Method of discriminating RFID tags in motion from static RFID tags

1. A method for processing a plurality of RFID tags, intended to discriminate between moving tags and stationary tags, the method comprising:

transmitting, from first and second remote reading stations placed along the path of the moving tags, a first and a second radiofrequency interrogation signal to said plurality of RFID tags,

determining whether said first and second reading stations receive response signals from tags in response to said first and second radiofrequency interrogation signals, respectively, in such a way that, for a current tag, if only one of said first and second reading stations receives a response signal from said current tag, said current tag is a stationary tag, and if said first and second reading stations each receive a response signal from said current tag, a k-nearest neighbors method, k being an integer greater than or equal to 1, is used to determine whether the current tag is stationary or moving,

which k-nearest ‘neighbors’ method includes:

learning on a population of N tags, called learning tags, the state of which, stationary or moving, is known, wherein a signal representative of the response signals received by said first and second reading stations for each of said N learning tags are recorded,

selecting the k learning tags having the recorded response signals that are the closest to the response signals received for a current tag according to a predetermined distance criterion, and

determining the stationary or moving state of the current tag on the basis of the state of the k learning tags selected.

2. The method according to claim 1 , wherein k is an odd number.

3. The method according to claim 1 , wherein, in order to select the k learning tags, said response signals of the current tag and the response signals of the N learning tags are compared over a time window having a predetermined length synchronized on a predetermined time reference.

4. The method according to claim 3 , wherein the length of the time window is determined on the basis of the speed of the moving tags and/or their minimum distance with said first and second reading stations.

5. The method according to claim 3 , wherein the time reference is the crest of a peak of the RSSI level of the response signals and the time window is centered on said crest.

6. The method according to claim 3 , wherein the time reference is provided by a device upstream of said first and second reading stations, said time reference marking the beginning of the time window.

7. The method according to claim 3 , wherein the time window is divided into M sub-windows, the learning step further comprises, for each of the learning tags, a step of calculating characteristic values representative of the response signals received during each of said M sub-windows by said first and second reading stations, and the step of comparing the response signals of a current tag and the response signals of the learning tags comprises:

calculating, for the current tag, characteristic values representative of the response signals received during each of said M sub-windows by said first and second reading stations, and

calculating distance between the characteristic values of the current tag and the stationary values of each of the N learning tags,

the k learning tags selected being the k learning tags having the smallest calculated distances.

8. The method according to claim 7 , wherein the characteristic value calculated for each of the sub-windows belongs to the following group:

the average of the RSSI level of the response signals of the tag over said sub-window;

the maximum value of the RSSI level of the response signals of the tag over said sub-window;

the standard deviation of the RSSI level of the response signals of the tag over said sub-window;

the difference between the maximum value and the minimum value of the RSSI level of the response signals of the tag over said sub-window;

the phase difference between the first and second interrogation signals and the response signals received over said sub-window.

9. The method according to claim 7 , wherein the sub-windows have distinct lengths.