Abstract

Robocalls, spam texts and scam calls persist for one structural reason: in today’s telephone and messaging networks, knowing a number is enough to ring it. Once a phone number, SIP address or messaging handle leaks — through a directory listing, a marketplace enquiry, a data breach or a broker list — it works as a permanent, reusable key to its owner’s attention. Existing defences answer different questions. STIR/SHAKEN verifies who is calling, not whether that caller is allowed to reach this recipient now. Number masking hides the real number, but the masked alias usually stays dialable while active. Spam scoring, call labelling and do-not-call registries act after an attempt already exists or rely on after-the-fact enforcement. This paper describes an authorization-to-reach architecture that separates identifier possession from permission to contact. Every inbound call or message is held as a candidate request until a network enforcement point (SIP proxy, session border controller, application server or CPaaS gateway) validates a current grant bound to sender, recipient, purpose or transaction, channel, expiry and attempt quota. A valid grant releases only the bounded path for that one communication and is consumed atomically; a leaked, copied, expired or out-of-purpose handle produces no ring, no notification and no media session. The paper presents the end-to-end workflow, three deployment profiles from blocked-path to absent-path enforcement, interworking with legacy PSTN and emergency services, and worked examples covering marketplace callbacks and machine-speed AI-agent calling. As IMT-2030 (6G) networks introduce AI-native services and software agents able to originate signalling at machine scale, the gap between authenticated and authorized contact widens. The model complements STIR/SHAKEN and ITU-T work on countering spam and numbering misuse, and requires no new radio, core-network or cryptographic primitives. A runnable reference implementation is publicly available. The author is open to acquihire or strategic collaboration.

KEYWORDS FOR PATENT EXAMINATION

Core concept: callee consent, recipient consent, consent-based communication, permission to call, authorization to call, call authorization token, callee-issued token, recipient-issued credential, contact permission, reachability control, reachability authorization, call admission control, inbound call authorization, pre-call authorization, consent token

Scoped and consumable grants: time-limited token, expiring token, one-time token, single-use credential, use-limited token, call quota, attempt counter, purpose-bound token, transaction-bound token, context-bound authorization, scoped capability, revocable token, token revocation, nonce, replay protection, atomic consumption

Masked and temporary identifiers: number masking, anonymous number, proxy number, virtual number, temporary phone number, disposable phone number, ephemeral identifier, alias number, privacy relay, call relay, masked calling, anonymous communication session, rotating identifier, unlinkable identifier

Spam and robocall terms: unsolicited call, unwanted call, nuisance call, robocall, telemarketing call, spam over Internet telephony, SPIT, call screening, call blocking, call filtering, caller ID spoofing, spam call detection, whitelist, blacklist, do-not-call

Network enforcement: session border controller, SIP proxy, application server, call setup, SIP INVITE, pre-INVITE check, session establishment, call routing, terminating network, call rejection, SIP 403, signalling gateway, CPaaS, communication API gateway, policy enforcement point, policy decision point

Identity versus authorization: STIR, SHAKEN, PASSporT, caller authentication, attestation, originating identity, capability-based access control, bearer token, proof of possession, sender-constrained token, holder binding, non-bearer credential

Application context: marketplace contact, business directory, map listing, click-to-call, ride-hailing communication, delivery driver contact, AI agent call, automated dialer, machine-originated call, IMT-2030, 6G, AI-native network

CPC CLASSIFICATIONS H04M 3/436 - Screening of incoming calls H04M 1/663 - Preventing unauthorised calls to a telephone set H04L 65/1076 - Screening of IP real-time communications H04L 65/1079 - Screening of unsolicited session attempts (SPIT) H04L 63/10 - Network security, controlling access H04L 63/0807 - Token-based authentication H04L 63/108 - Time-limited access policies H04L 51/212 - Message filtering or selective blocking H04W 12/08 - Access security in wireless networks H04M 2203/6045 - Identity confirmation H04L 9/32 - Authentication cryptography

SEARCH STRING (callee OR recipient OR called-party) w/5 (consent OR permission OR authoriz*) AND (token OR credential OR grant) AND (expir* OR single-use OR quota OR revoc*) AND (call OR SIP OR INVITE OR message)

RELATED PRIOR ART RFC 5360 - A Framework for Consent-Based Communications in SIP RFC 5039 - The Session Initiation Protocol (SIP) and Spam RFC 3880 - Call Processing Language (CPL) RFC 8224 - Authenticated Identity Management in SIP RFC 8225 - PASSporT: Personal Assertion Token RFC 8588 - PASSporT Extension for SHAKEN 3GPP Communication Barring (CB) and Anonymous Communication Rejection (ACR) Masked-number and proxy-calling systems (ride-hailing, marketplace, call tracking)

DISTINGUISHING FEATURES 1. Grant bound to sender, recipient, purpose or transaction, channel, expiry and attempt quota 2. Atomic quota consumption per logical communication act; retransmissions and forks do not consume new authority 3. Absent-path enforcement: the visible handle alone resolves to no route, notification or media path 4. Non-bearer grant bound to the authenticated sender 5. Candidate-request hold before any ring, notification or media allocation 6. Emergency and legally mandated calls always bypass grant enforcement

Creative Commons License

Creative Commons License
This work is licensed under a Creative Commons Attribution-Noncommercial 4.0 License

Share

COinS