Research Article

Management of Sensor-related Data Based on Virtual TEDS

in Sensing RFID System
Wuming Luo1,2
wm_luo@126.com

Miodrag Bolic2*
mbolic@uottawa.ca

Jing Wang2
jwang226@uottawa.ca

Xin Qian2,3
lanyi923@163.com

1. Tianjin Vehicle and Ship Research Institute, Tianjin, China, 300161

2. School of Electrical Engineering and Computer Science, University of Ottawa, Ottawa, Canada, K1N 6N5
3. Suzhou Institute of Industrial Technology, Suzhou, China, 215000
*Corresponding author
Integration of multiple sensors into active Radio Frequency Identification (RFID) system is a technology trend.
Currently, sensing RFID (SRFID) systems face a number of challenges including description of characteristics of
non-smart sensors, management of sensor-related data and operational mechanism of sensor sampling. Based on
Transducer Electronic Data Sheet (TEDS), this paper presents a virtual description method of non-smart sensors
integrated into the active RFID tag. The objective is to improve interoperability and compatibility of the sensing
RFID system with off-the-shelf active RFID readers or novel sensing RFID open-loop systems in the future. This
paper also proposes a data table storage mode for management of sensor-related data in the system based on
virtual TEDS. The paper analyses the operational mechanism and procedure of the sensor sampling system that is
equipped with these two proposed techniques. Several experiments were carried out with a newly developed
container supply chain monitoring application system based on SRFID to examine the performance of the
system.

1. Introduction
Even though Radio Frequency Identification (RFID)
systems can rapidly collect information about items, they
cannot acquire dynamic status information of items, such
as temperature, humidity, and acceleration and so on. On
the other hand, sensor systems are capable of collecting
all kinds of dynamic status information of items and fill
the gap by RFID systems. In order to collect item data
rapidly and monitor the environment status around the
items, it is very important to integrate RFID with sensors.
We refer to the device that integrates RFID tag and
sensors, as SRFID (sensing RFID) tag. However,
designers are confronted with the several challenges
when they attempt to integrate sensors into an RFID tag,
especially when the SRFID tags are used in large scale or
open-loop systems.
The first challenge is the non-smart sensors
description. In future SRFID tags, the number and type
of sensors can be different according to specific
application requirements. In addition, models and
parameters of sensors from different manufacturers differ
among one another even if these sensors are of the same
type. The following problems are recognized when these
diverse sensors in the SRFID tags are identified and
configured by off-the-shelf readers in a standard way: 1)
unique differentiation of sensors in a SRFID tag by the
reader; 2) automatic recognition of types and basic

parameters of sensors; and 3) effective configuration of

sensors parameters. When the sampling datasets
collected by the sensors are transferred to the reader,
parsing and interpretation of these datasets represent
additional problems.
These problems are related to a sensors description.
Sensors description provides basic characteristics of the
sensor, and interoperability between SRFID tags and
readers or host applications. However, general-purpose
digital and analog sensors do not have such
self-description functions, and do not have the capability
to describe themselves.
The second challenge is the sensor-related data
storage in the SRFID tag. Potentially, there will be many
diverse sensors in a SRFID tag and the number of
sensors may vary in different SRFID tags. Therefore,
there is a need for simple, flexible and effective data
storage method. Because the interfaces of various
sensors are different, we need to unify API methods to
access all the sensor-related data and control all the
sensors in the SRFID tag by the reader. Besides, this
storage mode should allow a designer to conveniently
append a sensor into the tag or remove a sensor from the
tag.
The third challenge is the operational mechanism of
sensor sampling. Based on the sensors self-description
and data storage, it is necessary to setup an operational
mechanism of acquiring sensor parameters, configuring
sampling parameters, and collecting and parsing

sampling data.
In this paper, we will focus on sensor
self-description, sensor-related data storage management
and operational mechanism of sensor sampling for
sensors integrated with an active RFID tag. This study is
limited to self-description of a sensor and it does not
consider features of Plug and Play (PnP).
The remainder of this paper is organized as follows:
Section 2 gives an overview of integration of RFID and
sensors that is relevant to our work. Section 3 presents
sensor self-description design with virtual TEDS. Section
4 presents sensor-related data storage approach in the
SRFID tag. Section 5 discusses the operational
mechanism of SRFID system. Section 6 introduces the
implementation and experiment with the SRFID system.
Finally, conclusions are drawn in Section 7.
2. Related Work
This section covers two aspects of related work: a
sensors description and a management method of
sensor-related data in the sensor node.
2.1 Sensors Description
Sensors description is about sensors characteristic
parameters and it involves sensors plug and play
technology. There are two main technologies for device
PnP: Universal Plug and Play (UPnP) and Java
Intelligent Network Infrastructure (Jini) [1]. The UPnP
description for a device is expressed in XML and
includes vendor-specific manufacturer information like
the model name and number, serial number,

manufacturer name, URLs to vendor-specific Web sites,

etc. In the Jini, each network device is described as some
service, and these services can be published
automatically and be invoked by user. Implementation of
UPnP or Jini needs significant computing resources and
as such is not suitable for application in sensors. Sensor
PnP is also a kind of device PnP. Sensor plug and play
has six features, such as ability to announce a sensors
presence, automatic discovery of a sensors presence,
ability to describe sensors capabilities, sensor
self-configuration, control sensors function and sensors
event function. Sensors description is one of six features
of smart sensor Plug and Play.
For smart sensors PnP, there is a dedicated
industrial standard IEEE1451 that defines software and
hardware interface of the transducer (sensor or actuator).
A part of the standard, namely IEEE1451.4 is about a
sensors description technology with Transducer
Electronic Data Sheet (TEDS). Many legacy smart
sensors adopted TEDS to describe sensors characteristic
parameters, and almost all research papers in this field
are based on TEDS to describe sensors characteristic
parameters [2-7]. TEDS describes related parameters of
the sensors, such as the name of the sensor manufacturer,
the type and serial number of the sensor and other
self-descriptive data. TEDS data could help the
host-device identify the related parameters of the sensor
and help the sensor connect to the host-device, by
providing the self-description parameters of the sensor.
The TEDSs are usually stored in the built-in EEPROM
of the smart sensor node, and accessed via the 1-wire
protocol or other protocols.

IEEE 1451

Virtual TEDS
Legacy Sensor

Smart TEDS Sensor

Analog Signal
Sensor

Analog Signal

(bridge, ICP, mV, etc.)

Sensor
Sensor

Digital Signal
TEDS
TEDS
(EEPROM)
(EEPROM)

Digital

Virtual TEDS file

(extra line, or switched w/ analog)

Transducer Electronic Data

Sheet (TEDS)
Sensor manufacturer
Model number
Serial number
.
Measurement range
Transfer function
Calibration info
User Info
And more

WWW
WWW

TEDS
Database

Fig.1 Virtual TEDS System in IEEE1451.4

There are some non-smart sensors, such as legacy
analog sensors and digital sensors, which do not have
any TEDS datasets for self-description. Virtual TEDS
technology can help these sensors implement
self-description feature. These virtual TEDS datasets
could be stored in a local computer or online databases.
Fig.1 shows the block diagram of virtual TEDS system
defined in IEEE1451, with the smart TEDS sensor on the
left side and virtual TEDS sensors on the right. If the
host-devices need to identify the parameters of non-smart

sensors, they can get access to the related memory where

virtual TEDSs are stored.
Regarding integration of RFID and sensors, the
related standards include IEEE1451.7 [8], ISO/IEC/IEEE
21451-7[9] and the ISO/IEC24753[10]. The standard
IEEE1451.7 mainly specifies sensor security and data
structure, the interface between the sensors and the RFID
tag, the integrated architecture framework of the sensors
and the RFID command for the sensor operation and
RFID communication. This standard highlights the

integration of the RFID system and the full function

sensors. The standard ISO/IEC/IEEE21451.7 is similar to
the standard IEEE1451.7. The standard ISO/IEC24753.2
mainly specifies the application interface model for
integration of the RFID tag and the sensors, the data
structures, application commands and the corresponding
responses of a full function sensor and a simple sensor.
These three standards present integration solutions for
RFID and smart sensors and also describe how to
develop a smart sensor that can be integrated with RFID
system. However, they do not deal with integration of
existing analog or digital sensors into RFID tags. These
non-smart sensors need more consideration regarding
their integration with RFID tags.
2.2 Management Method of Sensor-Related Data
There are many models and description methods for
sensor-related data management. The Sensor Web
Enablement (SWE) initiated the Open Geospatial
Consortium (OGC) that has standardized web service
interfaces and data encodings as building blocks for the
Sensor Web and has proposed Sensor Modeling
Language (SensorML) [11]. The W3C Semantic Sensor
Network (SSN) incubator group put forward SSN
Ontology description [12, 13]. Berkeley Manufacturing
Institute, University of California, presented TinyML-Meta-data for Wireless Networks [14]. The Universal
Plug and Play (UPnP) Forum standardized sensor data
description model [15]. Sensor Descriptor Definition
Language (SDDL) is presented in [16] while the
metadata language for sensor descriptionStarfish
Fungus Language (StarFL) is described in [17]. However,
all these sensor-related data description methods are used
in the host computer or at the network level but not at the
sensor node level. These methods are based on the XML
language, which has too much format information,
consumes large memory space, and needs XML Parser to
resolve them. In addition, sensor nodes have limited
resources; therefore XML file is hard to implement the
data retrieval, data removal, data modification and the
like. Therefore, these methods are not suitable for
application in the sensor node, and we need a new
method to store sensor-related data in the SRFID tag.
In terms of sensor-related data management, there
are also many published methods. In [18], a data storage
framework for structured/unstructured RFID or sensors
data at the level of cloud computing is proposed. The
solution in [19] proposed data structures of heart rate
sensor and blood pressure sensor in the middleware and
[20] described three sensors data with XML language in
the Mobile Sensor Framework. Works presented in
[21-22] did not mention sensors description and sensors
data management when sensors were integrated into
RFID tag. In [23], a sensors description is presented but
without sensor data management solution. Work [24]
proposed a series of data models of sensor, sensor node,
session, sensor value, sensor location and sensor event.
The work in [25] divided the memory of a tag into
numerous 64-bit blocks and then further divided these
blocks into two parts, one for the counter and the other
for the sensor data. Solutions [24-25] presented
preliminary sensor data management solutions, but
without a sensors description and effective data

management, these solutions have poor expansibility in

an ultra large scale system or open loop system.
Therefore, better solutions are needed for sensor data
storage and management in the sensor node, especially
with multiple sensors at the sensor node level.
3. Sensor Self-Description Design with Virtual TEDS
Sensors self-description is a critical factor for
compatibility of SRFID tags in ultra large scale systems
or open loop systems and its standardization will also
benefit the storage of sensor-related data. In this section
we present a realization method of the sensor virtual
TEDS, as well as the design of virtual TEDS in the tag
and corresponding templates in the reader.
3.1. Hardware Architecture and Storage of Virtual TEDS
A traditional active RFID system consists of active
RFID tags, reader and host-computer application. When
physical sensors are integrated into the RFID system the
system becomes a SRFID system. The components of
SRFID system are shown in Fig. 2. The SRFID tag
consists of an active RFID tag and several sensors. It has
the functionality of active RFID tag and it is also in
charge of the sensors. It configures sensors parameters,
collects and stores sensory data, and transfers
sensor-related data to the reader. The reader receives and
processes all the data from the SRFID tag and then
transfers the processed data to the host-computer
application. The host-computer application receives and
further processes data, as well as executes some business
operations.
To minimize tag dimensions and cost, sensors
selected in the tag need to be small and common digital
or analog sensors, which are non-smart and therefore
cannot describe themselves. In order to describe
parameters of these non-smart sensors, we design virtual
TEDSs for every sensor, which stores parameters of
these sensors and all virtual TEDSs in the built-in
memory of the active RFID tag. The dataset contents of
TEDS will be introduced in this section and the storage
mode will be introduced in Section 4.
When the readers control sensors in the tag or
collect sensory data from the tag, they should access
virtual TEDS of the sensors first to get the parameters of
the sensors. These SRFID tags should be compatible
with existing readers and host-computer applications,
adopting the same air communication protocol.
Compared with wireless sensor network (WSN), SRFID
tag is similar to a sensor node in WSN, and SRFID
reader is similar to a sink node. SRFID system is
one-hop communication system while WSN system is
multi-hop system. Therefore, virtual TEDS, and some
ideas in the next section are universal, and can be used in
the WSN and other sensor networks.
3.2. Virtual TEDS Design of the Sensor
Fig.3 shows virtual TEDS in the SRFID system.
Virtual TEDS design of the sensor mainly involves two
parts: 1. sensors TEDS dataset in the EEPROM of
SRFID tag and 2. TEDS software in the reader to
decodes the TEDS dataset from the tag. TEDS dataset is

stored in compact binary stream format without any

structure information. After TEDS software receives the
TEDS dataset, it invokes the data structure templates and
data code mapping rule to parse the TEDS dataset in

Temperature
Sensor

Transceiver

Humidity
Sensor
Micro Control Unit

Accelerometer
Sensor
Radiation
Sensor

Active RFID Tag

Non-Smart
Sensors

order to obtain decoded sensor information. The

following contents will focus on these two parts.

Host
Computer
Application

Reader

EEPROM

Virtual TEDSSs are stored

Sensing RFID Tag

Fig.2 Component architecture of the SRFID system

Virtual TEDS(ID,Characteristic)
In the EEPROM of Tag
<0011010101>

Decoded Sensor
Information
Mfgr ID:
41
Model:
710LP
Serial Num: 0012113
Physical typeAcceleration
Range:
50g

TEDS Software
In the Reader

Templates

ID TEDS
Charact TEDS
Sampling Data Block1

TEDS type
Physical sensor type
Logical Sensor Map

Sampling Data Block2

Type Extended

Data Structure Templates

Data Code Mapping

Fig.3 Virtual TEDS in the SRFID system

TEDS dataset in the SRFID tag

Each sensor is designed with two virtual TEDS
datasets stored in the EEPROM of the tag, namely, the
sensor ID TEDS dataset and the sensor characteristic
TEDS dataset.
The main function of the sensor ID TEDS dataset is
to identify the physical sensor uniquely, and this dataset
is read-only. The sensor ID TEDS is formed by 64 bits
dataset and includes manufacturer ID, model number,

version letter, version number and serial number, which

are shown in Table 1. The manufacture ID has
corresponding manufacture ID code template table in the
reader or the host computer application. In most cases,
there is no need to parse the ID TEDS and it is only
regarded as the unique identifier of the physical sensor,
which is similar to the MAC address of the network
interface card.

Table 1. The Sensor ID TEDS(IEEE1451.4)

MANUFACTURER-ID

MODEL-NUMBER

VERSION-LETTER

VERSION-NUMBER

SERIAL-NUMBER

14 bits (17-16381)

15bits (0-32,767)

5-bits char(A-Z)

6bits (0-63)

24bits (0-16777215)

Table 2. The Sensor Characteristic TEDS

FIELD

NAME

LENGTH

FIELD

NAME

LENGTH

1
2

TEDS-TYPE
PHYSICAL-SENSOR-TYPE

3 bits
7 bits

7
8

SCALE-FACTOR-EXPONENT
SCALE-OFFSET-SIGNIFICAND

6 bits
11 bits

3
4

TYPE-EXTENDED
LOGICAL-SENSOR-MAP

5 bits
16 bits

9
10

SCALE-OFFSET-EXPONENT
DATA-PRECISION

6 bits
4 bits

5
6

DATA-TRANSFER-FORMAT
SCALE-FACTOR-SIGNIFICAND

3 bits
11 bits

11
12

SENSOR-CONFIGURATION
USE-FOR-FUTURE

1 bit
7 bits

The main function of the sensor characteristic

TEDS dataset is to describe the physical sensor type and
its basic parameters. The reader can identify the sensor
and configure the sensor parameters automatically. The
sensor characteristic TEDS consists of 80-bit dataset and
the length of every field is shown in Table 2. For a new
non-smart sensor integrated into the tag, we can read the
product specification to get sensor-related parameters in
the TEDS and construct the sensors virtual TEDS.
TEDS-TYPE is to realize the TEDS extension capacity
as
specified
by
IEEE1451.7
[8].
PHYSICAL-SENSOR-TYPE and TYPE-EXTENDED
are used to encode different physical types of sensors.
For example, the sensor code for temperature is 31, and
sensor type code is mapped by specific code mapping
table in the reader. LOGICAL-SENSOR-MAP is mainly
to map the various logical functions defined by the
physical sensor, so that one physical sensor can carry out
several tasks of logical sensors, and the logical sensor
has specific code mapping table in the reader. It also
relates to some data structure templates of the sampling
data of sensor which will be discussed in the following
sections. DATA-TRANSFER-FORMAT defines the
length of the unsigned number in sensors transmission,
and has specific data structure template in the reader.
SCALE-FACTOR-SIGNIFICAND,
SCALE-FACTOR-EXPONENT,
SCALE-OFFSET-SIGNIFICAND
and
SCALE-OFFSET-EXPONENT define the measuring
range of the sensor and its calculation method is
supported by specific examples in IEEE1451.7. The
calculation process is carried out by the reader.
DATA-PRECISION formulates the relative error of the
sampling data and it has specific code template table in
the reader. SENSOR-CONFIGURATION defines
whether the parameters of a digital sensor can be
configured or not, where 0 stands for non-configurable
and 1 stands for configurable. USE-FOR-FUTURE field
is for extending the length of characteristic TEDS, and it
is reserved for the future.
Templates in the SRFID reader
The reader has two kinds of templates for parsing
and interpreting compact virtual TEDS datasets. The first
kind of templates is the compact TEDS data structure
templates, which mainly provides two data structure
templates for decomposing two TEDS datasets, namely,
the sensor ID TEDS dataset and the sensor characteristic
TEDS dataset. The data structure templates of two
datasets are stored in the EEPROM space of the reader,
and their data structures are shown in Table 1 and Table
2.
The second kind of template is the code-mapping
tables which function is to match the code values with
the actual physical meaning and some additional
information. There are many code-mapping tables, such
as
PHYSICAL-SENSOR-TYPE
table,
TYPE
EXTENDED table, LOGICAL-SENSOR-MAP table,
DATA-TRANSFER-FORMAT
table,
DATA-PRECISION table and so on. Every
code-mapping table has a serial of code value referred to
the
IEEE1451.7
standards
[8].
The
PHYSICAL-SENSOR-TYPE table defines the physical

quantity type of sensor, and the fields of the table include

physical sensor type code, corresponding to physical
quantity, and unit of physical quantity which relies on the
unified international standard. The TYPE EXTENDED
table encodes the chemical substance sensors, and its
fields include type extended code and corresponding
chemical
substance
sensor.
The
LOGICAL-SENSOR-MAP table includes two fields, the
logical sensor code and logical sensor function. The
DATA-TRANSFER-FORMAT table includes two fields,
the data transfer format code and the transfer format
definition. The DATA-PRECISION table includes two
fields, the data accuracy code and data accuracy
definition.
4. Sensor-related Data Management
4.1. Requirements Analysis of Sensor-related Data
Management
Sensor-related Data
For every sensor in the SRFID tag, there are four
kinds of sensor-related basic data. They are sensor
identification data, sensor parameter data, sensor
sampling-configuration data and sensor sampling data.
Sensor identification data is to uniquely identify each
sensor in the world, and sensor ID TEDS dataset can be
used as an identifier of a sensor, as it is shown in Table 1.
Sensor parameter data is used to describe basic
performance parameters of a sensor and sensor
characteristic TEDS can be used for describing sensor
parameter data, which is shown in Table 2. The sensor
sampling data is the dataset that the sensor acquires
during the period of monitoring.
These four kinds of sensor-related basic data should
be stored in the memory of the tag. Firstly, the RFID tag
is not always in the coverage of fixed readers with
network. When the RFID tag leaves the reading zone of
the reader and the reader cannot acquire data from the tag
for a long time, some sampling data from the sensors in
the tag must be stored temporarily in the tag. Secondly,
network is not always available. In container supply
chain application, when the container with mounted
SRFID tag is in the container yard or in-transit where
there is no available network, handheld reader cannot
parse the sampling data without characteristic and
configuration parameters. Therefore, characteristic and
configuration parameters are stored in the tag. Thirdly, in
some systems, there are many types of SRFID tags and
each type of sensing RFID tags can have different
sensors with each sensor having different parameters and
configuration. In order to interpret the sensor-sampling
data from the SRFID tag, readers must get quickly basic
information of sensors in the tag, such as sensor features,
configuration and the like. Therefore, additional
computing and communication resources are required in
order to respond to many readers quickly if large number
sensors characteristic and configuration parameters are
stored on the Web. However, if these characteristic and
configuration parameters are stored in the SRFID tag,
they consume small memory space in each tag and have
quick response.

The sensor sampling configuration data is used to

control sampling and set alarms for the corresponding
sensor. The parameters include time stamp when the
sensor is configured, time instant when the sensor starts

sampling, time span between two sampling instances

(sampling period), trigger threshold of the alarm, settings
for alarm conditions and alarm action and so on. Sensor
configuration-sampling data are shown in Table 3.

Table 3. Sensor sampling-configuration data table

FIELD

NAME

LENGTH

NOTE

1
2

CONFIGURATION-TIME
SAMPLING-BEGIN-TIME

32 bits
32 bits

UTC timestamp
UTC timestamp

3
4

INTERVAL-OF-SAMPLING
ALARM-SETING

16 bits
8 bits

Time span between two samplings

Settings for alarm condition and alarm action

5
6

UPPER-ALARM-THRESHOLD
LOWER-ALARM-THRESHOLD

32 bits
32 bits

Trigger threshold maximum

Trigger threshold minimum

Storage Mode of Data in the Tag

The sensor-related data may be stored in the
EEPROM space of the SRFID tag in two modes, namely,
the compact data storage and the data storage with
format information and contents. The former can save
memory space through applying binary data stream. The
latter mode is more flexible regarding organizing and
parsing data. For example, the IEEE1451 standards
provide the TEDS in text format and allow the
manufacturer to use XML language. Although the
storage capacity of active SRFID tag is much larger than
that of the passive RFID tag, its space is still limited and
is in the range of several kilobytes to hundreds of
kilobytes. The tag allocates a large space not only for
static feature information, but also for sampling data
from multiple sensors. In addition, if sensor-related data
is stored with significant formatting overhead, this would
result in occupying more memory and consuming more
power to transfer data to the corresponding reader. For a
mobile device, memory and power are limited resources,
and it is important to reduce transmission throughput as
much as possible. Therefore, the compact storage mode
is adopted without format information for sensor-related
data in the SRFID tag.
Interoperability
Even though there are several various sensors in the
SRFID tag with different sensors parameters and
sensor-sampling data formats, we need to unify API
methods to access all the sensor-related data and control
all the sensors in the SRFID tag. APIs of sensors from
various manufacturers are different so that we cannot
provide a dedicated API for each sensor because this
would require also changes in the reader for every tag
with new sensors. Therefore, we must encapsulate all the

different API methods in the tag, and standardize the

access method of sensor-related data. It is important to
design a simpler, more flexible and more effective data
storage method to deal with sensor-related data in the
SRFID tag.
Besides, the storage method of sensor-related data
should be convinient for the integration of new sensors
and the removal of integrated sensors, and the existing
readers can access this data without any change of the air
protocol between tags and readers.
4.2. Data Structure of Sensor-related Data Storage
In this section, a new management method of
sensor-related data with data table is proposed.
Design the data structure
This paper proposes a data table method to store
sensor-related data in the SRFID tag. In this table, each
sensor has a record with four fields for four kinds of
sensor-related data described in Section 4.1. Each field
stores sensor-related data, and data value in each field is
in compact structure storage mode. The reader can access
the sensor-related data and control the sensors through
the reading or writing data directly to/from this table
without knowing the access and control methods of
individual sensor, as shown in Fig. 4. The data table of
sensor-related data acts like a bridge which connects the
reader and all the sensors, and hides the details of an
individual sensor from the reader. Therefore, the reader
can access all the sensor-related data and control all the
sensors by operating the data table, without knowing the
details of every sensor and their operation commands.

Table of Sensor-related Data

ID

Host
Computer

CHARACTER

CONFIG

DATA

Temperature
Sensor
Humidity
Sensor

Reader
Accelerometer
Sensor
Radiation
Sensor
SRFID Tag

Fig.4 Table of sensor-related data in SRFID tag

In this data table, there are four fields as shown in
Table 4, namely, SENSOR ID TEDS, SENSOR

CHARACTERISTIC TEDS, SENSOR SAMPLING

CONFIGURATION and SENSING DATA ADDRESS.

These four fields are matched with sensor identification

data,
sensor
parameters
data,
sensor
sampling-configuration data and sensor sampling data
respectively. The field SENSOR ID TEDS is totally 64
bits in length as shown in Table 1, the field SENSOR
CHARACTERISTIC TEDS is totally 80 bits as shown in

Table
2,
field
SENSOR
SAMPLING
CONFIGURATION is totally 152 bits as shown in Table
3, and field SENSING DATA ADDRESS is 16 bits. The
field SENSING DATA ADDRESS only stores the
address that points to the sensor-sampling data in the
EEPROM of the tag.

Table 4. Records of the sensor-related data

SENSOR ID TEDS
Sensor1 IDTEDS
Sensor2 IDTEDS
Sensor3 IDTEDS
Sensor4 IDTEDS

SENSOR CHARACTERISTIC TEDS

Sensor1 Characteristic TEDS
Sensor2 Characteristic TEDS
Sensor3 Characteristic TEDS
Sensor4 Characteristic TEDS

SENSOR SAMPLING CONFIGURATION

Sensor1 Sampling Configuration
Sensor2 Sampling Configuration
Sensor3 Sampling Configuration
Sensor4 Sampling Configuration

This table of sensor-related data is stored in the

SRFID tag. Due to the limitation of storage capacity in
the tag and the increase of sensor sampling data, a
strategy of memory rollover is adopted. The maximum
memory of sampling data is defined, and the earliest
record is overwritten on a First In First Out (FIFO) basis
at the instant when the memory becomes full.
The TEDS dataset templates and the code mapping
table involved in the TEDS dataset are stored in the
reader as illustrated in Section 3. The readers are
responsible for parsing the compact TEDS datasets and
matching the codes in the TEDS datasets, which will be
illustrated in Section V. The tag is responsible for reading
the overall sensing data from its memory according to
the first data address and the reader is responsible for
analyzing this data.
Pros and cons
This data table storage method has the following
advantages for multiple sensors in the SRFID tag. In the
data table every sensor has a corresponding record and
all the sensor-related data of this sensor is stored in this
record. All the commands from readers to sensors can be
executed by operating the table in the tag. Adding a new
sensor or removing a sensor from the SRFID tag only
requires a record in the table to be appended or deleted.
Accessing the sensor-related data can be done by reading
the corresponding record of the sensor from the table.
Controlling the sensor is done by rewriting the
corresponding record of the sensor without operating the
sensor directly. By utilizing the data table, we can rely on
the original air protocol between RFID reader and active
tags. In this way, it is possible to unify the interface of
APIs of different sensors. On the other side, it is still
possible to integrate some smart sensors into active
RFID tag.
However, this storage method requires the tag to
control the table. In the tag, we need to implement all the
universal operation methods of the table, which requires
more processing capacity of the Micro Controller Unit
(MCU). In addition, if we integrate a smart sensor into
the RFID tag, there will be a redundant TEDS datasets in
the data table. These drawbacks are negligible for
interoperability or integration in an ultra large scale
system.
5. The Operational Mechanism of SRFID System
Based on virtual TEDS and the storage method of
sensor-related data, this paper sets up the operational

SENSING DATA ADDRESS

Sensor1 Sample Data Address
Sensor2 Sample Data Address
Sensor3 Sample Data Address
Sensor4 Sample Data Address

mechanism and procedure for reader to control sensors

and to manage the sensor-related data. For the SRFID
tags that readers have not accessed before, there are three
stages for the readers to get access to the sensor-related
data. The first stage is to acquire the TEDS dataset from
the tag; the second is to configure the sensors in the tag;
and the last is to collect sampling data from the tag,
which is shown in Fig.5. The sequence of operational
procedure is illustrated by number designation in this
figure. The following content will focus on the
information flow procedure, while neglecting the
security authentication procedure between the tags and
the readers.
5.1. Acquisition of Parameters of Sensors
Firstly, the reader sends a wireless sensor table
query command with some sensors parameters (such as
the sensor type, or all sensors) to the tag, to select a
suitable sensor. The MCU in the tag queries the records
of the table in the tag according to the command of the
reader, and returns the query result with sensor ID
(Sensor ID TEDS Field) and its parameters (Sensor
Characteristic TEDS Field) to the reader.
After the reader receives the TEDS datasets of the
sensors, the MCU in the reader invokes the characteristic
TEDS dataset structures template pre-stored in the tag, to
parse the characteristic TEDS dataset and decompose the
separate value of every parameter of sensor. Because
values of some data unit are code data, the MCU in the
reader continues to invoke the relevant code-mapping
tables pre-stored in the reader to match these codes with
actual physical meanings or values. In this way, the
values of the sensors parameters are obtained by the
reader.
5.2. Configuration of Parameters of Sensors
After the reader gets all the sensors parameters
parsed as mentioned above, these parameters would be
shown on the screen of terminal. The users can select the
sensor of interest and set the value of the sensors
sampling parameter and the sensors parameters
separately. Then the reader packs these parameters into
TEDS datasets according the TEDS data structure
templates and transfers them to the SRFID tag. The tag
finds the corresponding record and updates its fields in
the table. After that, the tag will trigger the sensor to
sample according to the start time instant and conditions
of sampling.
5.3. Collection of Sensor sampling data

When the reader sends an air command with certain

parameters to the tag to collect the sampling data, the tag
queries the table with the parameters and gets the
beginning address of sensor-sampling data in the field
SENSING DATA ADDRESS. The tag reads the whole
sampling datasets and transfers them to the reader
wirelessly.
A series of data structure templates are stored in the
reader. When the reader receives the sampling data, it
invokes the corresponding data structure template first,
and parses the whole sampling datasets into many
individual data units. Then the reader obtains features of
these data units, such as physical meaning, data type and
length, data precision and so on, through mapping them
with parsed TEDS datasets. For some codes in the data
Data Structure
Templates Parsing

Data Code
Mapping

Parsed Sensors
Parameters

Parsing Sensor
sampling data

unit, the reader needs to invoke the code-mapping table

to match these codes and to get actual physical meaning
and measurement unit of the sensor physical quantity.
After the reader parses the sampling datasets
acquired by the sensors, it will re-pack the sensor ID
TEDS dataset, the sensor characteristic TEDS dataset,
sensor sampling dataset and other related data into the
XML files according to the format of Sensor web
Enablement standards developed by OpenGIS [12], and
will upload them to the host computer application
according to the application requirements. The host
computer application can effectively interpret these XML
files and implement subsequent procedure, which
includes displaying data on the screen, storing the XML
file into the relevant XML database directly and so on.

Sensor ID TEDS
Sensor
Characteristic TEDS
Sensor sampling
configuration
Sensor sampling
data

Temperature
Sensor
Humidity
Sensor
Accelerometer
Sensor

And more

Radiation
Sensor

Sensing RFID Tag

Reader

Fig.5 The process chart of the data collected by sensors

6. Experiments
According to the application requirements, a novel
sensing RFID tag and reader system, which is used in the
container supply chain, is developed and tested.
6.1. Application Requirements
With the development of information technologies,
Automatic Identification Technology and Internet of
Things can meet growing requirements for monitoring
environmental conditions in smart container supply chain.
In container transportation it is important not only to
know the static attribute information of goods such as
types and quantities, but also to measure dynamic
information of goods such as temperature, humidity,
acceleration and the status of container door. For
example, when seafood, fruits or vegetables, precision
instruments, oil, drug, chemicals, dangerous cargo,
high-value goods and so on are transported, people often
need to monitor environment status of these goods to
prevent them from being damaged or stolen. There are
four main requirements in containers supply chain,
namely, monitoring status, tracking location, identifying
goods and acquiring information about goods rapidly. In
supply chain, there are many solutions where RFID and
sensor technologies are integrated [26-31].
In this project, we designed a novel SRFID system,
which is used for monitoring condition of goods loaded
in container and status of container doors during the
transportation of goods in the container. This system not

only has the performance of an active RFID system, but

it also senses the environmental condition in the
container.
There are two types of data structures on static
information of items in the container. One is serialized
data structure, such as container shipment information,
and the other is data table structure, such as detailed
information of all the items in the container. The
sensor-related data, which is dynamic environment status
information of items, is also stored in the data table. The
sensing RFID system adapted the standard ISO18000-7,
and this standard can provide some commands to operate
these two data structures. Therefore, two types of data
structure can coexist well.
6.2. Development of Tag and Reader
In this SRFID system, the tag contains active tag
module and sensors module, and these two modules are
wired together. The sensors module contains temperature
and humidity sensor, accelerometer sensor and light
sensor as shown in Fig.6. As a mechanic device, the
container door switch sensor is not on the board. We
implemented partially the active RFID air interface
standard based on ISO/IEC 18000-7 [32], which has
provided a large number of commands for table
operations, such as create table, delete table, insert a
record, delete a record, append a record, read a table,
write a table, query a table and so on. However, the
ISO/IEC 18000-7 standard was not used to acquire all

sensor-related data but only for items static data. We

implemented all the functions related to table operations
to handle data table.
We used Protel software to design and develop the
hardware of the novel SRFID tag, used Keil software
development tools to develop embedded software in the
tag and reader, and used Visual C++ and C# to
implement DLL and Demo App in host computer
application software.
In this SRFID project, we analyze and design virtual
TEDS and data table for sensor-related data, based on the
capabilities of table operations in the tag. We constructed

Fig.6 The integrated circuit board of the sensor module

Fig.8 The sensing tag

6.3. Testing
After the SRFID system was developed,
experiments, product testing and certification were
carried out. We carried out the experiments in four
container supply chain scenarios including container
embarkation point, in-transit point, entrance to yard and
container yard.
We mounted the SRFID tag on the container door,
as shown in the Fig. 7. The sensors module of the tag
was inside the container, as shown in Fig. 8, and it was
responsible for sampling environmental conditions in the
container and transferring sampled data to RFID tag
module. The RFID tag module was outside the container
and was responsible for accessing the table of the
sensor-related data and transferring the data to the fixed
SRFID reader or handheld SRFID reader using active
RFID protocol.

the virtual ID TEDS and characteristic TEDS for each

sensor. All the sensors were described with virtual TEDS
in the memory of SRFID tag according to the description
in Section 3. These virtual TEDS datasets and other
sensor-related data of each sensor are organized and
stored in the corresponding record of the data table as
illustrated in Section 4. This table is created in the
EEPROM of the tag by the microcontroller. The
procedures for sensors samples and parsing sensor data
are implemented according to the analysis in Section 5.

Fig.7 RFID tag attached to the container door

Fig.9 The screenshot of demo app Interface

At the container embarkation point, we configured
all the sensors in the SRFID tag, and write the items
information into the tag with handheld reader. During the
container
transportation,
the
sensors
sampled
temperature, humidity, vibration and light inside the
container and recorded time when the container door was
opened. A fixed reader was placed at the in-transit
crossroad, and it was used to read data about container
shipment and receive the alarm signal. The reader
received this data and parsed them before these data were
transferred to the host computer application. When the
container arrived at the entrance of the container yard,
another fixed reader read all the information of items in
the container and alarm information before the container
came into the yard. When the container was placed in the
yard, a handheld reader detected the container status data,
managed or found some items in the container. The data

from the readers is transferred to the host computer

application and shown on the display. The screenshot of
demo application software interface for sensors
configuration and sampling is shown in the Fig. 9.
At the same time, testing and certification was also
performed by Chinese National RFID Product Quality
Testing Center in January, 2012. The testing activities
included indoor and outdoor testing. Part of the testing
result is the following. The data rate between the reader
and the SRFID tag is 250kbps, and the storage capacity
of the SRFID tag is 64K bytes. The maximum
communication distance between tags and the reader is
125 meters when the tag is mounted on the container
door; and it is 280 meters when the tag is put on the
plastic racks. When the tag is mounted on the container
door, the fixed reader can read the tag ID number from
the tag which speed is 70 km/h. The sensing RFID tag
can sample the status values of temperature, humidity
and vibration inside the container and it can also alarm
when the container door is opened by an unauthorized
person.

[2]

[3]

[4]

[5]

[6]

[7]

7. Conclusion
Integration of sensors into RFID system is not only
a trend, but also a need nowadays. However, there are
several challenging problems related to integration of
sensors and RFID tags including parameters description
of a non-smart sensor, management of sensor-related
data and operational mechanism of sensor sampling. This
paper focuses on addressing these problems in SRFID
system. It proposes virtual TEDS technology to describe
parameters of non-smart sensors in the tag, puts forward
the method with the data table to manage the
sensor-related data effectively, and analyzes the
operational mechanism of sensor sampling. Finally,
validity of these ideas is verified in an experiment that
deals with the container supply chain. In the future,
sensor plug and play technology and semantic sensor
network will be studied. If the function of sensor
Plug-and-Play is implemented completely, the TEDS
data will be constructed automatically by the SRFID tag
without any intervention when new smart sensors are
plugged into an existing SRFID tag.

[8]

[9]

[10]

[11]

Acknowledgments
This paper is a partial work of Tianjin Key
Technology R&D Program No. 09ZCKFGX03700
funded by Tianjin Municipal Science and Technology
Commission, China.

[12]

[13]

Conflict of Interest
The author(s) declare(s) that there is no conflict of
interests regarding the publication of this article.
References
[1]

Igor Sharovar. Smart sensor with Network Plug and

Play Capabilities. M.A.Sc.thesis, University of
Ottawa, Ottawa, Canada, May 2004.

[14]

Virgilio Mattoli, Alessio Mondini, Barbara Mazzolai,

Gabriele Ferri and Paolo Dario. A Universal
Intelligent System-on-Chip Based Sensor Interface.
Sensors. 2010, 10, 7716-7747.
Arne Broring, Patrick Maue, Krzysztof Janowicz,
Daniel
Nust
and
Christian
Malewski.
Semantically-Enabled Sensor Plug & Play for the
Sensor Web. Sensors, 2011, 11, 7568-7605.
Sergio D.Vera, Alberto Bayo, Nicols Medrano,
Beln Calvo and Santiago Celma. A Programmable
Plug & Play Sensor Interface for WSN Applications.
Sensors, 2011, 11, 9009-9032.
Syed Alamdar Hussain, Deniz Gurkan. Management
and Plug and Play of Sensor Networks Using
SNMP. IEEE Transactions on Instrumentation And
Measurement, 2011, 60(5), 1830-1837.
Jorge Higuera, Jose Polo, Manel Gasulla. A ZigBee
wireless sensor network compliant with the
IEEE1451
standard.
2009-IEEE
Sensors
Applications Symposium. New Orleans, LA, USA February 17-19, 2009.
Xiaohui Zhou. A plug and play framework for an
HVAC air handling unit and temperature sensor auto
recognition technique. Mechanical Engineering
Graduate Thesis and Dissertations, Iowa State
University, Iowa, USA, 2010.
Standard for a Smart Transducer Interface for
Sensors and Actuators Transducers to Radio
Frequency
Identification(RFID)
Systems
Communication
Protocols
and
Transducer
Electronic Data Sheet Formats. IEEE std 1451.7,
2010, 1-87.
Information technologySmart transducer interface
for sensors and actuators Part 7: Transducer to
radio frequency identification(RFID) systems
communication protocols and Transducer Electronic
Data Sheet (TEDS) formats, ISO/IEC/IEEE Std
21451-7, 2011, 1-82.
Information
TechnologyRadio
Frequency
Identification(RFID) for Item Management
Application Protocol: Encoding and Processing
Rules for Sensors and Batteries, ISO/IEC Std FDIS
24753:2010, 2010, 1-72.
A. Brring, J. Echterhoff, S. Jirka, I. Simonis, T.
Everding, C. Stasch, S. Liang, and R. Lemmens.
New Generation Sensor Web Enablement. Sensors,
2011, 11(3), 26522699.
M. Botts, OGC Implementation Specification
07-000: OpenGIS Sensor Model Language
(SensorML). Open Geospatial Consortium, Inc.,
2007. Accessed on 07/15/2014, Available from:
http://www.opengeospatial.org/standards/sensorml.
Michael Compton, Payam Barnaghi, Luis Bermudez
et al. The SSN Ontology of the W3C Semantic
Sensor Network Incubator Group. Web Semantics:
Science, Services and Agents on the World Wide
Web. 2012, 17, 2532.
Nathan Ota, William T.C. Kramer. TinyML:
Meta-data for Wireless Networks. Accessed on
07/15/2014, Available from:
http://www.cs.berkeley.edu/~culler/cs294-f03/finalp
apers.

[15]

[16]

[17]

[18]

[19]

[20]

[21]

[22]

[23]

[24]

UPnP Forum. Sensor Management Sensor Data

Model Service (Version 1.0). 2013. Accessed on
07/15/2014, Available from:
http://upnp.org/specs/smgt/.
University of Florida. Sensory Dataset Description
Language (SDDL) Specification (Version 1.0).
2009. Accessed on 07/15/2014, Available from:
http://www.icta.ufl.edu/persim/sddl.
Christian Malewski, Ingo Simonis, Andrew
Terhorst, Arne Brring. StarFL a modularized
metadata language for sensor descriptions.
International Journal of Digital Earth, 2014, 7(6),
450-469.
Lihong Jiang, Li Da Xu, Hongming Cai, Zuhai
Jiang, Fenglin Bu, and Boyi Xu. An IoT Oriented
Data Storage Framework in Cloud Computing
Platform. IEEE Transactions on Industrial
Informatics. 2014, PP(99) , 1-8.
Jun Tang. A Middleware of Wireless sensor
networks for Home healthcare monitoring
application - Design and implementation based on
multiple module of data acquisition. Masters thesis,
Mid Sweden University, Sundsvall, Sweden, June
2013.
Mats Mortensen. Collecting and Distributing Sensor
Data using the Argos middleware platform. Masters
thesis, University of Troms, Troms, Norway,
June 2007.
Zhi Zhang. Networked RFID Systems for the
Internet of Things. Doctoral dissertation, The Royal
Institute of Technolgoy, Stockholm, Sweden 2013.
Filip Zherdev. Storing and reading sensor data from
battery assisted passive RFID. Uppsala University,
Uppsala, Sweden, November 2011.
Luo Wuming, Han Pingyang, Zhao Ruilin. Study on
Design and Application of Wireless Sensor Network
Based on Communication of Radio Frequency
Identification System. The Proceedings of 5th
International
Conference
on
Wireless
Communications,
Networking
and
Mobile
Computing, Beijing, China, Sept.2009, pp.1-6
Eric L.L.C.Hansen. A system for Storing and
Reusing Data Sets from Sensor Networks. Masters

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

thesis, University of Oslo, Oslo, Norway, February

2012.
Tomas Unander, Johan Sidn, Hans-Erik Nilsson.
Designing of RFID-Based Sensor Solution for
Packaging Surveillance Applications. IEEE Sensors
Journal, 2011, 11(11), 3009-3018.
Lixing Wang, S.K.Kwok, W. H. Ip. A radio
frequency identification and sensor-based system for
the transportation of food. Journal of Food
Engineering 101(2010), 120-129.
E.Abad, F.Palacio, M.Nuin, etc. RFID smart tag for
traceability and cold chain monitoring of foods:
Demonstration in an intercontinental fresh fish
logistic chain. Journal of Food Engineering
93(2009), 394-399.
Han Ding, Rui Li, Shaoping Li, Jinsong Han,
Jizhong
Zhao.
MISS:
Multi-dimensional
Information Sensing Surveillance for Cold Chain
Logistics. 2013 IEEE 10th International Conference
on Mobile Ad-Hoc and Sensor Systems. Hangzhou,
China, 2013, 519-523.
Gaoxiang Wang, Wensheng Xu, Mei Wang. The
Grid-computing Based Instrumented Monitoring
Platform For Cold Chain Logistics. 2010
International Conference on Logistics Engineering
and Intelligent Transportation Systems (LEITS),
Wuhan, China, 2010, 1- 3.
Jekwan Park, Kyeongju Kim, Sungwoo Ahn, and
Bonghee Hong. Continuous Query Processing on
Combined Data Stream: Sensor, Location and
Identification.
2010
Seventh
International
Conference on Information Technology. Las Vegas,
Nevada, USA, April 12-14, 2010, 518-522.
Soohan Kim, Wooseok Ryu, Bonghee Hong.
Transaction Processing for Complete Reading of
Semi-Passive Sensor Tag. 2009 Third International
Conference
on
Sensor
Technologies
and
Applications. Athens, Greece, 2009, 683-688.
Information technology AIDC techniques Radio
frequency identification for item management Air
interface, Part 7: Parameters for active air interface
communications at 433 MHz (Revision), ISO/IEC
Std 18000-7, 2008, 1-75.