Research And Implementation Of Algorithms For High Sensitivity EAS Systems

With the progress of the retail mode,a zero distance contact between customers and goods is required for most of the shops and the superm arkets.Such an open sales mode created a relaxed and happy shopping experience and made shopping malls easy to manage,but also produced a new hidden danger,which invol ves that the shoplifting increases year by year.In order to solve the problem of shoplifting w ithout affecting the custom er's shopping experience,a variety of antishoplifting systems have been deve loped.Electronic Article Surveillance(EAS)system,an antishoplifting system based on radio frequency identification(RFID)and digita l signal processing(DSP)technology,have been rapidly developed because of its high sensitiv ity,anti-jamming capability and lower false alarm rate.The difference between integrated RF-EAS system and traditional EAS system is that the transmitter and the receiver are integrated in a hardware board.And both sides of the antenna are the detection area of the electronic tags.However,the launch of the sweep signal by tim e-division multiplexing technology is due to the fact that the transmitting antenna and the receiving antenn a are integrated toge ther.In addition,the transmitter and the receiver on a board will also bring the signal coupling and interference.Therefore,how to improve the detection rate with reducing the false alarm rate at the same time becomes a key research topic.In this paper,as the main controller,TMS320F28335 drives AD9834 to sweep in 1MHz span with the center frequency of 8.2MHz.This system uses two on-chip timers to respond to the launch timing sequence,processing timing sequence and antenna switching.The received signal is divided into orthogonal I / Q path by IQ demodulation chip called UP C2766 G,then sampled into digital signals by the on-chip 12 bit ADC.This paper makes full use of frequency selection characteristics of the LC circuit,intensity,periodicity,frequency and shape characteristics of the tag signal as identification basis.First of all,this ar ticle presents non-coherent cumulative algorithm for noise restraint,and then calculate the au tocorrelation coefficient for m ultiple cycles to make a preliminary judgment.And the transm ission frequency is fixed by locking the resonant point.Finally,the short-term average zero-crossing rate and CFFT algorithms which based on frequency and shape characteristics are lead up to identify the electronic tag.Environmental interference is also one of the m ain factors affecting system performance in study of the EAS systems.This pa per divides the interf erences into the stea dy-state interference and the transient interference.The first task of improving performance is to improve signal-tonoise ratio.In the aspect of software anti-interference,the comb filter technology and the peak tracking detection algorithm are introduced to increase the system's anti-jamming capability.Taking into account the installation environment is complex,the EAS system should have the ability to adapt to the environment.Therefore,we finally have improved the algorithm so that the relevant threshold can be adjusted according to the noise level of the surrounding environment.In this way,we can effectively reduce the false alarm rate and improve the detection rate.The experimental results show that under the normal no ise environment,the integrated RFEAS system can n ormally work,an d the detect ion distance is 120 ~ 130 cm.In the open fluorescent light or strong nois y jamming environment,the f alse alarm rate is 1 0%,and the detection distance will be shorten about 20 cm.