Fingerprint map searching method of steepest descent mode in fingerprint positioning operation

[0025] Example 1:

[0026] Step 1: First preset the search upper limit L=1; calculate the central reference point RP for the first search according to the prior information (a, b) ;

[0027] Step 2: Calculate RP (a, b) The Euclidean distance between the signal characteristic TP measured by the user and the search counter N c =0, set A=0;

[0028] Step 3: Take RP (a, b) As the center, search for the reference points RP around it, and calculate the Euclidean distance in the signal space between each reference point RP and TP;

[0029] Step 4: Record the 4 reference points RP closest to TP in the Euclidean distance of the signal space in step 3, and the 4 reference points RP are used to update the set A;

[0030] Step 5: Use the RP with the closest Euclidean distance in the signal space to TP in the set A as the new search center reference point RP’ (a, b) ;

[0031] Step 6: Observe whether there is any change in the two sets of sets A: If there is a change, return to step 3, N c =0; if there

[0033] Example 2:

[0034] In this embodiment, the preset search upper limit L is 5, the K value is 4, and the other steps are the same as the first embodiment.

[0035] Example 3:

[0036] In this embodiment, the preset search upper limit L is 10, and the K value is 4, and the other steps are the same as the first embodiment.

[0037] figure 2 The simulation results of the traditional fingerprint positioning method and the positioning accuracy obtained by the positioning method of the present invention are shown in Table 1. Table 1 is the simulation result of the positioning speed verified by the simulation. From figure 2 It can be seen that the positioning accuracy after clustering is the lowest and its variance is the largest; while the positioning accuracy of the present invention is comparable to the traditional method without clustering. When the single parameter selection is reasonable (when L=5 in the figure), the positioning accuracy of the present invention is even higher than the traditional fingerprint positioning algorithm without clustering. Since the speed of fingerprint positioning is mainly determined by the number of refe