CN102456117A

CN102456117A - Multi-tag radio-frequency identification anti-collision algorithm

Info

Publication number: CN102456117A
Application number: CN2010105095851A
Authority: CN
Inventors: 徐小科
Original assignee: SHANGHAI FULAI INFORMATION TECHNOLOGY Co Ltd
Current assignee: SHANGHAI FULAI INFORMATION TECHNOLOGY Co Ltd
Priority date: 2010-10-18
Filing date: 2010-10-18
Publication date: 2012-05-16

Abstract

The invention provides a multi-tag radio-frequency identification anti-collision algorithm aiming at overcoming the defects of tediously long data and low identification speed in the existing anti-collision algorithm. By using the algorithm to realize anti-collision, data required to be transmitted is fewer, the average identification times required for identifying one tag are fewer and the identification efficiency is greatly improved. In order to realize the goal, the following three conditions are needed: 1) pre-estimated quantity of tags, wherein the quality can be written in and transmitted to a read head through terminal equipment; 2) 8-bit counter which is arranged in each tag; and 3) zero or one random number generator which is arranged in each tag.

Description

A kind of many labels RF identification impact algorithm

Technical field:

The invention belongs to field of radio frequency identification, specifically, it relates to a kind of algorithm of many labels RF identification impact.

Background technology:

RF identification (English Radio Frequency Identification; Being abbreviated as RFID) technology is development in recent years technology very rapidly; Because it is contactless, need not optical visible, recognition speed is fast and can realize the advantage that multiple goal is discerned simultaneously, obtained in a plurality of fields using widely.When radio-frequency recognition system is worked, in the reach of read head, has the existence of a plurality of labels.In the radio-frequency recognition system of a plurality of read heads and many labels, exist the bumping method of two kinds of forms: a kind of is that same label receives the order that different read heads send simultaneously, and another kind is that a kind of read head is received the data that a plurality of different labels return simultaneously.First kind of situation will be avoided in actual use as far as possible, and the system of a read head and a plurality of labels is situation about occurring the most easily in the real system.In order to solve the collision problem of many tag recognition, the most flexibly with the most widely used binary search method that is to use.Concerning this method, in order from one group of label, to select one of them, read head will send a request command.Read head is through appropriate signal coding, can confirm the bit position accurately that bumps, thereby the data that electronic tag returns are made further judgement, sends the additional requests order, finally confirms all labels in the read head reach.The shortcoming of binary search method is, read head and label all need send the data with the identification code equal length at every turn in the identifying, and when identification code was longer, the time that data are sent was also long, and recognition speed has therefore slowed down.

Summary of the invention:

In order to solve the deficiency that data are tediously long, recognition speed is slow in the existing impact algorithm; The present invention proposes a kind of new many labels RF identification impact algorithm; Utilize this algorithm to realize impact; The data of required transmission are fewer, and it is also less to identify the required average identification frequency of label simultaneously, thereby has improved identification efficiency greatly.

The present invention needs 3 conditions for realizing above-mentioned target:

1, label estimate quantity, this quantity can write and pass to read head by terminal device;

2, be positioned at 8 digit counters of label;

3, be positioned at 0 or 1 tandom number generator of label.

Algorithm flow of the present invention is following:

1. label gets in the read head identification range, tag activation;

2. read head sends the number of labels N that estimates to label, and the size of N is (2 ⁿ-1), such as 7, binary form is shown 00000111;

3. label is according to initial self counting of the quantity N that estimates, and initial method is for doing the AND computing with scale-of-two N and tag recognition sign indicating number minimum 8.For example a tag ID code is 000321, and its binary form is shown 0,001 01000001, itself and N (00000111) are done the AND computing get 00000001, then 00000001 be this label initial count.

4. all countings are the identification code that 0 label sends oneself;

5. read head is judged the number of labels M that sends identification code:

M＞1; A. read head sends the failure order; All countings are that 0 label produces 0 or 1 random number, and with this random number as new counting; All countings are not that 0 label counting adds 1.B. return 4

M=1; Show unique label that identifies, a. read head with it after the exchanges data temporarily with its shielding.Read head sends success command then, and the counting of all labels subtracts 1.B. return 4

M=0; A. read head sends success command, and all label countings subtract 1.B. returning 4 present main anti-collision algorithms is the binary search method, compares with it, and this algorithm has following advantage:

1. the identifying information amount is few---and in the binary search method, read head all must send the match information that equals the tag recognition code length at every turn.The scope of tag recognition sign indicating number generally is very big, and for example in the museum, total registration number of its collection is generally 6 bit digital, the information of a 20byte of its binary representation.And in this algorithm, in the identifying only read head only sent some simple orders, it is many that its message length generally will be lacked, thereby improved recognition speed.

2. the time complexity of algorithm is adjustable---when estimate number of labels and actual quantity near the time, the efficient of this algorithm can be significantly increased.This useful information of experience that has added people's routine works in this algorithm has embodied people's subjective initiative equally.

Description of drawings:

Fig. 1 is the process flow diagram of algorithm according to the invention.

Embodiment:

Come the present invention is done further explanation below in conjunction with accompanying drawing.

Among Fig. 1, can find out that the present invention comprises following steps:

1. label gets in the read head scope, tag activation;

2. read head sends out the discreet value that the user sets;

3. label receives the discreet value that read head sends, the initialization of label counting device;

4. all labels judge whether self counts is 0;

5. the label counting device is 0 label transmission self identification code;

6. the read head statistics is sent the number of labels of information simultaneously;

7. statistical magnitude equals 1, explains and identifies an electronic tag, and then read head is set up with this label and communicated by letter; Reading of data, this label shielding that will read data then, last read head sends success command; Label counting in all identification ranges subtracts 1, turns back to for the 4th step;

8. statistical magnitude equals 0, and read head sends success command, and the label counting in all identification ranges subtracts 1, turns back to for the 4th step;

9. statistical magnitude is greater than 1, and read head sends the failure order, and all counters are to produce random number in 0 the label, give counter with this random number assignment then, and all counters are not that 0 label counting adds 1, turn back to for the 4th step then.

Claims

1. the impact of label RF identification more than kind algorithm, it is achieved in that and it is characterized in that, this algorithm need 3 must condition: 1) label estimates quantity; 2) be positioned at 8 digit counters of label; 3) be positioned at 0 or 1 randomizer of label.

2. a kind of many labels RF identification impact algorithm according to claim 1 is characterized in that the quantity of estimating of said label is write through terminal device by the user, can be stored in the read head.

3. a kind of many labels RF identification impact algorithm according to claim 1 is characterized in that the flow process of said algorithm is following:

A) label gets in the read head scope, tag activation;

B) read head sends out the discreet value that the user sets;

C) label receives the discreet value that read head sends, the initialization of label counting device;

D) all labels judge whether self counts is 0;

E) the label counting device is 0 label transmission self identification code;

F) the read head statistics is sent the number of labels of information simultaneously;

G) statistical magnitude equals 1, explains and identifies an electronic tag, and then read head is set up with this label and communicated by letter; Reading of data, this label shielding that will read data then, last read head sends success command; Label counting in all identification ranges subtracts 1, turns back to d);

H) statistical magnitude equals 0, and read head sends success command, and the label counting in all identification ranges subtracts 1, turns back to d);

I) statistical magnitude is greater than 1, and read head sends the failure order, and all counters are to produce random number in 0 the label, give counter with this random number assignment then, and all counters are not that 0 label counting adds 1, turn back to d then).