2120 IEEE SENSORS JOURNAL, VOL. 17, NO. 7, APRIL 1, 2017
An Indoor Multi-Tag Cooperative Localization
Algorithm Based on NMDS for RFID
Zheng Gao, Yongtao Ma, Member, IEEE, Kaihua Liu, Member, IEEE, Xinlong Miao, and Yang Zhao
Abstract—For radio frequency identification (RFID) system,
the location information is very important in many applications,
such as Internet of Things, logistics, library management, and
so on. Most of traditional RFID localization algorithms can
only locate one tag at a time. However, multitag localization
is desired in many RFID applications. In this paper, we propose
a novel range-free algorithm named nonmetric multidimensional
scaling (NMDS)-RFID(F), which combines NMDS algorithm and
the fingerprinting localization algorithm to achieve indoor RFID
multitag cooperative localization. The NMDS-RFID(F) algorithm
firstly uses received signal strength Euclidean distance based on
fingerprinting method to get the rank order of the distance for
all pairs of tags, and then NMDS algorithm is used to generate
the relative coordinates of tags. Finally, due to the coordinates
of reference tags are known, we can use the coordinate system
registration algorithm to get the absolute coordinates of tags.
Simulation results show that the NMDS-RFID(F) algorithm
outperforms some previous RFID localization algorithms.
Index Terms—RFID, indoor multi-tag localization, range-free,
fingerprinting, nonmetric multidimensional scaling.
I. INTRODUCTION
R
ECENTLY, radio frequency identification (RFID) has
been widely used because of its high performance
and low cost. Using RFID technology for indoor local-
ization has been widely concerned [1]. In the indoor
localization researches, the localization algorithms can be
classified to range and range-free methods. The ranging
techniques include received signal strength (RSS) [2], [3],
time of arrival (TOA) [4]–[6], time difference of arri-
val (TDOA) [7], [8], phase of arrival (POA) [9], [10],
phase difference of arrival (PDOA) [11], angle of arri-
val (AOA) [12], [13]. The localization accuracy of ranging
algorithm is determined by the ranging accuracy, and the dif-
ferent ranging methods that vary in performance. For example,
the TOA ranging method can achieve high accuracy of ranging
in the line-of-sight (LOS) and multipath environment [14],
but it needs accuracy of high clock synchronization which is
expensive between the transmitting end and receiving end [15].
Because the RSS is used widely for wireless signal, there
are many RFID localization algorithms use RSS to range [2],
Manuscript received December 8, 2016; revised February 1, 2017; accepted
February 1, 2017. Date of publication February 6, 2017; date of current
version March 8, 2017. This work was supported in part by the Natural
Science Foundation of China under Grant 61401301 and Grant 61671318, and
in part by the Tianjin Research Program of Application Foundation and
Advanced Technology under Grant 15JCQNJC41900. The associate editor
coordinating the review of this paper and approving it for publication was
Dr. Arindam Basu. (Corresponding author: Yongtao Ma.)
The authors are with the School of Microelectronics, Tianjin University,
Tianjin 300072, China (e-mail: mayongtao@tju.edu.cn).
Digital Object Identifier 10.1109/JSEN.2017.2664338
but the RSS has poor performance caused by the multipath and
non-line-of-sight (NLOS) propagation. The common range-
free method is fingerprinting [16], [17]. For example, the
well-known LANDMARC [18] is based on fingerprinting
localization algorithm, it used WKNN method to locate the
target tag based on the RSS. But it use active reference
tags to replace the fingerprint points for RFID localization,
because the reference tags and the target tags are in the same
environment, for the range-free metod, it can overcome the
multipath and NLOS to some degree. Because the RSS has
many good features such as it is easy and cheap to get, there
are a lot of RFID localization methods using RSS to locate
the tag [19], [20]. While the RSS can be imprecise because
of the environment factor. Therefore, how to achieve a high
localization accuracy using RSS is worth studying.
Until now many researches about RFID localization have
been proposed [21], [22]. The LANDMARC indoor localiza-
tion system based on fingerprinting is proposed to improve
the overall accuracy of locating objects using some reference
tags [18]. The classical multidimensional scaling localiza-
tion algorithm and the triangular method is used to fully
utilize the distance information of the inter-tag for active
RFID tags localization [2]. A Fingerprint-MDS localization
algorithm is proposed to improve a fingerprinting localiza-
tion algorithm with classical Multidimensional Scaling for
RFID [23]. A comprehensive characterization and model of
the UHF RFID channel has been proposed for tag ranging and
localization systems [24]. A multipath mitigation localization
algorithm based on MDS is proposed to improve localization
accuracy in multipath environment for passive UHF RFID
localization [11]. A precise RFID Indoor Localization method
assisted by Wireless Sensor Networks (WSNs) has been pro-
posed for a very large area [25]. A novel semi-passive indoor
UHF RFID localization system using backscatter modulation
has been proposed for precise indoor localization [26]. A pas-
sive UHF RFID location sensing system has been proposed to
offset the large-scale RFID base installation development [27].
In this paper, we proposed a NMDS-RFID(F) algorithm
to improve the localization accuracy and solve the multi-tag
simultaneously localization problem. This paper is organized
as follows: Firstly, We analyzed some problems of RFID
localization and proposed solutions to the problems. Secondly,
the indoor radio propagation model is presented to get the RSS
of tags. Finally, we combine the fingerprinting localization
technology and NMDS algorithm to get the relative coordi-
nates of tags, because the position of the reference tags were
known, then the coordinate system registration algorithm [28]
1558-1748 © 2017 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See http://www.ieee.org/publications_standards/publications/rights/index.html for more information.