Modeling Environments for Exploration of Business Dynamics within P-LEO Constellation Markets
Type of DegreeMaster's Thesis
MetadataShow full item record
Due to recent investments in the space industry, companies have begun to further explore the business feasibility for space-based systems. This rapidly expanding space resources economy includes proliferated low Earth orbit (P-LEO) satellite constellations which provide a variety of services such as broadband internet and telemetry to a diverse set of customer types, for multiple use cases. Although constellation operators hope this service is a financially lucrative business, uncertain market conditions coupled with limited resources create a techno-economic environment in which large capital expenditure and competition reduce the likelihood of profitability. Strategic development of P-LEO constellations is a central pillar that links orbital dynamics of constellations to downstream business performance. Since different constellation development strategies can lead to business success or failure, understanding the environment in which P-LEO constellations are developed is of paramount importance for any such business endeavors to be profitable. When considering factors such as space debris, competition, socio-economic variations, and geographical demand, this becomes an even more challenging environment to navigate. Although comparisons of P-LEO constellations have been performed and individual constellation designs have been critiqued, modeling competitive business dynamics and complex resource management strategies involving such large space systems have not been fully explored. To study these complex interactions and understand how satellite internet business strategies evolve with environment complexity, business environments with varying levels of modeling fidelity are created. Two different approaches are taken to model the complex P-LEO constellation environment: dynamic programming and gamification. A discrete, deterministic grid world formulation is used to develop simplified satellite communications business environments which utilize dynamic programming to generate optimal strategies. Such strategies gleaned from this environment demonstrate several concepts such as: optimal timing of investments, coupling of design variables with actions, and optimal resource allocations. However, this environment is limited in terms of modeling elements and strategy complexity, so a gamified, multi-agent environment is also developed. Using a mixture of gamification and modeling, the environment is constructed as a multi-player, real-time-strategy (RTS) simulation game to simulate business competition between constellation operators. This expanded environment provides opportunities to explore competitive and cooperative strategies in a high dimensional environment. As modeling fidelity increases, solving for an optimal strategy becomes increasingly infeasible. This is due to two factors: the rapid increase of the P-LEO environment's dimensionality and the increased complexity of processes modeled within the environment. Through qualitative comparisons of environments and investigations between action couplings, we can understand the dynamical, agent-environment system of P-LEO constellation markets at different levels of complexity.