Chapter 5
Optimization Algorithm and RFID
System Physical Anti-Collision
Radio-frequency identification (RFID) technology is a wireless communication technology that enables users to uniquely identify tagged objects or people [1–4]. The
passive RFID system, due to no direct power supply, is widely used in the field
of warehouse management, smart transportation, healthcare industries, and so on.
An important advantage of RFID technology is multi-target recognition at the same
time, but the problem of improving the reading performance emergences. Nowadays, the common method to the problem is anti-collision algorithms, which solve
the data conflict of multiple tags within the same radio frequency (RF) channel, such
as ALOHA algorithm and binary tree algorithm [5–8]. However, these algorithms
could not improve the reading performance to some extent in practice.
In recent years, the application of the camera on the target measurement has
been widely used [9]. In [10], Ma presented a line-scan CCD camera calibration
method in 2D coordinate measurement. In [11], Dong used single linear array CCD
to measure the vertical target density. In [12], Fahringer described a novel 3D, threecomponent (3C) particle image velocimetry technique which was based on volume
illumination and light field imaging with a single camera. Zhou presented a novel
model and the corresponding calibration approach which took the sensors as an
integrated structure with the viewpoint [13]. Chen proposed a novel non-contact,
full-field, 3D, multi-camera digital image correlation (DIC) measurement system.
In the proposed system, multiple cameras are combined as a single system [14].
Venkataraman used the images captured by camera arrays to measure the depth [15].
However, comparing with the two CCD cameras, using a single CCD camera to
measure the target 3D coordinates needs to constantly adjust the camera position
to obtain the same state of the object image from different angles. The operation is
complex, real-time poor and it is difficult to adapt to the requirements of modern
warehousing logistics. So, in this paper, two CCD cameras (vertical and horizontal
cameras) are used to measure the 3D coordinates of the RFID tags.
In the field of nonlinear modeling, neural networks have many advantages. The
neural networks do not need to construct mathematical function-based models. It can
approximate any nonlinear function or the nonlinear function relations of complex
© Science Press 2021
X. Yu et al., Physical Anti-Collision in RFID Systems,
https://doi.org/10.1007/978-981-16-0835-3_5
157
Optimization Algorithm and RFID
System Physical Anti-Collision
Radio-frequency identification (RFID) technology is a wireless communication technology that enables users to uniquely identify tagged objects or people [1–4]. The
passive RFID system, due to no direct power supply, is widely used in the field
of warehouse management, smart transportation, healthcare industries, and so on.
An important advantage of RFID technology is multi-target recognition at the same
time, but the problem of improving the reading performance emergences. Nowadays, the common method to the problem is anti-collision algorithms, which solve
the data conflict of multiple tags within the same radio frequency (RF) channel, such
as ALOHA algorithm and binary tree algorithm [5–8]. However, these algorithms
could not improve the reading performance to some extent in practice.
In recent years, the application of the camera on the target measurement has
been widely used [9]. In [10], Ma presented a line-scan CCD camera calibration
method in 2D coordinate measurement. In [11], Dong used single linear array CCD
to measure the vertical target density. In [12], Fahringer described a novel 3D, threecomponent (3C) particle image velocimetry technique which was based on volume
illumination and light field imaging with a single camera. Zhou presented a novel
model and the corresponding calibration approach which took the sensors as an
integrated structure with the viewpoint [13]. Chen proposed a novel non-contact,
full-field, 3D, multi-camera digital image correlation (DIC) measurement system.
In the proposed system, multiple cameras are combined as a single system [14].
Venkataraman used the images captured by camera arrays to measure the depth [15].
However, comparing with the two CCD cameras, using a single CCD camera to
measure the target 3D coordinates needs to constantly adjust the camera position
to obtain the same state of the object image from different angles. The operation is
complex, real-time poor and it is difficult to adapt to the requirements of modern
warehousing logistics. So, in this paper, two CCD cameras (vertical and horizontal
cameras) are used to measure the 3D coordinates of the RFID tags.
In the field of nonlinear modeling, neural networks have many advantages. The
neural networks do not need to construct mathematical function-based models. It can
approximate any nonlinear function or the nonlinear function relations of complex
© Science Press 2021
X. Yu et al., Physical Anti-Collision in RFID Systems,
https://doi.org/10.1007/978-981-16-0835-3_5
157