BOUNDS ON SIGNAL PERFORMANCE REGARDING MULTIPATH-ESTIMATING DISCRIMINATORS

After lengthy negotiations, the US and the EU agreed on having BOC(1,1) as common signal for the open service of E1/L1, achieving thus the best possible degree of interoperability in this band [1]. During this long process of consultations and analyses on both sides, not only the concerns on the interference with the GPS M-Code and Galileo PRS played an outstanding role, but as it could not be different, the performance of the potential studied solutions was one of the main drivers since the very beginning. As we know, the positioning performance of a signal depends on the contributions of many different sources of error, among which the ionospheric error and multipath are the most relevant. But with a big difference: while the ionospheric error can be removed to a high degree with dual-frequency measurements, the multipath is completely unpredictable. Multipath is indeed the last frontier when we talk about mitigation of the different sources of error. Thus, if we want to improve the positioning performance of a signal, the direction to go is finding optimized signal structures with high potential to eliminate multipath. If we take a look at the process how the different Galileo signals have been selected, we can clearly recognize that the search for the best possible performance has been the common denominator in all the works carried out in the last years. Clear proof of this is the resulting CBCS candidate signal for E1 OS that was proposed by the Galileo Signal Task Force of the European Commission in 2005 [2]. Nevertheless, since the Agreement [1] made it desirable for GPS L1C and Galileo E1 OS to have the same spectrum, a joint EC/US working group has recently presented an alternative solution that could be adopted by both parties [3][4][5]. In consonance with this and given the importance that the signal structure has played in the design of Galileo and the works that are already being done on the modernization of GPS and even for the next generation of Galileo, this paper will try to give answer to the fundamental question of setting bounds on signal performance and even more important, of establishing fair comparisons between signals. In a final step, we will prove that indeed, the MBOC signal waveform proposed in [3] is the best possible solution in terms of performance for the Open Service (OS) signals of E1/L1 under the constraints of the Agreement [1].