Building the Lunar Information Utility

Abstract

Lunar missions should not each carry a private communication and navigation architecture. I develop an information utility that combines direct Earth links, lunar relays, local mesh networks, delay-tolerant storage, position and timing services, and space-weather alerts through common interfaces. NASA, ESA, and JAXA's LunaNet Interoperability Specification provides a concrete foundation for cooperating service providers. I extend that logic into an operating architecture with explicit service levels, integrity bounds, traffic priorities, local autonomy, and degraded modes. Link margin alone is insufficient. The network must preserve important data through interruption, prevent one failed or misleading node from corrupting navigation, maintain a traceable time reference, and support vehicles that cross between providers. The desired result resembles terrestrial infrastructure in usability, but it must remain honest about delay, obstruction, radiation, power scarcity, and incomplete coverage.

Stop designing every mission as an island

A conventional spacecraft may establish a direct link to Earth, navigate through ground tracking and onboard sensors, execute a planned sequence, and store data until a contact. That model works for discrete missions. It becomes wasteful and fragile when landers, rovers, habitats, cargo vehicles, observatories, and crew all operate in the same region.

The lunar far side cannot see Earth directly. Polar terrain produces local occultation. Rovers cross ridges and craters. Directional antennas require pointing. High-rate science competes with command and health traffic. Ground networks have finite apertures and schedules. A shared service can aggregate demand, provide local routing, distribute time and navigation, and let users state requirements rather than own every relay.

LunaNet is important because it defines a framework for cooperating networks rather than a single government constellation. Its service concept includes communications, position, navigation and timing, and situational awareness. That makes interoperability the architectural center.

Layer the network by distance and urgency

I use four layers.

Layer	Primary links	Main function
Earth backbone	Ka-band, optical, and ground networks	High-rate return, command, and terrestrial time reference
Lunar relay	Orbiters and strategic surface towers	Coverage around terrain and the lunar limb
Local access	Radio and optical mesh	Rovers, suits, instruments, docks, and habitats
Delay-tolerant overlay	Persistent storage and scheduled forwarding	Data survival through partitions and changing routes

Optical communication can provide high rates but demands fine pointing and is affected by clouds at Earth terminals. Radio offers different availability and beam characteristics. I want multiple ground sites and complementary link types rather than one nominal channel.

The link equation screens received power:

$$P_r=P_tG_tG_r\left(\frac{\lambda}{4\pi R}\right)^2L,$$

where $L$ collects losses. Range loss grows with $R^2$, but a usable service also depends on bandwidth, noise temperature, coding, atmospheric conditions, pointing, interference, and availability. Every published rate needs an associated geometry and probability.

Networking must expect interruption

The terrestrial Internet often assumes a near-continuous end-to-end path. Lunar links will be scheduled, obstructed, power-limited, or unavailable. Delay and Disruption Tolerant Networking stores data at intermediate nodes and forwards it when a path appears.

That changes node design. Storage is part of the network, not a payload convenience. Custody policy determines when a sender may safely delete a bundle. Priority prevents bulk imagery from displacing medical telemetry or collision warnings. Expiration removes data that has lost operational value. Authentication and integrity protect commands and products across long custody chains.

I define four traffic classes:

• emergency voice, hazard, and survival telemetry;
• command, navigation, timing, and vehicle coordination;
• engineering health and time-sensitive science;
• delay-tolerant bulk science, software, and archives.

The network must reserve capacity for the first two classes. A high average throughput is meaningless if a bulk transfer blocks a time-critical alert.

Reach EarthRadio and optical backbone

Connects lunar traffic to diverse terrestrial terminals and archives.

Reach terrainOrbital and surface relays

Carry service around the limb, ridges, craters, and local shadows.

Preserve dataDelay-tolerant storage

Maintains custody and priority when no end-to-end route exists.

Know statePNT and alerts

Distributes position, navigation, timing, weather, and integrity.

Position needs an integrity statement

A coordinate without uncertainty can be dangerous. Rovers need local obstacle and route accuracy. Landers need high-integrity relative navigation near touchdown. Scientific instruments need traceable location. Orbiters need stable reference frames. Different tasks can use the same service only if its products carry uncertainty and alert limits.

For pseudorange measurement $i$, a simplified observation is

$$\rho_i=\lVert\mathbf{x}-\mathbf{x}_i\rVert+c(\delta t-\delta t_i)+b_i+\epsilon_i,$$

where user and transmitter positions are $\mathbf{x}$ and $\mathbf{x}_i$, clock offsets are $\delta t$ and $\delta t_i$, $b_i$ is modeled bias, and $\epsilon_i$ is residual error. Geometry determines how measurement error maps into position and clock error.

The utility should publish ephemeris, clock, frame, covariance, health, and integrity alerts. Users should fuse these signals with inertial measurement, terrain-relative navigation, optical landmarks, inter-vehicle ranging, and local beacons. No single constellation should be allowed to become a silent common cause.

Timing deserves equal discipline. Power grids, scientific observations, navigation, communication slots, distributed control, and event reconstruction depend on a common time scale. Holdover clocks must bridge outages, and their drift must be bounded. A restarted node cannot simply announce a plausible time.

Local autonomy is not disconnection

Earth-Moon propagation delay is modest compared with Mars, but ground availability and terrain still prevent continuous supervision. Vehicles should negotiate local traffic, avoid collisions, protect power, route around faults, and preserve data without waiting for Earth.

Autonomy needs bounded authority. A relay may reroute bundles and allocate ordinary capacity. It should not discard protected evidence, accept untrusted software, or change safety separation without controlled rules. Every automated decision should leave a trace that operators can reconstruct.

DetectLoss or misleading data IsolateSuspect link or source HoldPreserve data and local time RouteUse alternate service ReconcileMerge records after recovery

Measure a service, not a satellite count

I would publish coverage probability, latency distribution, assured minimum rate, delivery probability by traffic class, position error bound, integrity risk, time error, alert delay, outage recovery, and storage reserve. A constellation can have many nodes and still fail at a polar worksite. One excellent link can still be unavailable when its only ground station is under weather.

The utility becomes infrastructure when a new rover can declare its needs, authenticate, obtain time and position, discover routes, send protected telemetry, and preserve bulk data without a bespoke ground architecture. That capability will do more for sustained exploration than an isolated increase in peak downlink rate. It allows every later machine to begin with a shared sense of where it is, when it is, what is happening, and how to be heard.