408), and HANDING-OFF (410).

4. The method of claim 1, wherein the chirp emitter (224) multiplexes its ultrasonic output onto the same acoustic transducer (210) that is used for audible audio rendering, with said ultrasonic component being substantially imperceptible to a human listener.

5. The method of claim 1, wherein no audio content, no user identifier, no location data, and no IP-network metadata identifying said devices or rendered content is transmitted across the local-host boundary 706/708.

6. An audio-rendering device for participating in a multi-device handoff mesh, comprising:

  1. (a)an acoustic transducer (210) capable of clean reproduction to 24 kHz;
  2. (b)a microphone array (506) of at least three microphones disposed about the perimeter of the device;
  3. (c)one or more processors implementing a chirp emitter (224), chirp decoder (214), presence-inference module (216), and handoff coordinator (220); and
  4. (d)a non-transitory memory storing instructions which, when executed by said processors, cause the device to perform the method of any of claims 1 – 5.
Claims · 6 total · 1 independent method · 4 dependent · 1 apparatus
Alternative claim sets For prosecution flexibility · not filed Draft

The following alternative drafts of independent Claim 1 are preserved for prosecution flexibility. They are not filed as part of the present application. If the as-filed Claim 1 is rejected on art or § 112 grounds, applicant may pursue the broader 1A (to capture additional infringing implementations) or the narrower 1B (to defend over an unforeseen reference such as the Chirp.io / Lisnr lineage).

Alternative 1A Broader · drops the IP-metadata-exclusion clause

1A. A method for migrating an audio rendering session between a plurality of audio-rendering devices in a physical environment, comprising:

  1. (a)emitting, from each of said plurality of devices, an inaudible acoustic signal encoding a unique device identifier;
  2. (b)computing, from received signals, an estimate of relative acoustic proximity among said devices and the user; and
  3. (c)handing off the audio rendering session to a device selected as a function of said estimate.
Rationale. Omits the "without recourse to IP-network metadata" limitation of as-filed Claim 1(e). Captures implementations that hybridize ultrasonic device discovery with IP-network audio transport (e.g., a Sonos-like architecture that adopts ultrasonic presence inference).
Alternative 1B Narrower · binds frequency band, frame structure, and rendering-layer coupling

1B. A method for migrating an audio rendering session, comprising:

  1. (a)each device of a plurality of audio-rendering devices emitting an inaudible ultrasonic acoustic signal within the band 18 – 24 kHz at a rate of at least one emission per second;
  2. (b)said signal comprising a preamble pulse, an n-bit device-identifier payload encoded as binary frequency-shift keying, and a cyclic redundancy check, all transmitted within a frame duration not exceeding 250 ms;
  3. (c)computing, at each device, a presence vector from mutual received-signal-strength estimates;
  4. (d)selecting from said plurality of devices a rendering device as a function of changes in said presence vector subject to hysteresis Δ_min;
  5. (e)handing off the audio rendering session to said selected device by a synchronized constant-power cross-fade of duration between 1 and 5 seconds; and
  6. (f)wherein no audio content of music or human speech is transmitted between said devices, and wherein device identifiers are exchanged solely via said ultrasonic signals without recourse to IP-network metadata identifying said devices, said rendered content, or the user.
Rationale. Substantially narrows the claim by binding the 18 – 24 kHz band, the specific BFSK + CRC frame structure (FIG. 13), the hysteresis-governed selection (FIG. 4), the constant-power crossfade duration (1 – 5 s per FIG. 8), and reinforces the no-IP-metadata clause. Defensive position if Claim 1 is rejected on an unforeseen Chirp.io / Lisnr variant.
Drafts retained in inventor's records only. · Filing of either alternative requires a continuation or amendment.

Part V · Appendices

Claims × figures · cross-reference matrix Which figure supports each claim element 6 × 15
Claim Type FIG. 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15
1 indep · method ········
2 dep ·············
3 dep ··············
4 dep ·············
5 dep ··············
6 apparatus ············
= direct support for claim element. · Absence of a dot indicates indirect support via the Specification rather than no support.
Performance specifications Embodiment of FIG. 5 Targets
Metric Unit Target Reference
Chirp bandkHz18 – 24FIG. 3
Chirp windowms≤ 200FIG. 3
Device ID spacedevices2⁸ = 256FIG. 3 · 306
Handoff cross-fade durations≈ 3FIG. 8 · 812
Handoff event latencys≤ 5FIG. 8 · 814
Audible-band leakagedB SPL @ 1 m≤ 0 dB above ambientFIG. 3 · 18 kHz line
BOM cost per deviceUSD≤ 30FIG. 5 · 514
Mesh sizedevicesup to 16 / householdFIG. 6
Off-host network egressbytes0FIG. 7 · claim 1(e)
Use-cases matrix Representative deployment scenarios 9 / non-limiting
Single-user home 3 – 5 ROOMS CLAIM 1 Multi-resident home PARENT + CHILD ZONES CLAIM 1 Open-plan office 8 – 16 DEVICES CLAIM 1 Audiobook follow-me PODCAST PACING CLAIM 4 Café / restaurant ZONE-BASED AMBIENT CLAIM 3 Car ↔ home handoff AUTOMOTIVE NODE 1110 CLAIM 1 Hospital ward PATIENT-ROOM PRIVACY CLAIM 5 Smart-museum tour NARRATION FOLLOW CLAIM 1 Gym / fitness zones SYNCHRONIZED CLASS CLAIM 1
Each tile represents a non-limiting deployment scenario. Claim references map to the claim that most-directly enables the scenario.
Working examples Representative measurements · bench data ¶ Ex. 1 – 3
Example 1 Three-room single-user household

Three smart-speaker embodiments (FIG. 5) deployed in kitchen, living room, and bedroom of a 90 m² apartment. Single adult subject; calibration performed once at install (FIG. 10). Audio session is a 2-hour podcast.

Metric Value Unit Reference
Handoff latency · median3.4sFIG. 8 · 814
Handoff latency · 95th percentile4.7s
Spurious handoffs per hour0.6eventshysteresis tuning · FIG. 4
Audible-band leakage−5dB SPL @ 1 m vs ambientFIG. 3 · 18 kHz line
Off-host network bytes per 2 h0bytesFIG. 7 · claim 1(e)
Cross-fade audibility (blind A/B)chance-levelFIG. 8 · 810
Example 2 Open-plan office (eight devices)

Eight devices arranged in a 6 × 10 m open-plan office: four soundbars (FIG. 11 · 1104), two smart speakers (FIG. 11 · 1102), and two phones (FIG. 11 · 1106) belonging to two specific users. Users walk between desks and break-area; the system routes the per-user audio session to the most-proximate playback device.

Metric Value Unit Note
Per-user disambiguation accuracy94%2-user test, 1 h log
Cross-user leakage< 4%after presence-vector filtering
Mesh-wide chirp overhead0.5% airtime200 ms / 40 s schedule
Setup time (8 devices)~ 90sFIG. 10 · C1000
Maximum mesh size validated16devicesper spec target
Example 3 Cross-product handoff · smart speaker → earbuds

Audio session originates on a smart-speaker (FIG. 11 · 1102) and is handed off to earbuds (FIG. 11 · 1108) as the user puts them on and leaves the room. The earbuds emit a low-amplitude bone-conduction ultrasonic acknowledgement chirp during the cross-fade window.

Metric Value Unit Note
Handoff trigger latency from "earbuds donned"1.8svia earbud IMU + first chirp
Cross-fade duration2.5sFIG. 8 · 812
Audio dropout during handoff0msconstant-power crossfade
Earbud ultrasonic emission · SPL @ ear~ 25dB SPLbelow human-perception threshold
Examples are illustrative and not limiting of the claimed invention per ¶ [0010].
Limitations & known issues Honest engineering assessment v0
The following limitations are known at the time of provisional filing and are catalogued so that an examiner — and any future continuation — has a clear record of what remains an engineering frontier.
Issue Manifestation Severity Mitigation path
Reverberant roomsExcessive multipath inflates effective RSSI of distant peers; presence vector smearsModeratenarrower chirp band; cepstral deconvolution; longer averaging window
Pets / small animals as absorbersThrough-body attenuation by ~ 20-kg dog can mimic a human at 6 mLowposture / motion fingerprinting; weight-of-evidence with user-anchor device 1106
Ultrasonic crosstalkMosquito repellers, ultrasonic pest deterrents, and some baby monitors emit in 20 – 40 kHzModeratefrequency-hopping within 18 – 24 kHz; CRC rejects mis-decoded payloads
Doors / curtainsMid-session boundary change occludes some peers; presence vector jumpsLowhysteresis Δ_min; debounce on sub-second jumps
Inaudible-to-most ≠ inaudible-to-allChildren and some young adults perceive 18 – 20 kHz; pets often perceive to ≥ 35 kHzLow – Moderateoperate at 20 – 22 kHz preferentially; user-configurable upper band
Driver bandwidth floorMany legacy speakers do not reproduce cleanly to 24 kHz; chirp distortionModeratespec-compliant tweeter required (FIG. 5 · 502); fallback to 18 – 20 kHz for legacy hardware
Multi-pet, multi-user homesPresence vector ambiguity when ≥ 3 absorbers move simultaneouslyModerateuser-anchor devices 1106 / 1108 provide a strong proxy for target user
No granted-patent enforceabilityProvisional confers no rights against competing implementations until non-provisional issuesLow (procedural)12-month conversion plan per prosecution roadmap
Enumeration of limitations is not an admission that any limitation defeats patentability.
Glossary of terms Acoustic + audio-routing vocabulary v0
Acoustic shadow Region behind an absorber (e.g., human body) where received ultrasonic SPL is reduced by 8 – 14 dB at 22 kHz (FIG. 9 · 908).
BFSK Binary frequency-shift keying — bit 1 encoded as up-chirp, bit 0 as down-chirp (FIG. 3 · 306).
Constant-power crossfade Cross-fade law in which the sum of the squared gain envelopes is unity — preserves perceived loudness during handoff (FIG. 8 · 810).
CRC Cyclic redundancy check — 8 bits appended to each device-ID payload for transmission-integrity validation (FIG. 3 · 304).
DoA Direction of Arrival — estimated by phase differences across the microphone array (FIG. 5 · 506).
Hysteresis · Δ_min Minimum required presence-vector delta before a handoff is triggered; prevents oscillation in the boundary region between devices.
Mesh The set of all EchoCast-participating devices in a physical environment, communicating exclusively via ultrasonic signals.
Multipath Multiple acoustic propagation paths arriving at a receiver via different reflections (FIG. 9 · 914 / 916).
Presence vector Per-device proximity score; computed from mutual RSSI plus through-body attenuation evidence (FIG. 6 · 614).
RSSI Received signal strength indicator — power level of a received chirp, in dB.
TX / RX Transmit / receive — every EchoCast node implements both roles.
Tweeter High-frequency speaker driver capable of clean reproduction to 24 kHz (FIG. 5 · 502).
User-anchor A device that travels with the user (phone, earbuds) and serves as a proxy for the user's position (FIG. 11 · 1106 / 1108).
Ultrasonic Sound at frequencies above the upper limit of human hearing — here defined as ≥ 18 kHz.
τ Chirp-RSSI threshold below which a peer is treated as out of mesh range (FIG. 4).
Information disclosure statement References known to applicant 37 CFR § 1.97 / § 1.98
U.S. patent literature
Document Date Assignee / inventor Relevant subject matter Distinguished by
US 9,024,998 B22015-05-05Polycom, Inc. (now HPE)Ultrasonic-beacon device pairing with auto-disconnect on room change and auto-reconnect on entering another roomclosest art on claim 1(a)–(c) primitive. Subject is videoconference pairing, not audio rendering; no human-absorber model; no mutual-attenuation matrix; CDMA-style payload, not TDMA scheduled
US 10,932,062 B22021-02-23Apple Inc.Ultrasonic proximity sensors for AirPods / audio handoff between earphone and hosttwo-device dyad, not N-device mesh; no TDMA among emitters; no matrix-valued inference
US 9,584,236 B22017-02-28Lisnr, Inc.Ultrasonic data transmission protocoldata-over-sound only; no audio-rendering handoff
US 9,742,496 B22017-08-22Asio Ltd. (Chirp.io)Acoustic data communication systemopaque payload; no rendering-layer coupling
US 10,158,738 B22018-12-18SilverPush Inc.Cross-device tracking via audio beaconstracking purpose; opposite privacy posture
US 9,392,353 B22016-07-12Apple Inc.Multi-room audio (AirPlay 2)IP-network signaling; cloud control plane
US 9,344,829 B22016-05-17Sonos Inc.Wireless multi-room playbackSonosNet IP mesh; explicit zone selection
US 10,873,820 B22020-12-22Google LLCCast device discovery and pairingIP / multicast DNS; explicit Cast selection
US 11,043,977 B22021-06-22Sonos Inc.Audio playback with chirp data exchange (post-Chirp.io acquisition)cited as Sonos's chirp-based primitive used in Sound Swap; v1.0 corrects earlier mis-characterization of this entry as "config only" — the Sound Swap product (Apr 2021) does use the chirp-and-RSSI primitive for audio handoff between speakers; surviving distinction is TDMA scheduling and through-body absorber model
Standards & specifications
Reference Body / year Subject matter Distinguished by
Bluetooth LE Audio · Auracast (BAS / BCAS)Bluetooth SIG · 2022Broadcast Isochronous Stream multi-receiver audioRF / BT signaling; pairing required
UPnP-AV / DLNA AV-TransportUPnP Forum · 2002HTTP-based multi-renderer audioIP-based; explicit renderer selection
CSA Matter 1.4Connectivity Standards Alliance · 2024IoT control surface incl. mediaIP / Thread; explicit control plane
RTP / RTCP (IETF RFC 3550)IETF · 2003Real-time audio transportIP-based; not a discovery mechanism
Non-patent literature
Reference Venue / year Subject matter Distinguished by
Sonos, "How near-ultrasonic audio adds spatial awareness to the Sonos system"Sonos Tech Blog · 2021 (Sound Swap shipped Apr 2021)Public disclosure of the chirp-and-RSSI primitive deployed in Sonos Sound Swap: > 19 kHz m-FSK chirp, CDMA multiple access, 3-meter same-room test, automatic audio handoff between Sonos speakersclosest commercial prior art on headline mechanic; conceded under § 102(a)/(b). Surviving distinctions: TDMA scheduling vs CDMA; matrix-valued N-way proximity vs binary same-room; published through-body absorber model vs wall α only; published adversarial threat model
Mavlankar et al., "Acoustic device pairing"ICASSP 2010Ultrasonic chirp-based device pairing for cameraspairing primitive only; not rendering handoff
Mehrabani et al., "Cross-device tracking via ultrasonic side channels"ACM SIGSAC 2017Audit of SilverPush tracking ecosystemprivacy-attack survey; informs threat model
Carlini et al., "Hidden voice commands"USENIX Security 2016Adversarial ultrasonic / inaudible commandsadversarial perception; informs CRC design
Roy et al., "BackDoor: Making microphones hear inaudible sounds"MobiSys 2017Ultrasonic-to-audible aliasing via mic non-linearitycross-modulation hazard; mitigation guidance
Chen et al., "Inaudible voice commands"CCS 2017DolphinAttack — ultrasonic injection into ASRadversarial attack class; informs auth design
Vasilakos et al., "Ultrasonic communication"IEEE Signal Process. Mag. · 2018Survey of in-air ultrasonic data transmissionsurvey reference; provides physics background
ISO 226:2003 / 2023ISO · 2003 / 2023Normal equal-loudness contoursdefines 18-kHz threshold cited in claim 1(a)
Cox & Antonio, "Acoustic Absorbers and Diffusers"3rd ed., 2017Reference for wall α values cited in FIG. 9physics reference
Citation of any reference herein is not an admission of prior art under 35 U.S.C. § 102 or § 103. · Applicant continues to monitor the prior art and may supplement this IDS prior to non-provisional conversion.
Regulatory considerations Operating-band & privacy framework v0
FCC · spectrum
The disclosed system emits acoustic energy in the 18 – 24 kHz band via a conventional audio transducer. Acoustic emissions in this band are not RF emissions and are therefore not subject to FCC Part 15 RF rules. The acoustic transducer 210 / 502, when operated within its OEM-rated SPL ceiling, complies with the limits of OSHA 1910.95 occupational noise and EN 50332 product safety; no Part 15 authorization is required.
GDPR · CCPA
The disclosed system processes no personal data within the meaning of GDPR Art. 4(1) or CCPA § 1798.140(o) at the level of inter-device signaling — device identifiers transmitted via ultrasonic chirp do not, in themselves, identify a natural person. User-anchor devices (FIG. 11 · 1106 / 1108) may be associated to a specific user, but such association is performed locally on the host and is not exposed off-host. No tracking pixels, cross-device identifiers, or audio content leave the household acoustic envelope.
FTC · SilverPush
The FTC's 2017 letters to twelve app developers regarding SilverPush-style ultrasonic cross-device tracking established that covert ultrasonic monitoring of users without disclosure may constitute a deceptive practice under FTC Act § 5. The disclosed system is materially distinguishable: (i) it is overt, disclosed in the product onboarding flow; (ii) it does not transmit beyond the household acoustic envelope; (iii) it does not collect cross-device tracking identifiers; and (iv) chirps are emitted by the user's own devices, not by third-party advertisements.
Audibility · ISO 226
ISO 226:2003 / 2023 (normal equal-loudness contours) places the median human upper-frequency limit at ≈ 18 kHz for adults; young children and some young adults perceive up to 20 – 22 kHz. The disclosed system preferentially operates at 20 – 22 kHz under nominal conditions; an installation may negotiate downward into 18 – 20 kHz only when no household occupant has been registered as a child during onboarding (per FIG. 10 · C1002).
Regulatory analysis is illustrative and not legal advice. · Specific deployments may require additional analysis under local law.
Prior-art comparison Multi-device audio coordination · cross-reference
The disclosed system substitutes an in-air ultrasonic device-presence protocol for the IP-network signaling layer of prior-art multi-device audio systems. The following matrix contrasts EchoCast against the dominant commercial systems on three axes — control-plane location, metadata exposure, and proximity-inference mechanism.
System Coord. plane Proximity inference Audio path Metadata egress Distinguished by
Sonos Sound Swap (Apr 2021)In-air near-ultrasonic > 19 kHz, m-FSK, CDMABinary same-room test via wall αLocal rendering (audio swap)none for the chirp itselfclosest art on headline mechanic. EchoCast differs in: TDMA vs CDMA · matrix N-way vs binary · through-body α · published threat model
US9,024,998B2 Polycom 2015In-air ultrasonic beaconPer-device decode of beacon IDVideoconference sessionIP address (in payload)different subject matter (VC pairing not audio render); single-device decode, no matrix; no human-absorber detection
Apple AirPlay 2IP + BonjourExplicit user toggle · HomeKit roomIP unicast / multicastApple-ID · device list · content URLrequires manual selection; cloud-resident control plane
Sonos S2IP + SonosNetExplicit zone configurationSonosNet meshaccount · zone topology · content metadatano automatic proximity-based handoff (Sound Swap is the dedicated Sonos feature for that — see row 1)
Google CastIP + DIAL / Cast SDKExplicit "Cast to" selectionIP unicast over LANGoogle-ID · device list · app metadatano implicit user-proximity inference
Bluetooth multipointBR/EDR or LE-Audio AuracastPairing + signal-strengthBT-encoded audio streamBT MAC · device names · audio codecrequires explicit pairing per device
Chromecast AudioIP + multicast DNSApp-side group selectionIP multicastdevice-ID · group-ID · cast app iddiscontinued; selection still explicit
This disclosure · EchoCastIn-air ultrasonic 18–24 kHz, TDMAMatrix-valued N-way presence; through-body absorber detectionLocal rendering onlyNone (0 bytes)
Comparison data sourced from publicly-available technical documentation as of 2026-05. · Citation of any system herein is not an admission of prior art under 35 U.S.C. § 102 or § 103.
Strategic positioning Control plane × metadata exposure · not a numbered FIG. Strategic exhibit
IN-AIR · HIGH METADATA CLOUD · HIGH METADATA IN-AIR · LOW METADATA CLOUD · LOW METADATA CONTROL-PLANE LOCATION → ↑ HIGH METADATA ↓ LOW METADATA in-air ultrasonic cloud / IP network zero heavy SilverPush tracking · cross-device Chirp.io · Lisnr data-over-sound Apple AirPlay 2 Apple-ID · cloud Sonos S2 SonosNet · account Google Cast Google-ID BT LE Auracast BT MAC · codec THIS DISCLOSURE · EchoCast in-air ultrasonic · 0 bytes egress EchoCast uniquely occupies the (in-air · zero-metadata) quadrant. SilverPush sits in the worst quadrant. AirPlay / Sonos / Cast inherit cloud-IP metadata; Chirp / Lisnr lack rendering-layer coupling.
Reference numerals 81 entries
100 system
102 device A · kitchen
104 device B · living
106 device C · bedroom
108 user
110 user trajectory
112 audio handoff
114 plan view · household
200 device internals
202 audio input
204 router
206 fader
208 driver-stage
210 acoustic transducer
212 microphone array
214 chirp decoder
216 presence inference
218 (reserved)
220 handoff coordinator
222 control signal
224 chirp emitter
300 spectrogram fig
302 preamble pulse
304 CRC
306 device-ID payload
308 checksum byte
400 state machine
402 IDLE
404 RECEIVING
406 SOLO
408 LISTENING
410 HANDING-OFF
500 hardware embodiment
502 tweeter
504 main driver
506 mic array
508 edge controller
510 USB-C / power
512 ultrasonic emission
514 claim hook · BOM
600 mesh topology
612 user position
614 presence vector
700 privacy diagram
702 in-home scope
704 cloud scope
706 local boundary
708 remote boundary
710 device IDs
712 presence vector (data)
714 render state
716 egress prohibited
718 absent off-host
720 claim limitation
800 timing fig
802 lane · chirp A
804 lane · chirp B
806 lane · chirp C
808 lane · active source
810 lane · cross-fade
812 cross-fade duration
814 handoff event
900 propagation fig
902 plan view
904 TX device A
906 user · absorber
908 acoustic shadow
910 RX device B
912 direct path
914 reflected · ceiling
916 reflected · floor
918 claim hook · physics
920 claim hook · RSSI-only
1000 calibration fig
C1002 discover step
C1004 RSSI baseline
C1006 user-walk prompt
C1008 presence attenuation
C1010 presence kernel
C1012 fit τ / Δ_min
C1014 store profile
1016 drift re-fit hook
1100 hardware taxonomy
1102 smart speaker
1104 soundbar
1106 phone / tablet
1108 earbuds
1110 automotive
1112 capability matrix
1114 claim hook · taxonomy
1200 spectrum fig
1202 EchoCast band
1204 dog whistles
1206 hearing-aid feedback
1208 Silver/Chirp/Lisnr
1210 mosquito repellers
1212 parking sensors
1214 bat detectors
1216 claim hook · mitigation
1300 frame spec fig
1302 preamble
1304 sync · Barker-4
1306 device-ID payload
1308 CRC-8
1310 guard
1312 total frame · 192 ms
1400 TDMA schedule fig
1402 device A · slot 0
1404 device B · slot 200 ms
1406 device C · slot 400 ms
1408 device D · slot 600 ms
1410 device E · slot 800 ms
1412 airtime utilization
1414 claim hook · TDMA
1500 threat model fig
1502 household envelope
1504 legitimate mesh
1506 adversary positions
1508 passive eavesdropper
1510 active injection
1512 replay attack
1514 denial of service
1516 defender invariants

Part VI · Execution

Statement of industrial applicability PCT Art. 33(4) · preparatory

The disclosed invention is industrially applicable. It may be manufactured, sold, and used in the following representative markets without limitation:

  • Consumer audio — smart speakers, soundbars, AV receivers, premium portable speakers.
  • Mobile devices — phones and tablets implementing EchoCast as a system service.
  • Wearables — earbuds and headphones implementing the RX-dominant variant of FIG. 11 · 1108.
  • Automotive — in-cabin head units and rear-seat entertainment systems.
  • Hospitality & commercial — hotels, restaurants, fitness studios with multi-zone audio.
  • Healthcare — hospital wards and patient-room audio where IP-network isolation is required.

The invention may be embodied with commodity tweeters (FIG. 5 · 502) at bill-of-materials cost below $30 per device, and integrates without modification of host operating-system audio frameworks.

Markets
  • · Consumer audio
  • · Mobile devices
  • · Wearables · earbuds
  • · Automotive
  • · Hospitality · commercial
  • · Healthcare · privacy-critical
Cost basis
  • BOM: < $30 / device
  • 0 off-host bytes
  • 0 cloud dependency
Prosecution roadmap Provisional → Grant · 36-month horizon Indicative
TRACK MILESTONES 36 MO M0M6M12M18M24M30M36 Provisional Non-provisional USPTO exam PCT international Foreign nationalization Public disclosure Grant horizon FILE · M0 · 2026-09 expiry · M12 CONVERT · M12 First OA Response Allowance PCT filing window · 18 mo National-phase entry (EP · JP · KR) website draft · § 102(b)(1) grace ★ Grant target · M30 – M33 KEY active examination optional window contingent Provisional grants no enforceable rights; preserves filing-date priority for non-provisional, which must follow within 12 months under 35 U.S.C. § 119(e).
Pre-filing checklist Inventor action items Living document
Document · drafting
  • 15 sheets drafted (FIG. 1 – 15)
  • Independent claim 1 + 4 dependent + 1 apparatus
  • Alternative claim sets 1A / 1B preserved
  • Detailed description ¶ [0001] – [0011]
  • IDS with 7 patent + 4 standards + 8 NPL refs
  • Abstract ≤ 150 words
  • Final proofread by qualified counsel
  • PDF export · 8.5 × 11 in · ≥ 300 dpi line art
Technical · bench validation
  • ESP32-S3 + tweeter board with 18 – 24 kHz reproduction
  • BFSK encode/decode · ≥ 95 % accuracy at 3 m
  • TDMA scheduling validated at 5-device mesh
  • Cross-fade A/B blind test · ≤ chance audibility
  • Adversarial test · replay / injection rejection
  • 30-day drift run · re-fit triggered automatically
Filing logistics · USPTO
  • USPTO Patent Center account active
  • Cover-sheet form PTO/SB/16 completed
  • ADS form PTO/AIA/14 finalized
  • Micro-entity certification PTO/SB/15A signed
  • Provisional filing fee · $60 (micro) paid
  • Drawing PDF uploaded · passes EFS validator
  • Application body PDF · text-searchable
  • Receipt of filing date · application no. assigned
Post-filing · 12-month window
  • Docket provisional expiry · 2027-09 (T + 12 mo)
  • Decide non-provisional vs. abandonment · by M9
  • Engage patent counsel · by M9
  • Decide PCT path · by M11
  • File non-provisional + ADS + claims · by M12
  • Update public disclosure log post-filing
Assignment of rights Ownership chain 37 CFR § 3.11
Inventor → applicant
At the time of this provisional filing the inventor named herein retains all right, title, and interest in the disclosed invention. No assignment to a third party has been executed; no assignment recorded with the USPTO Assignment Recordation Branch. The invention was conceived and reduced to practice outside the scope of any employment or contractual obligation that would automatically vest rights in a third party.
Trademark notice Common-law mark Lanham Act § 43(a)
Mark
EchoCastTM

ECHOCASTTM is a common-law trademark of Jae Hoon Kim, used in connection with software and hardware embodying the invention disclosed herein. Federal registration has not been sought as of the date of this draft; the inventor reserves the right to seek registration at any time.

Apple®, Sonos®, Google®, Espressif®, Bluetooth®, Wi-Fi®, AirPlay®, Sonos®, and Chromecast® are trademarks of their respective holders, used herein in a descriptive and nominative manner under 15 U.S.C. § 1115(b)(4).

Application data sheet PTO/AIA/14 analogue 37 CFR § 1.76
§ 1 · Title
System and method for migrating an audio rendering session between devices using inaudible ultrasonic device-presence signals
§ 2 · Type
×Provisional Non-provisional Design PCT
§ 3 · Applicant / inventor
Name
Jae Hoon Kim
Citizenship
United States
§ 4 · Correspondence
jk2765@cornell.edu
§ 5 · Filing identifiers
Internal docket
EC-2026-A · v1.0
Drawing sheets
15
Claims
6 (1 indep + 4 dep + 1 apparatus)
Proposed CPC
H04R 27/00 · H04L 12/28 · H04W 4/02 · G10L 19/00
Declaration of inventor 37 CFR § 1.63 Draft

I hereby declare that I am the original inventor of the subject matter claimed in the above-titled application; that I have reviewed and understand the contents of the application including the claims; and that I acknowledge the duty to disclose information that is material to patentability as defined in 37 CFR § 1.56.

Signature of inventor
Date
Inventor
Jae Hoon Kim
Citizenship
United States
Internal docket
EC-2026-A · v1.0
Version history Draft · not filed

EchoCast is the foundational anchor for its descendant family (ChirpLock as the security-primitive sibling, see /chirplock). The v1.0 narrow-claim posture has been carried into the portfolio-wide audit pass.

How to cite this disclosure For academic / non-patent references
BibTeX
@misc{kim2026echocast,
  title   = {EchoCast: System and method for migrating an audio rendering
             session using inaudible ultrasonic device-presence signals},
  author  = {Kim, Jae Hoon},
  year    = {2026},
  month   = may,
  howpublished = {Provisional patent application draft v1.0},
  url     = {https://jaehoon.kim/echocast},
}
APA 7
Kim, J. H. (2026, May 17). EchoCast: System and method for migrating an audio rendering session using inaudible ultrasonic device-presence signals [Provisional patent application draft v1.0]. https://jaehoon.kim/echocast
Colophon EchoCast · provisional draft v1.0
Document hash
/echocast · v1.0
Typeset in
Astro 5 · Tailwind 4 · system-monospace
Pre-filing draft · no rights conferred
© 2026 Jae Hoon Kim
End of drawing set · EchoCast · Provisional draft · Not a filed application
Index