Link Budget Considerations for Relayed Map Data
The Link Budget Stack From Pit Floor to Earth
A Mars cave cartography team typically thinks about three link segments: rover-to-relay (Electra UHF proximity), relay-to-Earth (X-band or Ka-band to the DSN), and the DTN layer that holds data when either of those segments is unavailable. Each segment has an independent failure mode, and a robust concept has to account for the joint probability that any single segment is degraded during a given pass window. The DESCANSO MRO telecommunications chapter reports Electra proximity rates from 1 kbit/s to 2 Mbit/s depending on geometry. NASA's DTN overview describes the store-and-forward architecture that makes these intermittent, variable-rate links usable for science return.
The Mars Relay Network page describes data returned roughly 12 times per week at gigabit-class aggregate volumes, but a single rover's share of that capacity is small. The Wikipedia page on the NASA Deep Space Network lays out the ground architecture and the scheduling constraints that a cave mission inherits. Inside a pit, the rover-to-relay link itself attenuates fast: arXiv work on path planning and navigation inside off-world lava tubes and caves shows that in-cave link attenuation forces a multi-hop relay even for modest cave depths.
The DSN-side scheduling constraint compounds the rover-side bandwidth limit in ways that surprise concept teams used to surface-mission planning. The Mars-orbiter constellation shares a finite number of DSN dish-hours per week, and a cave concept arriving with a new payload competes against existing missions for that allocation. Concept teams that assume their proposal will receive median DSN allocation typically discover during Phase A that competing missions have higher historical priority, and the available pass schedule looks tighter than the planning numbers suggested. Sizing the cartography pipeline against the worst-case rather than the median DSN allocation produces a more defensible Phase B operations plan and avoids the late-stage rebaseline that has cost several recent cave concepts substantial schedule slip. EchoQuilt's Mars relay features descriptor format was specifically designed to scale gracefully across this allocation range without requiring a different decoder for the worst-case versus nominal scenarios.
How EchoQuilt Sizes Quilt Patches for the Multi-Hop Link
EchoQuilt sizes patches against a target Electra pass rather than against a maximum bit rate. The default configuration assumes a 7-minute pass at a working rate of 128 kbit/s, which carries about 6.7 MB per pass. A typical traverse produces hundreds of quilt patches at 8-32 kbit each, which fits comfortably inside that budget with margin for retransmits. The pipeline scores patches by information gain, queues them in priority order, and ships them during the pass window. When the rover moves deep into the cave and the direct rover-to-relay link fails because of regolith and rock attenuation past the skylight, a secondary hop is needed.
ScienceDirect's RoboCrane paper on power and communication links between lunar surface and caves describes a crane relay positioned at the skylight to bridge the attenuation gap. EchoQuilt integrates with this topology: a sky-side relay node aggregates quilt patch queues from in-cave nodes over short-range link (BLE or UWB), holds them under DTN semantics, and forwards them during the next Electra pass. The quilt survives long gaps because patch deltas are small, idempotent, and prioritized by information gain, which means a multi-sol comm blackout does not destroy the early science return; it just delays it.
The multi-hop architecture also lets a single skylight relay serve multiple in-cave platforms simultaneously. A Marius Hills concept that envisions both a CADRE rover formation and a tethered Axel descent unit can share the same sky-side relay, with each platform's patch queue prioritized independently. This sharing is invisible to the Electra link, which sees a single uplink from the relay node, and to the DSN, which sees a single Mars-side correspondent to schedule against. The architecture has the side benefit of letting a follow-on mission inherit the existing relay node without replacing it.
A companion post on DTN constraints covers the specific DTN bundle layer EchoQuilt uses, including bundle expiry rules and custodial transfer between in-cave and sky-side nodes.
A practical implication of the DTN bundle layer is that operations teams can adjust the priority weights for queued patches mid-mission. If the science team decides that a particular cave region warrants higher coverage density (because an early patch suggested an interesting wall feature, for example), a single command on the next uplink updates the patch scoring weights, and the rover begins prioritizing that region's patches in subsequent downlinks. This responsiveness is hard to achieve with a fixed-size return queue and impossible with a raw-data pipeline; it relies on EchoQuilt's patch-as-decoded-science design and the asynchronous queue management that the DTN layer enables. JPL operations teams accustomed to multi-week ground turn-around appreciate the shorter loop because it lets the science team steer the campaign in something close to real time, even with a multi-sol round-trip light-time on top.
Advanced Tactics For Link-Budget-Aware Campaign Planning
Three tactics improve mission return under a tight link budget. First, log patch ages and track the distribution of time-to-downlink across the traverse. A campaign where patches arrive on Earth within 48 hours of acquisition produces tighter science loops than one where patches sit for a week in the DTN queue waiting for a clear pass. EchoQuilt reports this distribution as a first-class metric so campaign leads can see bottlenecks early.
Second, model the link budget with real pass geometries from the target mission rather than nominal values. Electra passes at Marius Hills latitude differ from passes at an equatorial Arsia Mons site in duration, elevation profile, and frequency of unfavorable geometry. The quilt pipeline accepts a pass ephemeris file derived from the actual orbiter ephemerides and sizes its queue accordingly across the planned mission window. Concepts that skip this step routinely overestimate their return capacity by 30-50 percent and have to rebaseline mid-Phase B.
Third, plan a reduced-fidelity ground mode for degraded link conditions. If the working rate drops to 1 kbit/s, the quilt should still return a minimal spine (major apertures, rough wall positions) rather than blocking until a better pass. EchoQuilt's pipeline supports named fidelity tiers that trade patch detail for patch count under tight budgets. This is the single feature most often requested after a team's first bandwidth-starved analog campaign.
A fourth tactic that compounds the previous three is to instrument the relay link with a quiet-channel probe during off-pass intervals, drawing on the data-logger pairing cadence bat hibernacula teams use when retrieval windows are unpredictable. Even when the primary Electra link is unavailable, the rover can transmit a low-power beacon that a passing orbiter may opportunistically detect during otherwise unallocated time slots. JPL's heritage with opportunistic relay returns from CubeSat constellations and the deep space Bundle Protocol Version 7 testing on the EM-1 mission has shown that these unscheduled returns can carry several percent of additional science data per sol, and EchoQuilt's bundle layer handles them transparently because the patches are already idempotent and DTN-aware.
Concept teams that include opportunistic relay handling in their mission ops plan typically see a 10-20 percent uplift in usable science return on bandwidth-constrained missions, which is large enough to be worth defending in a Phase B operations review.
CTA
EchoQuilt is pairing with JPL telecom planners, ESA LunaNet teams, MatISSE proposers, and NIAC flight concept leads who need a cartography payload that fits inside an Electra UHF relay schedule, a DTN-mediated DSN return, and a multi-hop in-cave relay topology without losing science return to scheduling friction. We provide the link-budget emulator, the pass ephemeris integration, the opportunistic relay handler, and the fidelity-tier configuration library. Join the Waitlist for Planetary Analog Researchers to validate a link-budget profile against your mission's actual relay schedule and DSN allocation, with worst-case as well as nominal scenarios. Teams working on Artemis-era surface infrastructure, LunaNet integration, or Mars Sample Return-adjacent cave concepts get priority review and direct integration support from our telecom engineering team.