GO NTN

Goal Oriented Schedulers for Non-Terrestrial Networks

Purely software based capacity boost for non-terrestrial networks: more useful data per pass, more customers per satellite, no new spacecraft.

Scalable NTN service without additional orbital investment and value-based differentiated data services for enhanced data monetization.

*1.5x–20x: GO NTN value proposition, more customers without additional orbital investment; the exact gain will be quantified experimentally in the PoC. 49%: goal oriented bundle management over FIFO; 30×: age-optimal preemption, heavy-tailed delay regimes (GO SPACE).

A new kind of decision problem

3GPP 5G Rel. 19 has specified the store and forward operation mode for base stations in orbit. In this mode, while a satellite is disconnected from the ground segment, it stores users’ traffic, and forwards it once a feeder link becomes available.

To use this window efficiently, the network must determine what is transmitted, in what order, and what is no longer worth transmitting at all.

Store and forward (S&F) · Rel. 19

Store and forward Users on the ground send traffic to a satellite storing node running GO NTN. The node stores the traffic and forwards it to the ground segment once a feeder link becomes available. Animated: updates fill the on-board buffer while the feeder link is down, then drain to the ground when the feeder window opens. Storing node · GO NTN Users Ground segment feeder link
It stores users’ traffic, and forwards it once a feeder link becomes available.

Two kinds of scarcity

5G Rel-19 stores and queues, but does not decide

Sparse constellations · Scarcity

The window is short

Traffic waits on board for minutes to hours, then only a brief feeder window opens.

3GPP Rel-19 store-and-forward · TR 23.700-29

GO NTN decides what leaves the satellite in a short contact window.

Mega-constellations · Scarcity

The beam is crowded

Thousands of satellites and ground nodes. Coverage is continuous, but every beam is shared by thousands of nodes sending tiny packets.

3GPP Rel-19 IoT-NTN uplink capacity enhancements

GO NTN decides which nodes are served first in a crowded beam.

The problem. In both cases, traffic is served in arrival order or by static priority. The small number of static priority levels cannot differentiate among large volumes of stored messages, which then compete equally for scarce resources. With commercial reality at ≪10 messages per device, per day (e.g., 50-byte limits), what gets sent matters more than how much.

Mapping GO SPACE results to deployable decisions

  1. Retain and replace

    Bounded admission, replacement and eviction rules for S&F buffers.

  2. Hold versus release

    Optimal-stopping structures adapted to uncertain contact opportunities.

  3. Harvest idle space

    Released opportunities reassigned to throughput-oriented traffic.**

  4. Contact selection

    Supply availability-aware routing can improve data freshness.

**Patent pending: TR2026/006400, Systems and Methods for Harvesting Self-Generated Idle Space in Goal-Oriented Communications (FRESHDATA).

GO NTN Decision Engine

When capacity is restricted, GO NTN sends the information that is most useful first.

Decision functions at the storing node

  1. Pending updates on board, or waiting at thousands of nodes
  2. Value/age scoring class, age, size, forwarding budget
  3. GO scheduler ordering, admission, eviction

Each stored item is scored from its class, its age, its size and the forwarding budget available in the coming window, producing a utility used for ordering, admission and eviction.

A portable library, bounded in memory and deterministic in execution time. Onboard decision engine for Rel-19 store-and-forward (NB-IoT/eMTC NTN, regenerative payload). **Patent pending: TR2026/006400.

What the operator gets

Scale up revenue without scaling up orbital investment

Goal oriented SLAs allow differentiated pricing

For the end user, the value is a class of service-level agreement (SLA) based on operational utility rather than raw delay or throughput—distinguishing, for example, a routine container status update that can wait from a cold-chain excursion whose value falls rapidly with delay. An operator today can sell coverage but cannot define tiers within it.

Example. GO NTN ensures a cold-chain alarm gets prioritized over routine container updates.

For the licensee

A component a payload or network software vendor can license and adopt as a software-only integration.

Portable core library

Admission, ordering, replacement, expiry and budget-constrained selection; deterministic execution path, bounded memory.

Reference integration

Integrated through a host-specific adapter without changing the decision core or external protocol behaviour.

Performance envelope

The break-even boundary, the high-gain operating region, and where conventional forwarding remains sufficient.

What we offer

Decision-engine licence

GO NTN, to vendors implementing Rel-19 store-and-forward.

Operator services

Goal-oriented SLA design, joint integration and pilots.

Pathway. B2B licensing and joint integration with NTN payload, base-station and network-software vendors.

Baselines: FIFO, the strongest implementable static-priority scheduler, and a priority-based release window with randomised spacing.

Interested in GO NTN?

info@freshdatatechnology.com

Contact Us →