Connectivité Bitcoin hors réseau

Réponse rapide

Sortir Bitcoin du réseau, c'est deux problèmes distincts, et un seul est résolu. RECEVOIR est facile et gratuit : Blockstream Satellite diffuse toute la blockchain depuis quatre satellites géostationnaires vers n'importe quelle parabole à ~150 $, si bien qu'un nœud reste en pleine synchronisation de consensus sans aucun Internet, pour toujours. DIFFUSER votre propre transaction signée est la moitié difficile — une liaison descendante en réception seule ne peut physiquement pas émettre, il vous faut donc une liaison montante, et chaque option est un compromis. Starlink offre du haut débit complet mais c'est un fournisseur unique avec KYC qui tire 50-100 W ; Rogers Satellite (Starlink direct-to-cell) transforme n'importe quel téléphone en canal de diffusion de tx à 15 $ CA/mois ; un seul message Iridium pousse une transaction depuis le pôle Nord littéral pour quelques sous ; et une transmission radioamateur JS8Call n'a besoin que d'une radio et d'un ami licencié. La pile robuste est asymétrique à dessein : liaison descendante satellite gratuite pour les blocs, plus la plus petite liaison montante payante que vous tolérez pour les dépenses.

Recevez gratuitement, diffusez petit. Pour un site canadien reculé en 2026 : Starlink comme lien principal, Rogers Satellite sur votre téléphone comme canal de diffusion de poche, un récepteur Blockstream Satellite comme flux de blocs à l'épreuve des pannes, et Iridium comme la seule chose qui fonctionne encore dans le Haut-Arctique. Évitez Swarm (mort en mars 2025) et n'attendez pas AST SpaceMobile (pas d'opérateur canadien, repoussé à 2027).

Option	Direction	Adéquation Bitcoin	Couverture	Coût	Hors-FAI
Blockstream Satellite (block/tx downlink)sat-broadcast-rx · operational	receive-only	Full-node sync: receives complete blocks + new txs 24/7, so a node validates the chain and confirms incoming payments fully offline. It CANNOT transmit — you cannot broadcast your own signed tx through it.	Near-global: Galaxy 18 (North America incl. most populated Canada), Eutelsat 113, Telstar 11N (EU/Africa), Telstar 18V (APAC). High-Arctic Canada marginal (GEO look angle).	US$0/month — free and unmetered forever once you own the receiver	yes
~100-200 kbps continuous downlink (keeps pace with 10-min blocks; a full IBD over satellite takes weeks — most users sync once via internet/USB then hold sync via satellite) · Seconds behind internet propagation (GEO one-hop ~250 ms); real-time for confirmations · 45-90 cm dish + Ku/C-band LNB + receiver (RTL-SDR ~US$30-50, or TBS/Novra pro); DIY ~US$100-200; kits historically ~US$250-500 (2026 kit pricing unverified). Confirmed operational into 2025-2026 (Telstar 11N restored, new frequencies, API/GUI updates). Receive-only is the defining constraint: you still need SOME uplink to broadcast a spend.
Iridium (SBD / Certus / sat phones)sat-phone-data · operational	two-way	Purpose-fit for the broadcast problem: one ~250-vbyte signed tx fits in a SINGLE 340-byte SBD message to a relay endpoint (a RockBLOCK 9603 does it for well under a dollar). Certus 100/200 can even limp a pruned node or Lightning gossip.	100% of Earth including the entire Canadian Arctic — the ONLY option here with no northern-latitude caveat	SBD line ~US$12-20/month + per-credit usage; GO! plans ~US$55-160/month (reseller figures)	yes
SBD 340 B/message; GO! 2.4 kbps; Certus 100 ~22-88 kbps; Certus 700 up to 704 kbps · Seconds (real-time network, no pass-waiting — unique among small-data options) · RockBLOCK 9603 ~US$250-300; Iridium GO! ~US$800; 9555 phone ~US$1,100; inReach ~C$300-500 (messaging only, no raw-tx API — unverified). The professional-grade answer to 'broadcast a tx from anywhere on the planet, today, with certainty.' Expensive per byte, but the byte you need to send is small.
Starlink (LEO broadband)LEO-broadband · operational	two-way	Everything: run a full archival node, IBD in hours, serve peers, run Lightning with normal routing, remote-manage an off-grid mining site (Stratum traffic is a few kbps per fleet). The default answer for off-grid mining in Canada.	Effectively all of Canada incl. territories and high latitudes (polar shells); global land coverage in licensed countries	Canada 2026: ~C$110-140/month residential; Roam C$70 (100 GB) or C$189 unlimited (verify on starlink.com/ca)	partial
~100-300+ Mbps down / 10-30 Mbps up typical in Canada · ~20-50 ms (Canada typical ~36 ms) · Standard kit ~C$649; Starlink Mini ~C$249 (2026 aggregator pricing); needs ~50-100 W continuous — a real load on a small battery bank. Avoids terrestrial ISPs but is still one company's KYC'd, shut-offable service with ground-station backhaul — resistant to local outages, not to SpaceX or a state ordering disconnection. Pair with a Blockstream Sat receiver for a path that survives Starlink loss.
Starlink Direct to Cell (Rogers Satellite / T-Satellite)direct-to-cell · operational	two-way	Enough to broadcast a signed tx from your pocket: paste a raw tx to an SMS-to-broadcast gateway, or push it through a whitelisted messaging app to a relay bot. Cannot run a node; general TCP data is gated until V2 satellites (5G-class, targeted 2H 2027).	Canada-wide via Rogers Satellite (5.4M km² beyond terrestrial networks); US via T-Mobile; cross-border roaming live since Apr-May 2026	Canada: C$15/month add-on, bundled free into many Rogers 5G+ plans in 2026; US: US$10/month	yes
Text/MMS-class today; shared per-beam LTE — kbps-per-user for app data · Seconds for messages; app data usable but bursty · Any recent unmodified 4G/5G smartphone — zero extra gear. 🔴 The single biggest 2025-26 shift here: every Canadian with a Rogers add-on now carries a satellite uplink able to move a signed tx — the natural back-channel companion to a receive-only Blockstream Sat node. Data stays app-gated on V1; 'run a wallet normally' is 2027+ (V2).
Amateur radio HF data (JS8Call / packet)ham-radio · niche	two-way	A ~250-byte signed tx encodes into a JS8Call transfer taking several minutes at ~10-40 bps; the receiving station (with internet) rebroadcasts it. Nothing more than tx-sized payloads is practical — no blocks, no node.	Global-capable via skip, weather/solar dependent; works anywhere in Canada	US$0/month forever	yes
~6-40 bps effective (JS8 normal/turbo); HF packet ~300 bps · Minutes per message + needing a listening counterparty and open propagation · HF transceiver + antenna + computer: ~US$500-1,500 new (used QRP + wire antenna <US$300); requires an amateur licence (ISED Basic in Canada, free exam). 🔴 Legal gray zone: amateur rules ban encrypted content and 'pecuniary interest' traffic — a signed tx is public (not encrypted), but routine financial use arguably violates the commercial-traffic ban; treat as emergency-only. Zero-infrastructure is the last-resort property nothing else here has.
Blockstream Satellite API (pay-to-broadcast down)sat-broadcast-rx · operational	send-only	Delivers data TO offline satellite receivers worldwide — useful to REACH an off-grid node. But the API request + Lightning payment must go over the internet, so it does NOT solve the off-grid uplink; it is the mirror image of it.	Broadcast globally over the same GEO beams as the Blockstream Satellite downlink — any receiver in those footprints hears it; sending only needs internet access	Millisatoshis per byte via Lightning (spot rate; historically ~1 msat/byte floor — current unverified)	no
Message-oriented; a ~250-byte tx is trivial; queue-based · Seconds to minutes depending on bid/queue · None to send (any internet device); a Blockstream Sat receiver on the far end to hear it. Frequently misunderstood: it broadcasts DOWN to receivers, it is not an uplink FROM an off-grid site. Real use: a connected friend/service relaying data to your disconnected receiver.
Traditional GEO VSAT (Viasat / Hughesnet / Xplore)GEO-broadband · operational	two-way	Can run a full node (slow sync) and broadcast txs normally. Lightning is degraded (600+ ms RTT sluggish HTLC/gossip) but functions. Fine as mining-site backhaul; painful for anything interactive.	Southern Canada (GEO beams degrade above ~60-65N); practical Canadian incumbent is Xplore GEO (details unverified)	~US/C$65-150/month with priority-data caps (specifics unverified)	partial
Viasat Unleashed ~25-60 Mbps real-world; Hughesnet ~8-20 Mbps · ~600-750 ms round trip (GEO physics; cannot improve) · ~75 cm dish + modem, usually pro-installed/leased. A dying category: EchoStar disclosed (Nov 2025) it is exiting consumer satellite internet, referring HughesNet customers to Starlink. Only choose GEO VSAT where Starlink is unavailable or as a dissimilar-path backup.
Sateliot (5G NB-IoT from LEO)LoRa-satellite · beta	two-way	Small-packet two-way: adequate to uplink a signed tx to a relay endpoint and receive short confirmations. Nowhere near node or Lightning capable.	Global-by-design incl. Canada, but sold through operator agreements — consumer availability in Canada unverified	unverified (wholesale per-device IoT, reportedly a few US$/month class)	yes
NB-IoT class: hundreds of bits to a few kbps in short sessions · Minutes to hours per pass (early constellation); improves as satellites launch · Standard NB-IoT NTN module/dev kit (~US$20-100 class); sold wholesale via MNO/MVNO, not retail. The most credible future 'cheap tx uplink' because it uses commodity 3GPP silicon, but in mid-2026 it is early-commercial and not something a Canadian pleb can just sign up for. Someone must still run the ground-side relay that hands your tx to the network.
Lacuna Space (LoRaWAN LR-FHSS to LEO)LoRa-satellite · niche	two-way	Marginal: uplink-oriented, 50-byte messages mean a ~250-byte signed tx must be fragmented across ~5+ messages and reassembled by a custom ground relay — a hack, not a product. No node/Lightning.	Polar LEO orbits give global reach incl. northern Canada; individual availability unverified	unverified (B2B; now also licensing the tech to other operators per MWC 2026)	yes
~50-byte messages, min 6/day; ~300 bps-class LR-FHSS bursts · Hours-scale (pass-dependent, store-and-forward) · Standard LoRa/LR-FHSS module (Semtech LR11xx class, ~US$10-50) + antenna. Interesting because your existing Meshtastic-class LoRa hardware family can, with LR-FHSS, reach orbit — the conceptual bridge between the terrestrial-mesh dataset and this one. Fragmenting a tx across six daily 50-byte slots is survivalist-grade.
EchoStar Mobile LoRa (GEO S-band, Europe)LoRa-satellite · operational	two-way	Real-time small-packet two-way that could relay a signed tx to an internet gateway in seconds — technically the best LoRa-satellite fit for tx broadcast, but only inside Europe.	Europe only — no Canada/North America service	unverified (B2B via partners + Swisscom)	yes
LoRa-class: tens to hundreds of bytes per message, low duty cycle · Near-real-time (GEO ~600 ms path) · EchoStar/Semtech S-band LoRa module (EM2050 class) via IoT partners; cost unverified. Listed for category completeness; irrelevant to Canadian off-grid users today. EchoStar's 2025-26 spectrum turmoil makes long-term continuity worth watching (unverified).
AST SpaceMobile (BlueBird direct-to-device)direct-to-cell · announced	two-way	None yet (pre-commercial). If it delivers claimed multi-Mbps to handsets it would support full wallet apps and lightweight nodes from a phone — but that is a 2027+ promise, not a 2026 capability.	Initial intermittent US coverage first; Canada unannounced (no disclosed Bell/Telus deal — unverified)	unverified (future carrier add-on)	yes
Claimed ~20-120 Mbps per beam to standard phones (demo figures; commercial rates unverified) · LEO-class, tens of ms · Any unmodified smartphone. Track, don't rely on: timeline has slipped repeatedly (late-2026 to early-2027) and Canadian carrier partnerships are absent. Starlink Direct-to-Cell is the one you can actually buy in Canada today.
Swarm (SpaceX VHF IoT) — DISCONTINUEDLoRa-satellite · discontinued	two-way	Historically could carry a ~250-byte signed tx in two packets to an internet-side API for rebroadcast. Now: NONE — the network is gone.	Was global incl. polar	Was US$5/month per device	yes
Was ~1 kbps bursts, 192-byte packets, store-and-forward · Was minutes to hours (waiting for a pass) · Swarm M138 modem (~US$119) — now e-waste for connectivity. 🔴 Included as a WARNING: guides written 2021-2024 still recommend Swarm for off-grid tx broadcast; that path died March 2025. Its spiritual successor is Starlink Direct-to-Cell.

Le problème de la diffusion. Receiving Bitcoin off-grid is solved and free — Blockstream Satellite pushes every block and mempool tx to a ~$150 dish setup forever. Broadcasting is the hard half: a signed ~250-byte transaction still needs SOME uplink, and every option is a trade — Starlink (full broadband, ~C$110+/mo, one KYC'd provider), Starlink Direct-to-Cell texting from a bare phone (C$15/mo, the cheap new 2025-26 answer), a single Iridium SBD burst (works from the North Pole for pennies-per-message on a ~$15/mo line), a LoRa-to-satellite hack, or a JS8Call HF transmission to a connected ham. The robust off-grid stack is asymmetric by design: free satellite downlink for blocks + the tiniest paid uplink you can tolerate for spends.

La pile hors-réseau canadienne. For a remote Canadian cabin or mining site the realistic 2026 stack is: Starlink as primary (full-Canada coverage incl. territories, runs nodes/pools/Lightning outright), Rogers Satellite (C$15/mo Starlink direct-to-cell on any modern phone, free on many 2026 Rogers plans) as the pocket tx-broadcast channel, a Blockstream Satellite receiver on Galaxy 18 as the outage/censorship-proof block feed, and Iridium (inReach/RockBLOCK/GO!) as the only layer that still works in the High Arctic where GEO look angles fail. Swarm is dead (Mar 2025), GEO VSAT (Hughesnet) is in managed decline, and AST SpaceMobile has no Canadian carrier and slipped to 2027 — none of those belong in a Canadian plan today.

Données ouvertes (CC BY 4.0) : CSV · JSON · API : /wp-json/dc/v1/off-grid-connectivity

Produits, réparations et guides connexes

Dernière révision: 27 juillet 2026.