METHOD FOR TRACKING A TRANSMITTER BY MEANS OF A SYNTHETIC SPARSE ANTENNA NETWORK The invention relates to a method of locating one or more transmitters stationary or mobile on the ground on the basis of the running of a carrier and of an onboard sensor producing the associated direction vectors. It is applied for example with an array of synthetic sparse antennas. The prior art describes various procedures for locating one or more transmitters on the basis of running carriers. Figure 1 illustrates an example of airborne locating. The transmitter 1 to be located is at the position (x0,yo,z0); the carrier 2 at the instant & is at the position (x*,y/(,zk) and sees the transmitter at the incidence (0((it,x0,y0,Z0), J(/tox0,yo,zo)). The angles 0(t,Xo,yo,Zo) and A(t,Xo,yo,Zo) evolve over time and depend on the position of the transmitter as well as the trajectory of the carrier. The angles flftx0,yo,Zo) and A(t,Xo,yo,z0) are for example labeled as shown by figure 2 with respect to an array of N antennas that may be fixed under the carrier. There currently exist several families of locating techniques making it possible to determine the position (xm,ym,zm) of a transmitter on the basis of the direction vectors. These locating techniques differ through the geometric constraints or characteristics of the antennal array; they are classed into several categories cited hereafter. Use of unambiguous, paired, standardized and calibrated arrays and of goniometry techniques These techniques are in most cases based on 1D azimuthal goniometry. The azimuths 6U= 6(tkfxm,ym,zm) associated with the mth transmitter are measured for various instants tk. By using the position (xk,yk,zk) of the carrier at this instant k, a position (x.mk,ymk,Zmk) of this transmitter is estimated through a ground intersection. The position (xk,yk,Zk) of the carrier is given by a GPS. Its orientation is given by a compass in the case of a terrestrial carrier, and by a navigation platform in the case of an aircraft. On the basis of all the positions (xm^,y^,zm^) an extraction of data is performed making it possible to determine the dominant position (xm,ym,zm) of the incident transmitter. Locating is done by triangulation or by ground intersection (2D goniometry). The drawback of triangulation techniques is that they require a significant transit. Moreover, in antennal constraint terms, these goniometry techniques must use an unambiguous array of sensors, and require calibration, pairing and standardization of the channels. Use of sparse array of paired antennas The applicant's patent application FR 03/13128 describes a method which makes it possible to effect direct estimation of the position (xm,ym,zm) of the transmitter on the basis of a multichannel parametric analysis of the direction vectors at various instants ^ over a duration At. This method requires a pairing of the channels as well as a correction of the distortions in phase and in amplitude of the receivers. The present invention relies on a different approach which does not require, during normal operation, correction of the distortions of the receivers provided that the receiver exhibits a substantially constant response over the locating durations considered. The invention relates to a method of locating one or more transmitters on the basis of an array of sensors moving with respect to the transmitters characterized in that it comprises at least the following steps: o determining the direction vectors a^ corresponding to the response of the array of sensors to a source with incidence (9, A) as a function of the incidence parameters 9, A, and of the parameter p related to the distortion of the phases on the sensors, o transforming this vector a/c so as to eliminate the unknown parameter p, into a transformed vector c , c' ion km o using the transformed vector to obtain the position of the transmitter using a maximized locating criterion. The method according to the invention exhibits the following advantages in particular: o It allows direct estimation of the positions of each of the transmitters on the basis of a multichannel parametric analysis at various instants t, o ft makes it possible to use of arrays sparse sensor that are unpaired, or even uncalibrated (large-aperture array), o It makes it possible to circumvent the pairing as well as the calibration in amplitude and in phase of the reception channels, o It is possible to take into account a model on the variation in the signal reception level, o it makes it possible to introduce any observation model deduced from the direction vectors at different observation instants, o it is possible according to a variant to consider the whole set of channels, without taking a particular reference channel. Other characteristics and advantages of the invention will be better apparent on reading the description which follows of an example given by way of wholly nonlimiting illustration with appended figures which represent: o figure 1 an exemplary locating of a transmitter by an airplane equipped with an array of suitable sensors, o figure 2 an array of 5 antennas and the angles of incidence of a transmitter. In order to better understand the principle implemented in the present invention, the example which follows is given by way of wholly nonlimiting illustration for a system such as shown diagrammatically in figure 1, comprising an airplane 2 equipped with an array of N sensors and with a processor adapted for executing the steps of the method according to the invention and with a transmitter 1 on the ground to be located. In the presence of M transmitters, the airplane receives at the instant t at the output of the N sensors of the array, the vector x(f). Around the instant tk, the vector x(t+tk) of dimension Wx1 is the mixture of the signals of the M transmitters: M M m=l for \t\