Mission Planning Around Delay Tolerant Networking Constraints

The Problem

A lunar analog campaign generated 220 GB of passive acoustic data across a 12-sol push at Kilauea but had only 180 minutes of simulated relay contact per sol to deliver the resulting quilt back to Earth. The team discovered on sol 6 that roughly 70% of the quilt data they were generating could not fit inside the scheduled DSN-analog passes, which meant critical patches would not reach analysts until days after they were collected. The decision tree of which patches to ship, which to defer, and which to compress had no formal representation in the mission plan. The science lead ended up making ad hoc calls during each pass, and two of those calls deferred patches that turned out to be the highest-value data of the campaign.

This is the standard mission planning failure around Delay Tolerant Networking. NASA's DTN stack is store-and-forward by design, which means Earth never sees a continuous data stream from Mars or the Moon; it sees bundles delivered during contact windows. The NASA DTN tutorial documents the Bundle Protocol layer that governs how patches become deliverable units, and JPL's Interplanetary Overlay Network provides the reference ION implementation. None of this matters if the mission plan does not explicitly allocate patch-to-bundle assignments against DSN schedule windows.

The upstream scheduling bottleneck is DSN. The Deep Space Network's published operations show antenna passes are pre-allocated weeks ahead across competing missions, and JPL's Service Scheduling Software formalizes how missions bid for antenna time against each other. A team that generates data faster than its pre-allocated bandwidth cannot simply ask for more capacity in real time.

The bundle scheduling problem also intersects directly with teleoperated delay planning, since supervisory commands use the same DSN contact windows that science bundles compete for. A campaign that schedules supervisory commands without coordinating against the bundle plan ends up forcing science patches to defer to a later pass, which compresses the supervisory feedback loop the operator depends on.

The same constraint exists at every layer of the stack. Mars Relay Network proximity links between surface assets and orbiters like MRO and MAVEN allocate kbps-class capacity in scheduled passes, with the Electra UHF radio's published throughput defining the per-pass bundle envelope. Ground-side, DSN antenna pass auctions clear weeks before the data they will carry has even been collected, which means the bundle plan has to forecast generation rates against scheduling constraints that were locked in before the campaign started. A campaign that discovers mid-execution that its bundle plan exceeds its pass capacity cannot simply add capacity; it has to choose which patches to defer, and the deferral decisions ripple into the science narrative that the campaign produces. EchoQuilt's planner makes those tradeoffs explicit at planning time so they do not become emergency decisions at execution time.

The Solution: DTN-Native Bundle Planning

EchoQuilt's DTN-native bundle planner treats DSN passes as hard scheduling constraints and assigns quilt patches to outgoing bundles during mission planning rather than mid-execution. The planner consumes the DSN pass schedule (parsed from standard SSS exports) and the projected quilt generation rate from the mission's sol plan. It then produces a bundle plan that allocates each expected patch to a specific DSN pass window, with alternates for patches that might generate late.

The bundle plan is not rigid. It carries priority metadata on every patch, so if the sol produces an unexpected high-priority patch (for example, a newly detected skylight signature), the planner can bump a lower-priority patch to a later pass and re-pack the current bundle. The re-packing is deterministic and logged, so science leads reviewing the campaign can trace why each patch arrived when it did. This matches the discipline NASA formalized in its high-rate DTN work for future laser-comm missions: treat the network as a scheduled resource, not a best-effort pipe.

Each patch in a bundle carries compression and stitching metadata that lets Earth-side operators unpack the patch and fit it into the master quilt without waiting for adjacent patches. This is critical when bundle delivery gets reordered due to contact failures, because the Earth-side quilt has to be able to absorb partial updates without becoming inconsistent. The approach draws from JPL's ION implementation, which demonstrated custody transfer and bundle expiration semantics that EchoQuilt's planner respects natively.

The planner also handles DSN contention gracefully. When an antenna pass allocation shrinks due to higher-priority mission contention, the bundle plan degrades predictably: patches near known scientific features stay on the schedule, and peripheral or redundancy patches move to the next pass. This beats the default DTN behavior of dropping whatever happens to be at the tail of the queue.

DTN-native planning depends directly on link budgets because bundle-size decisions depend on link capacity calculations, and the link-budget analysis has to feed bundle planning rather than running as a downstream sanity check. EchoQuilt's bundle planner consumes link-budget outputs as a constraint layer at planning time.

Advanced Tactics

Pre-stage bundle fallback profiles for contingency passes. If a nominal 4-hour DSN pass is cut to 1 hour due to solar conjunction or antenna maintenance, the planner should have a prebuilt 1-hour bundle plan ready, not a real-time recomputation. EchoQuilt stores up to three degraded profiles per pass: nominal, 50% reduced, and emergency-only. Science leads approve each profile during supratactical planning, so the decision never has to be made during a reduced-contact window.

Tag patches with sol-local recency. A patch recorded 90 minutes before a DSN pass is more likely to contain time-sensitive information than a patch recorded 20 hours earlier. The planner weights recency into the priority sort so that late-breaking science gets into the next pass by default, which reduces the number of operator overrides needed during execution.

For analog teams running federated campaigns with ESA or JAXA assets, negotiate shared bundle schemas up front. DSN scheduling bottlenecks will intensify as Artemis-era missions compete for antenna time, and teams that have pre-agreed bundle formats can borrow each other's pass allocations in contingencies. EchoQuilt's bundle format is designed to be portable across agency stacks; adopting it during analog campaigns is the cheapest way to validate interoperability before flight.

Build a bundle replay log that captures every actual delivery against the planned schedule. The log records which patch went into which bundle, which bundle rode which pass, when the bundle was acknowledged by Earth, and any custody-transfer events along the way. Across a 30-sol campaign this log becomes the audit trail for the full data pipeline and lets science teams reconstruct exactly when each patch became available for analysis. The log also feeds back into the planner: patches whose bundles experienced delivery anomalies get flagged for re-validation when the next pass clears, which catches silent data corruption before it propagates into the master quilt.

Account for solar conjunction blackout windows in long-horizon campaigns. Mars-Earth solar conjunction occurs roughly every 26 months and produces a 2-3 week window when DSN cannot reliably command or receive from Mars assets. Campaigns that span a conjunction need to pre-plan stockpile bundles that will sit in the spacecraft buffer until conjunction clears, which means the bundle planner has to reserve bundle storage capacity that would normally cycle through delivery. EchoQuilt's planner accepts a conjunction-aware mode that pre-stages stockpile slots and re-prioritizes patch delivery as the conjunction window approaches, so the campaign continues collecting science even when downlink is impossible.

A cross-domain analogue from terrestrial conservation provides useful operational testing for the bundle planner. Our statewide networks work covers federated sensor networks that face similar store-and-forward constraints at smaller scales, and the patterns proven there for prioritizing data delivery across constrained network capacity transfer naturally into the planetary DTN problem. The terrestrial network operators have years more operational experience with store-and-forward scheduling than any planetary mission has, and EchoQuilt's planner inherits several of their practical heuristics.

Treat custody-transfer acknowledgments as observability events. The DTN bundle protocol's custody-transfer model means that when a bundle moves from one node to the next, the acknowledgment confirms storage responsibility has handed off, which is informational beyond a simple delivery confirmation. EchoQuilt's planner records these acknowledgments and uses them to track bundle flow through the relay network in real time. When custody transfers stall — bundles sitting in a relay's queue past their expected handoff window — the planner flags the relay for review and can re-route subsequent bundles to alternative paths if available. This kind of visibility is what makes a DTN-native pipeline operationally useful rather than just standards-compliant.

Ready to Plan Your Campaign Around DTN?

JPL and ESA mission planners scoping Mars or lunar analog campaigns need a mapping tool that respects DSN pre-allocation and DTN bundle semantics instead of assuming continuous downlink. EchoQuilt's bundle planner is built for that reality, with SSS schedule parsing, ION-compatible bundle formats, pre-staged fallback profiles, and conjunction-aware stockpile slot management for the 2-3 week solar conjunction blackout windows that recur every 26 months. Each pilot ships with three pre-built degraded bundle profiles per pass (nominal, 50 percent reduced, emergency-only), a custody-transfer acknowledgment monitor that tracks bundle flow through the relay network in real time, a sol-local recency weighting module sized to MRO and MAVEN proximity link cadences, and a bundle replay log format that captures every actual delivery against the planned schedule for post-campaign audit.

Pilot teams shape the conjunction-aware mode defaults, the cross-agency bundle schema that the 2027 reference release will adopt for ESA and JAXA federation, and the custody-transfer observability stream that NASA Glenn SCaN's high-rate DTN program is rolling into Artemis-era flight planning. Priority goes to JPL Mars cave concept teams scoping multi-sol DTN-native pipelines, NIAC PIs targeting Marius Hills or Mare Tranquillitatis pit campaigns, MatISSE proposers preparing TRL 5 advancement within DSN-contested windows, and ESA LunaNet relay designers coordinating Artemis-era proximity link standards. Funded analog teams running 12-sol or longer campaign concepts receive direct integration support from our DTN field engineering team during their 2026-27 deployment windows. Join the Waitlist for Planetary Analog Researchers to pilot bundle planning against your upcoming campaign schedule and help scope the interagency federation work on our roadmap.