Erick Lansard

Global design of satellite constellations: a multi-criteria performance comparison of classical walker patterns and new design patterns

Abstract Basically, the problem of designing a multisatellite constellation exhibits a lot of parameters with many possible combinations: total number of satellites, orbital parameters of each individual satellite, number of orbital planes, number of satellites in each plane, spacings between satellites of each plane, spacings between orbital planes, relative phasings between consecutive orbital planes. Hopefully, some authors have theoretically solved this complex problem under simplified assumptions: the permanent (or continuous) coverage by a single and multiple satellites of the whole Earth and zonal areas has been entirely solved from a pure geometrical point of view. These solutions exhibit strong symmetry properties (e.g. Walker, Ballard, Rider, Draim constellations): altitude and inclination are identical, orbital planes and satellites are regularly spaced, etc. The problem with such constellations is their oversimplified and restricted geometrical assumption. In fact, the evaluation function which is used implicitly only takes into account the point-to-point visibility between users and satellites and does not deal with very important constraints and considerations that become mandatory when designing a real satellite system (e.g. robustness to satellite failures, total system cost, common view between satellites and ground stations, service availability and satellite reliability, launch and early operations phase, production constraints, etc.). An original and global methodology relying on a powerful optimization tool based on genetic algorithms has been developed at ALCATEL ESPACE. In this approach, symmetrical constellations can be used as initial conditions of the optimization process together with specific evaluation functions. A multi-criteria performance analysis is conducted and presented here in a parametric way in order to identify and evaluate the main sensitive parameters. Quantitative results are given for three examples in the fields of navigation, telecommunication and multimedia satellite systems. In particular, a new design pattern with very efficient properties in terms of robustness to satellite failures is presented and compared with classical Walker patterns.

Satellite Constellation Design: Searching for Global Cost-Efficiency Trade-Offs

For a long time, the optimization of satellite constellations has been formulated as the minimization of the number of satellites which satisfy a given geometrical coverage criterion (eg: continuous coverage of the Earth above a given minimum elevation threshold). Many authors theoretically solved this problem for single and multiple satellite coverage but without considering any other criterion. With the advent of large navigation and commercial telecommunication satellite constellations, cost considerations have become mandatory for any constellation designer eager to achieve the best global cost/efficiency trade-offs between the user requirements and the services provided by the operator. The paper proposes and discusses a multi-criteria approach that is highlighting and simultaneously handling three driving criteria in any constellation optimization process: coverage performances, operational availability and life-cycle costs of the system. A reference telecommuniction mission is taken as a case study to illustrate the power of the proposed approach with respect to the classical optimization process.

The Skybridge Constellation Design

The SkyBridge space segment, based upon Low Earth Orbit satellites, had to face a challenge: though using the same frequency band as various geostationary systems in order to profit by well-proven and cost-efficient techniques — i.e. Ku Band -, SkyBridge satellites must not in any way interfere with geostationary satellites. This “frequency sharing constraint” forces SkyBridge satellites to stop emitting (resp. receiving) towards (resp. from) any portion of the ground as soon as they are seen in alignment with the geostationary orbit from the point of view of this earth portion.

Mission analysis of clusters of satellites

Abstract An innovative satellite system that provides high precision localisation of beacon positions consists of a cluster of satellites, i.e. a group of satellites that maintain assigned positions at relatively short distances from each other. Compared to a single satellite, the interest of such a cluster lies in its ability to synthesise antenna bases much longer than those who can be physically mounted on one satellite. Each satellite of the cluster measures the time-of-arrival of the signal transmitted by the beacon. The derived time-differences-of-arrival (TDOA) are processed to estimate the beacon position. At first, this paper summarises the investigations performed on the localisation accuracy that have yielded the optimal cluster geometry. In a previous paper [E. Frayssinhes and E. Lansard, AAS paper 95-334 (1995)], Alcatel Espace has proposed a mathematical formulation relying on a strong analogy with GPS geometrical characterisation of navigation performances. The effects of geometry are expressed by geometric dilution of precision (GDOP) parameters. Such parameters are obtained by solving the TDOA measurement equations for the beacon position using an iterated-least-squares procedure. Then, the paper focuses at the system level on the peculiar problems that arise when such a satellite cluster system is dealt with, and more particularly the launch and early operations phases, the station-keeping strategies of manoeuvres, and the relative localisation and clock synchronisation of the satellites. In particular, it is shown that even with the “civil” C A GPS measurements, differential techniques can yield respective accuracies better than 5 m r.m.s. and 15 ns r.m.s.

How to improve the orbit of Doppler tracked satellites

Abstract In this paper, results obtained from the actual SEASAT data (Doppler and crossover residuals) are presented and discussed. In particular it is shown on the basis of external altimetric crossover tests that the achieved radial accuracy on 1 day should yield 30 cm-rms on average, which represents a significant orbit improvement with respect to classical semi-dynamical orbit computations.