École Nationale Supérieure de l'Électronique et de ses Applications (also known as ENSEA) is a graduate school (grande école) of electrical engineering and computer science, located in Cergy (in Val d'Oise department) close to Paris in France. It was founded in 1952 under the name of ENREA and became ENSEA in 1976.
This paper addresses the evolving role of Nongeostationary (Non-GSO) satellite systems, specifically focusing on homogeneous constellations in either low Earth orbit (LEO) or medium Earth orbit (MEO). These systems, known for their low latency due to their proximity to Earth, incorporate advanced developments in payload and antenna technologies including adaptive radio resource management and beam forming capabilities. The study investigates key issues in multi-beam satellite systems, including beam footprint design, user-to-beam assignments, and the optimal distribution of radio resources. We introduce a novel problem aimed at optimizing beam placement and balancing user loads in Non-GSO satellite systems, which results in a mixed-integer non-convex formulation. For practically appealing applications, we relax the binary variables and propose the continuous relaxation problem using the inner convex approximation framework and a penalty method to further speed up the convergence. The proposed method addresses a complex, previously unexplored non-convex and NP-hard optimization problem, offering a practical and efficient polynomial-time solution. Experimental results underscore the effectiveness of our approach, demonstrating significant improvements over existing benchmarks and providing new insights into the optimization of satellite communication systems
Finding the cost-optimal number and placement of ground stations of non-geostationary satellite constellations is a complex problem whose solution significantly impacts the quality of the satellite service. The problem is made even more complex when accounting for the presence of rain attenuation and inter-satellite links. In this work, we present a methodology to address this problem, which consists of the following. First, a model to compute the average terminal capacity depending on the maximum inter-satellite link hop count is presented for a specific ground segment placement. Then, a Monte Carlo sample generator and estimator are proposed to model the effect of rain attenuation. The optimal set of ground segment placements that balances deployment cost and average terminal capacity is found using a multi-objective genetic algorithm called the non-sorted genetic algorithm II. Finally, numerical results for worldwide coverage in V band are provided. The proposed method showed satisfying convergence despite the computational complexity of the scenario. The results demonstrate that increasing the maximum intersatellite link hop count from 0 to 4 strongly reduces the ground segment cost for a given average terminal capacity (by up to a factor of 30) and significantly increases the average terminal capacity for a given ground segment cost (by up to 108%). Moreover, the increase in the maximum inter-satellite link hop count also significantly decreases the terminal capacity disparities between high-demand regions and low-demand regions, assuming inter-satellite links are used for load balancing.
Phased arrays offer high gain for User Terminals (UTs) in Non-Terrestrial Networks (NTNs), but they suffer from prohibitive costs and complexity. Recently, Time-Modulated Arrays (TMAs) have been proposed as a low-complexity alternative; however, they face challenges in harmonic control and limited gain and bandwidth. Motivated by the LEO satellite direct-to-UT scenario and the need for mass deployment of low-cost UTs, this work presents a low-complexity, high-gain TMA. We study multilevel switching (MLS) TMA, aiming to enhance both gain and effective bandwidth while minimizing complexity. We design a discrete sinusoidal MLS integrated with a Quadrature Time-Delay (QTD) feeding network to maintain continuous element activation and focus power on the main beam. Further, we introduce an exponential MLS and optimize it to enhance gain. The sinusoidal fit offers superior spectral purity and the exponential fit provides better gain. The proposed TMA offers a low-cost, low-complexity antenna solution, with only a 1.5 dB gain deficit compared to an equivalent phased array, which is promising for emerging UTs.
Activities such as the design and optimization of satellite constellations for telecommunications, communication solutions engineering, or GEO/non-GEO satellite inter-system coexistence applications benefit from methods to compute changes in Earth station equivalent noise temperature, a product of the dynamic nature of link geometry and the propagation channel. This paper addresses said requirement. Starting with a review of the user terminal G/T and Earth station noise temperature, methods are developed to compute the change in noise temperature introduced by varying elevation angle and propagation channel conditions. These methods are useful for engineering analyses when the noise contributions from each component of the Earth station receive chain are not known, especially considering the increased impact of clouds and gasses on GEO and non-GEO links operating in frequencies of the Q, E, and W bands. Examples of use and benefits are provided, as well as considerations on the applicability and validity of the methods.
Due to its proximity to Earth and cost-effectiveness, Low Earth Orbit (LEO) satellite constellations have emerged as an attractive infrastructure for high-throughput and global connectivity. However, the rapid satellite motion and limited visibility windows require agile beamforming, low-complexity, low-power-consumption antenna solutions on the User Terminal (UT) side. Traditional Phased Antenna Arrays (PAA), while powerful, are often complex, costly, and power-hungry. In this work, we present an alternative based on Time-modulated array (TMA) to reduce the cost, complexity, and power consumption of UT. We first present the fundamentals of TMA for UT. We next compare it with the PAA, considering performance and power consumption trade-offs, and provide insights into optimal TMA array size design. Finally, this work lays a roadmap for future research related to TMA-UT, including advanced switching, pulse shaping, and AI-driven optimization approaches.