Quantum technologies using laser-cooled atoms, such as cold atom interferometers

and atomic clocks, have revealed their potential in ground-based instruments. Un-

doubtedly, space-based quantum devices could provide orders of improvement under

micro-gravity environments. However, current instruments, including both terrestrial

and space-based, rely on complex laser systems, resulting in a high Size, Weight, and

Power (SWaP) and high-cost design. Overcoming this bottleneck could enable quantum

technologies for low-cost experiments and missions using small satellites. This would

open the door to new types of scientific experiments and missions, such as constella-

tions of quantum sensors, deep space optical communication networks, and cold atom

gyroscopes.

This research develops the Software Defined Laser (SDL), a compact laser system

architecture that applies Software Defined Radio (SDR) concepts to optical frequency

control. By combining reconfigurable Field-Programmable Gate Array (FPGA)-based

electronics with Radio Frequency (RF) techniques and optical modulators, SDL reduces

the number of seed lasers and optical components while maintaining the performance

required for quantum applications. We derive quantitative requirements for SDL by

analysing existing ground-based and space-based quantum instruments and develop a

highly configurable system-on-chip firmware to support diverse missions.

A Python-based SDL model is developed to evaluate RF technologies, frequency-

tuning topologies, and noise performance. Comprehensive noise analysis shows that all

investigated RF sources contribute less than 100 Hz to laser linewidth, satisfying Raman

and optical clock requirements. A novel frequency-tuning method is proposed, main-

taining sub-1 kHz linewidth while reducing tuning dead time from 5.8 ms to 200 µs.

To improve robustness beyond conventional PID control, SDL enables a multiple-

input multiple-output (MIMO) laser controller synthesised using H2 robust control.

Both single-input single-output (SISO) and MIMO controllers are evaluated under re-

alistic disturbances. The MIMO controller maintains laser frequency noise below 1 kHz

and achieves 0.02 % power stability under 17.5 K temperature variations over 6000 s,

outperforming state-of-the-art systems. To enable autonomous adaptation, an FPGA-

accelerated H2 synthesis module is developed, reducing optimisation time by 30 % and

enabling real-time controller reconfiguration during operation.