ASI-407 Galassia 5 Payload Radio Subsystem Team
(CDE4301 Final Report)

1. Galassia 5 Mission Overview

1.1 Problem Introduction

In today’s satellites that are used for earth observation (EO), data latency limits the scope of activities that EO data can be used for. Existing missions typically rely on a dedicated ground facility - often the size of a building - which receives the raw data from the satellite and processes it. This processing often consists of various corrections and extracting the relevant features from the image. The extracted features must be then sent to the end user. End to end, this can be a lengthy process and it may be difficult to send information to users that are in remote areas (like in the middle of the ocean). This limits real-time responsiveness in applications such as disaster monitoring or maritime surveillance.

Furthermore, the link used to send the raw data from the satellite to ground often occurs in X band, which is increasingly being congested due to the number of satellites in orbit.

To address these challenges, Galassia 5 is designed to demonstrate an AI-enabled, real-time Earth-observation capability, supported by a direct to user Ku-band downlink system.
The mission aims to overcome traditional data-handling bottlenecks by performing on-orbit image processing and establishing direct high-speed downlinks to portable VSAT terminals.

1.2 Program Heritage

The Galassia satellite program is the flagship series of undergraduate built CubeSat missions by the National University of Singapore (NUS) with the goal of: 1) Developing a talent pipeline for the local space sector, 2) Serving as testbed for frontier technologies that may be of interest to the national ecosystem. The heritage of the program is as follows:

• Galassia (2015):
  A 2U CubeSat (~2 kg) launched for ionospheric observation and quantum entanglement experiments using the SPEQS payload.

• Galassia 2 (2023):
  A 3U CubeSat (~6 kg) carrying a multispectral agricultural Earth-sensing payload for precision imaging.

• Galassia 5 (Planned 2027):
  A 6U CubeSat (~12 kg) featuring on-orbit AI image processing and a Ku-band radio payload.

  1.3 Mission Objectives

The primary objectives of the Galassia 5 mission are:

1. Demonstrate real-time imaging and onboard AI analytics in a 6U cubesat
2. Demonstrate the use of a Neural Processing Unit in support of AI enabled real time imaging.
3. Validate a direct-to-VSAT Ku-band downlink for high-speed, low-latency data transfer between the satellite and portable ground stations.

These goals collectively aim to advance Singapore’s capability in responsive, data-driven space operations.

1.4 Value Proposition and Possible use cases

By developing the real time imaging technology, and facilitating direct to user downlinks, Galassia 5 seeks to unlock the following value added features:

1. Smart Earth Observations

   Onboard AI analytics extract only essential insights from captured imagery, drastically reducing redundant data and enabling faster decision-making.

2. Accessibility and Spectrum Efficiency

   Operating in the Ku-band allows for smaller ground antennas for an equivalent gain and reduces reliance on congested X-band channels. This increases portability and the possibility of taking an antenna to

3. Reconfigurability

   Galassia 5’s AI payload can be reprogrammed in orbit through a high speed payload data uplink, which enables the satellite to be used in accordance to varying mission needs.

The combination of reconfigurability and smart, near real time earth observation has the potential to support multiple real-world applications:

• Humanitarian Rescue: Rapid identification of collapsed infrastructure can assist first responders in targeting their rescue efforts, saving valuable time and maximising the chance of survival for victims

• Forest Fire Surveillance: Real time detection of fire outbreaks can help identify which areas should be prioritised for fire fighting efforts, and may also provide responsive data to predict the future spread of the fire.

• Anti-Piracy Surveillance: Real time maritime monitoring for situational awareness over large oceanic regions can help vessels identify the possibility of piracy attacks. Data may be correlated with Automatic Identification of Ships (AIS) beacons for an added layer of assurance.

1.6 Concept of Operations (CONOPS)

Galassia 5 is scoped around a possible piracy surveillance use case. The simplified Concept of Operations is as follows:

1. Galassia 5 overflies a user vessel equipped with a Ku-band ground terminal.

2. The vessel transmits an Area of Interest (AOI) request to the satellite.

3. The satellite captures and processes imagery of the AOI.

4. The onboard AI identifies surrounding vessels and their positions.

5. The processed results are downlinked directly to the vessel via Ku-band.

A detailed mission CONOPS is available in Appendix B. In the current implementation of the mission, it is not possible to perform a Ku band uplink from the user terminal, as the hardware around which the Ku band subsystem does not support uplink on the Ku band side. Instead, real time imaging operations will be scheduled during a pass over a ground station with an uplink capability. In addition to the mission mode, a housekeeping operation is also provided for general administration of the satellite

1.7 Payload System Architecture

The payload is the portion of the satellite which performs the sensing, AI acceleration, and data transmission functions. This is the core area that the NUS team focuses on building. The key components of the payloads are:

• Payload Computer (PLC): MPSoC-based single-board computer integrated with a Hailo 8 AI accelerator (PCIe Gen 2 ×4).

• Optical Imager: The optical imager is the Simera Monoscape100, and is the main sensing instrument used for the mission. The imager is primarily interfaced over Spacewire, with an Secondary I2C control interface and a USART data interface

• Payload Radio: Consists of the Ku band payload downlink, as well as the secondary X band, and S band links. The radio system is interfaced to the rest of the satellite through a Spacewire or RS422 data interface. Control may be provided through a CAN bus.

• Bus Interfaces: Electrical, mechanical, and data pathways connecting the payload to the satellite bus. These provide the necessary supporting infrastructure.

  1.8 Stakeholders and Partners

The Galassia 5 program works with various project stakeholders and partners. In general, program stakeholders are organizations that support the program financially or in kind, and drive its scope and key objectives. Project partners provide support either out of contractual obligation or goodwill, but generally do not drive the outcomes of the program.

1.8.1 Project Partners

1.8.2 University Partners

1.9 Ku-Band Development Rationale and Motivation

Operation in Ku band was given as a mandate from MINDEF FSTD as a core technology to be demonstrated. As part of that mandate, an outcome that was expected was that the ground segment should consist of equipment that could be man portable.

In today’s paradigm, X band is used primarily for high speed payload data transfer. The high speed of transfer typically means that most X Band signals are of a higher bandwidth, and the increasing number of satellites result in increasing spectrum congestion. To maintain a stable link, parabolic reflectors (what would be normally called a ‘satellite dish’) are typically used to increase the gain and hence the link stability of the payload data downlink. As the gain of a dish increases proportionately to the frequency of the received signal, a Ku band signal will be able to produce a higher gain for the same dish size. The ability to miniaturise the dish allows the ground segment to be portable, which is in line with MINDEF FSTD’s mandate.

To identify a suitable frequency in the Ku region to operate in, the Infocomm Media and Development Authority (IMDA) was consulted. As the spectrum regulator in Singapore, IMDA is required to approve and grant the mission the required spectrum licenses needed. To facilitate cost reductions, IMDA allocated the mission a small range in the Earth-Exploration Service (EES) downlink band between 13.93 and 13.99 GHz. Operating in this band, in contrast to the usual Commercial SATCOM band (11.5 - 12.75 GHz) allows the frequency coordination fee - which is nominally 80K SGD - to be waived, therefore achieving significant cost savings in the realization of the mission.

The key drawback to operating in the EES region is the lack of readily available commercial hardware that can be used to support the mission. This is true of both the space and ground segments. An example of what is presently available on the market for both space and ground segments respectively are shown above. For downlink, these hardware are configured, designed, and licensed to operate in the commercial SATCOM band, rather than the EES band, which makes them unsuitable for the present use case. Furthermore, given that operating in the higher frequency microwave band is relatively novel for the case of cubesatellites, there is also a general lack of companies that make quality space segment hardware. This is also compounded by the fact that higher microwave frequencies are harder to work with - and typically require higher powers and greater care in their design.

As a result, a decision was undertaken to design, build, and integrate space and ground segment hardware that would operate in the required frequency bands that were allocated to us. In addition,