Direct-to-Cell

 

Direct-to-Cell (D2C) - Procedures

A phone has lost terrestrial coverage and a satellite signal is available. What happens before a message can be delivered? The radio connection, subscriber access and application transaction are separate stages, and each can fail independently.

This page uses a simplified conventional LTE/EPC example. It is not a captured trace from a vendor, and it does not prescribe an identical procedure for every D2C deployment. See Device for compatibility prerequisites.

High-Level Access and Service Sequence

The sequence below groups the main steps of an initial connection. The radio endpoint may be on the satellite or on the ground behind a relay. The core column groups several functions rather than representing one physical box.

Illustrative LTE access, attachment and application exchange over satellite accessPhoneSatellite access / eNBEPC functionsServiceCell synchronization / broadcast informationRandom access and RRC establishmentNAS Attach Request via eNBAuthentication / NAS security exchangesContext and bearer setup (grouped)Attach Accept / radio configuration (grouped)Attach completion (grouped)Application request over established pathApplication response / delivery indicationConceptual ordering; grouped exchanges, omitted internal messages, no timing scale.

Figure 1 : Radio access and network attachment precede the illustrated application exchange

The baseline is TS 36.300 for radio access and TS 23.401 for EPS attach. Authentication and context reuse depend on state. This example assumes initial attachment with a packet-data bearer; a previously attached device may use a service request instead.

Finding and Selecting a Cell

Before diagnosing a failed message, establish whether the phone found a usable cell. A status-bar label is helpful for the user, but an engineering test needs the selected network, frequency and available radio measurements.

The handset must acquire the cell and read the information needed for access. Network selection and subscriber policy determine whether it can use that cell. Satellite visibility alone does not satisfy these checks.

Record the time terrestrial service disappears, the time the satellite cell is first detected and the time access begins. A long interval between these events may be search or selection delay rather than slow packet transport.

Radio Access and Network Attachment

In the ordinary contention-based LTE example, random access includes preamble, response, scheduled uplink and contention resolution. RRC establishes the radio connection. NAS attachment then registers the subscriber with EPC, with security and bearer handling as applicable.

Where Did the Attempt Stop ?

Use the last successful stage to narrow the investigation. The table is a diagnostic guide, not a one-to-one mapping between a symptom and its cause. Several failures can produce the same visible behavior.

Observation

Useful next checks

No cell detected

Band, sky visibility, receiver signal and search configuration

Cell detected; no access response

Preamble reception, uplink power, frequency/timing error and response timing

Radio connection succeeds; attach rejected

NAS cause, subscriber provisioning and permitted network

Attached; application times out

Bearer, routing, service policy, server availability and coverage continuity

A reject message is different from silence. Preserve the decoded cause when available. If no response arrives, check the corresponding network log before deciding whether the message was never received, rejected internally or lost on its return path.

Sending and Receiving Messages

A messaging interface hides several possible delivery paths. SMS and an IP messaging application do not necessarily use the same network service. The product must specify which path is available and what its user-visible acknowledgements mean.

For a test message, record four separate times: user presses send, device submits the message, network or server acknowledges it, and the destination receives it. A local sent indicator may refer to an earlier stage than final delivery.

Incoming Reachability

An incoming transaction requires the network to reach the device or retain the transaction until it can be reached. Coverage gaps, idle behavior and application server policy all affect when the user receives it.

Do not assume every satellite stores messages onboard. Buffering may occur at the application server, messaging platform or another network node. Identify the actual queue and expiration policy before attributing delayed delivery to satellite store-and-forward operation.

What Happens When Coverage Moves ?

A stationary phone can face changing radio conditions because the satellite moves. But a changing satellite or transport path does not necessarily mean the phone sees a new logical cell.

LTE distinguishes idle-mode reselection from connected-mode handover. In a conventional handover example, measurements inform the network decision; the source commands the move and UE accesses the prepared target. The actual implementation determines which transitions are exposed to the handset.

If service disappears without a usable target, continuity cannot be assumed. The phone may need to recover the radio connection or reacquire service. An application may retry afterward, but successful retry is not the same as an uninterrupted call.

Worked Example: Delivery Time Is Not Radio RTT

Consider an illustrative message submitted at t = 0 s while no usable service is available. A usable cell appears at t = 40 s; search and access take another 8 s, and delivery then takes 2 s.

Event

Time

Message submitted by user

0 s

Usable coverage begins

40 s

Access and service readiness completed

48 s

Destination receives message

50 s

The end-to-end delivery time is 50 s. Only 2 s occurs after service readiness in this example. Reporting the result as a 50 s satellite propagation delay would be incorrect. The 40 s coverage wait also must not be omitted when describing the user experience.

The values are invented to separate the stages. Use synchronized clocks or a clearly defined single-clock measurement in a real test. Continue with Performance for delivery ratios, timeouts and latency reporting.

References