TARASEC FOR INTERNET PROVIDERS

Direct affected traffic through TaraSec. Keep the normal path for everyone else.

TaraSec explores how cooperating internet service providers can identify affected connections, share useful threat information and help customers recover. A proposed selective-routing deployment lets normal traffic continue through existing ISP infrastructure.

CAPACITY & LATENCY

Size the TaraSec path for affected traffic.

In this proposed design, only traffic selected by the ISP's security policy goes through the TaraSec gateway. Normal traffic stays on the existing path. Additional gateway processing and latency therefore apply primarily to redirected connections.

Capacity depends on affected connection numbers, traffic volumes, the routing design, restrictions and how long customers remain on the security path. Prompt notification and practical help may shorten that time. Once cleared, customers can return to their normal path—an incentive to resolve the problem.

Measure throughput, connection capacity and latency in a pilot, including peak and failure conditions. The selection mechanism on existing infrastructure also needs capacity testing. TaraSec is experimental; this architecture is a deployment proposal, not a claim of ISP-scale performance.

PROPOSED ISP ARCHITECTURE

Identify, steer, assist and restore.

01

Identify the affected connection

The provider correlates evidence with its subscriber session or, where possible, an individual unit. A shared home gateway may identify only a household; with carrier-grade network address translation (CGNAT), a public address alone may represent multiple subscribers. Use the ISP's own mapping and connection records.

02

Redirect a newly identified connection

Existing routers or subscriber-management systems can steer traffic toward the TaraSec gateway using policy-based routing (PBR), subscriber policies or another supported mechanism. The selected scope and confidence should reflect the evidence.

03

Maintain a separate path

A virtual local area network (VLAN) separates a network segment. A virtual routing and forwarding instance (VRF) provides a separate routing table. Either or both may be useful; a subscriber policy may be more practical on some access networks. Immediate redirection and a longer-lived policy can lead to the same gateway.

04

Recover and return

Allow appropriate access to updates, support and recovery resources. After remediation or correction of a false attribution, apply the provider's clearance process and restore normal routing. Users need a practical way to request help or challenge a classification.

ADDRESSING & OPERATIONS

A separate public IP pool is an option.

The TaraSec gateway may use a designated public IPv4 pool with network address translation (NAT), or another routing and address-assignment design. A public address identifies a path, not proof of infection or a specific person. Receiving networks remain responsible for their own policies.

Changing a public address can interrupt existing connections. Plan for return traffic, stateful firewalls, NAT and connection tracking, IPv6, route loops, capacity, redundancy, gateway failure and rollback. Integration, tagging and subscriber steering require development and validation for the provider's equipment; they are not implied to be automatic today.

CUSTOMER RECOVERY

Inform people and help them stay secure.

Providers can promote the TaraSec App as a way to view available security information about a unit and use recovery functions where permitted. AI-assisted guidance is being developed to explain warnings and help users investigate, clean their devices and maintain security. Immediate interactive assistance for every affected customer is a goal, not a service guaranteed today.

The goal is to address threats without requiring the ISP to inspect unrelated private activity. The ISP retains its subscriber mapping; technical identifiers and metadata still need protection. Read about privacy and access boundaries.

Business customers may choose a managed security service from their ISP or a specialist, with agreed responsibilities, authorised access and data handling. Larger customers may prefer to keep detailed investigations within their own environment.

DEMO 4 · PARTNER ROUTING

Test the selective-path principle.

Demo 4 compares clean and selected tagged traffic to a configured test destination. It aims to show the selected traffic taking a partner egress path with a recognizable public address while ordinary traffic uses its normal route. The live viewer compares the observed source addresses.

The demonstration tests a limited routing scenario, including a hotspot behind NAT or CGNAT. It does not establish automatic ISP subscriber quarantine, a production VLAN or VRF integration, remediation, or production-scale performance. A pilot must test those separately.

PILOT WITH TARASEC

Test the architecture on your network.

We welcome internet providers, researchers and security partners who want to examine subscriber identification, selective routing, customer recovery and measured performance in a controlled pilot.

Contact TaraSec