1246
6, DECEMBER 2016
A Survey of Emerging M2M Systems:
Context, Task, and Objective
Yang Cao, Member, IEEE, Tao Jiang, Senior Member, IEEE, and Zhu Han, Fellow, IEEE
Abstract—Machine-to-machine (M2M) systems enable
machines or devices to collect data, exchange information, and
act on the environment without direct human intervention.
A device in an M2M system not only collects data for its own
usage but also shares the data with other devices automatically
to achieve certain goals. Therefore, emerging M2M systems
for civil transportation, electric power grid, medical treatment,
industrial automation, etc., can be set up based on the net-
working of devices. In this survey paper, we first introduce the
general architecture and communication networks for M2M
systems. Then, we categorize emerging M2M systems according
to the types of M2M context, M2M task, and M2M objective.
We further survey recent solutions for M2M systems from both
academia and industry with the insights of the M2M system
categorization. Finally, we summarize challenges in developing
M2M system solutions.
Index Terms—Categorization, civil transportation, electric
power grid, Internet of Things (IoT), machine-to-machine (M2M)
systems.
I. I NTRODUCTION
NTERNET of Things (IoT) can be defined as a highly
I interconnected network of heterogeneous entities such as
tags, sensors, embedded devices, hand-held devices, and back-
end servers, which provides new services and applications for
civil transportation, electric power grid, medical treatment, and
industrial and home automation [1]. While the IoT is focused
on the end-points and on the interconnection of physical
objects with each other and with humans, the machine-to-
machine (M2M) system is connectivity centric and it refers to
automated systems involving devices that automatically col-
lect data (temperature, humidity, speed, position, heartbeat,
etc.) from a remote source, exchange information, and act on
the environment according to control messages, without, or
with only limited, human intervention through a public net-
work infrastructure. Therefore, the M2M system will play an
Manuscript received April 26, 2016; accepted June 12, 2016. Date of
publication June 20, 2016; date of current version January 10, 2017.
This work was supported in part by the National Science Foundation of
China (NSFC) under Grant 61428104, Grant 61471177, Grant 61471163, and
Grant 61428101, Joint Specialized Research Fund for the Doctoral Program
of Higher Education and Research Grants Council Earmarked Research
Grants under Grant 20130142140002, the Key Project of Hubei Province
in China under Grant 2015BAA074, and in part by the U.S. NSF under
Grant ECCS-1547201, Grant CCF-1456921, Grant CNS-1443917, and Grant
ECCS-1405121. (Corresponding author: Tao Jiang.)
Y. Cao and T. Jiang are with the Wuhan National Laboratory for
Optoelectronics, School of Electronics Information and Communications,
Huazhong University of Science and Technology, Wuhan 430074, China
(e-mail: ycao@hust.edu.cn; tao.jiang@ieee.org).
Z. Han is with the Department of Electrical and Computer Engineering,
University of Houston, Houston, TX 77004 USA (e-mail: zhan2@uh.edu).
Digital Object Identifier 10.1109/JIOT.2016.2582540
2327-4662 c 2016 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.
Authorized licensed use limited to: KU Leuven Libraries. Downloaded on December 29,2023 at 11:25:06 UTC from IEEE Xplore. Restrictions apply.
IEEE INTERNET OF THINGS JOURNAL, VOL. 3, NO.
important role in the realization of the IoT [2]. Communication
suppliers such as Cisco and Ericsson estimate that 50 billion
devices will be networked via M2M systems by 2020 [3].
Moreover, Forrester Research predicts that the global sales
volume of M2M systems will be $50 billion by 2020 [4].
Generally, M2M systems can be heterogeneous, unstruc-
tured, and loosely-knit by using standardized communication
technologies, such as IEEE 802.11 (WiFi), IEEE 801.15.4
(ZigBee, 6LowPAN), long term evolution (LTE), IEEE 802.16
(WiMAX), Ethernet, power line communications, and IEC
61850 [5]. Using standardized technologies allows easier
device interoperation in M2M systems and facilitates using
mass-produced, standards-compliant equipments, which makes
implementation less expensive [2]. The recent trend is that
wireless technologies can play more and more important role
in M2M systems, and thus most devices or machines can
be networked via radio chips or modules. Although M2M
systems have great potential in the future, it is still in its
infancy since the implementation and management of M2M
systems are facing many challenging issues, and there is still
a long way for M2M systems to enter every corner of the
society.
There are some recent surveys about the emerging M2M
systems. Lu et al. [6] mainly surveyed the communication
requirements of real-time monitoring M2M systems, i.e., the
energy efficiency, reliability, and security. Some solutions for
secure communications in M2M systems were introduced.
Taleb and Kunz [7] studied some of the standardization activ-
ities by the 3rd Generation Partnership Project (3GPP), and
focused on M2M systems based on wireless cellular networks.
Some solutions for tackling issues like congestion and over-
load control based on M2M communication features were
discussed. Wu et al. [8] investigated M2M systems from
the view of future embedded mobile Internet. Specifically,
they discussed the requirements and challenges associated
with mass-scale M2M systems, and described potential sys-
tem architectures and deployment options that can enable the
connectivity of billions of low-cost devices. Moreover, the
progress of standardization efforts for M2M systems was also
summarized. Kim et al. [1] surveyed architectures, issues, and
enabling technologies for M2M service platforms, which sup-
port M2M systems such as energy M2M system in smart grid,
e-health system, environmental monitoring system, supply
chain management system, and so forth. Fan et al. [9] provided
an overview of several standardization activities for M2M
communications, and analyzed some of the enabling M2M
technologies and applications from an industry perspective.
CAO et al.: SURVEY OF EMERGING M2M SYSTEMS: CONTEXT, TASK, AND OBJECTIVE
Chen et al. [10] provided a survey of recent developments
in home M2M system. As an emerging M2M system, home
M2M system requires specific architecture, radio technologies,
and quality-of-service (QoS) provisioning.
The above literatures have provided valuable insights into
the high-level network architectures, communication tech-
nologies, standardization efforts, service platforms, and key
applications supported by M2M systems. To solve challeng-
ing issues in the implementation and management of M2M
systems, we need to properly categorize M2M systems into
different types, which reveals the essential characteristics of
enabling M2M systems. In this paper, we fully investigate dif-
ferent types of M2M systems and develop an M2M system
categorization. The contributions of this paper are summarized
as follows.
1) To sort highly segmented solutions for different M2M
systems, we categorize M2M systems according to the
type of M2M context, M2M task, and M2M objective,
which can be effective to evaluate solutions for M2M
systems.
2) We survey most recent M2M system solutions from
both academia and industry systematically based on
the proposed M2M system categorization. Specifically,
we use a context-task-objective format to represent the
sub-categories of M2M systems.
3) We track the future directions of M2M systems,
especially main challenges including enabling massive
concurrent data transmissions for M2M systems, config-
uration and management of M2M systems, coexistence
of M2M systems and legacy systems, and user satisfac-
tion for M2M systems.
The remainder of this paper is organized as follows.
Preliminary of M2M systems is provided in the following
section. In Section III, we propose the M2M system cate-
gorization. In Section IV, recent solutions for M2M system
are surveyed with the proposed M2M system categorization.
Section V summarizes the major challenges in the M2M sys-
tems, and tracks future directions. Finally, conclusions are
drawn in Section VI.
II. P RELIMINARY OF M2M S YSTEMS
A. General Architecture of M2M Systems
The architecture of M2M systems serves as the founda-
tion of supporting M2M applications and services, which
is one of the key parts in major M2M standardiza-
tion activities. Recently, the European Telecommunications
Standards Institute (ETSI), the 3GPP, the Internet Engineering
Task Force (IETF), the Alliance for Telecommunications
Industry Solutions, the China Communications Standards
Association (CCSA), and the Open Mobile Alliance (OMA)
have launched standardization activities on M2M [11].
Specifically, the ETSI focuses on the general M2M archi-
tecture, especially for its components, and the interactions
between its two domains, i.e., the device-and-gateway domain
and the network domain. The 3GPP M2M standard is basi-
cally consistent with the ETSI M2M standard, but places
extra emphasis on cellular-based communication solutions for
Authorized licensed use limited to: KU Leuven Libraries. Downloaded on December 29,2023 at 11:25:06 UTC from IEEE Xplore. Restrictions apply.
1247
Fig. 1. Illustration of ETSI M2M architecture.
M2M systems, i.e., how can cellular networks support M2M
applications and services efficiently. In 2012, multiple M2M
standardization bodies formed into a single organization, i.e.,
“oneM2M,” to attract and actively involve organizations from
M2M-related business domains [12].
According to ETSI M2M standard, the general architec-
ture of M2M system is composed of multiple key elements
as follows [13].
1) M2M device connects to the network domain via direct
connectivity or M2M gateway. In the first case, the M2M
device connects to the network domain via the access
network, which performs the procedures such as regis-
tration, authentication, authorization, management, and
provisioning with the network domain. In the second
case, the M2M device connects to the M2M gateway
using the M2M area network.
2) M2M gateway acts as a proxy between M2M devices
and the network domain. As an example, an M2M
gateway can run an application that collects and treats
various information (e.g., contextual parameters) from
sensors and meters.
3) M2M server works as a middleware layer to pass data
through various application services.
4) M2M area network provides connectivity between M2M
devices and M2M gateways.
5) M2M communication network provides connection
between the M2M gateways/devices and the M2M
servers. Usually it contains two parts: a) the access
network and b) the Internet.
As depicted in Fig. 1, these elements constitute the two
domains of M2M system specified by ETSI, that is, M2M
device, M2M area network, and M2M gateway constitute the
device-and-gateway domain, while communication network
and M2M server constitute the network domain.
B. Communication Solutions of M2M Systems
Communication solutions that support M2M systems play a
critical role, which are named as the machine type communica-
tion (MTC) in the 3GPP M2M standard. 3GPP defines features
and requirements for MTC [14]. There are some key elements
defined in 3GPP M2M architecture, namely, MTC device,
MTC feature, MTC group, MTC server, MTC user, and MTC
subscriber. It is worth noting that MTC features are functions
to optimize the communication network for M2M applications,
1248
Fig. 2. Illustration of 3GPP M2M architecture, in which solid lines rep-
resent the user plane communication and dotted lines represent the control
plane communication. Here, eNB denotes evolved node B (the base station of
the LTE network). PDN denotes packet data network. MME denotes mobil-
ity management entity. HSS denotes home subscriber server. IWF denotes
interworking function.
for example, low mobility, location-specific trigger, infrequent
transmission, and group-based MTC features.
As depicted in Fig. 2, MTC devices communicate with
one or more MTC servers, and the network operator pro-
vides network connectivity to MTC servers. The MTC servers
can be controlled by the 3GPP network operator or con-
trolled by the third-party service provider over the Internet.
To support MTC, the 3GPP core network is required to com-
municate with the MTC Server. An intermediary entity may be
used for the control plane communication (for signaling and
control) for topology hiding or protocol translation purposes.
Specifically, the interworking function (IWF) is introduced to
the control plane. An MTC-IWF can be a stand-alone entity
or a functional entity of another network element. Moreover,
the MTC-IWF hides the internal public land mobile net-
work (PLMN) topology and relays or translates signaling
protocols used over MTCsp1 to invoke specific functionality
in the PLMN. Moreover, 3GPP introduces the key issues and
solutions for fulfilling the MTC service requirements [15]. The
key issues include the lack of Internet protocol (IP) addresses
and congestion of control/data signals for a large number of
MTC devices. The solutions attempt to resolve the key issues
by: 1) using IPv6 address or dual-stack address and 2) using
group-based management for M2M devices.
1 It is the reference point an entity outside the 3GPP system uses to
communicate with the MTC-IWF related control plane signaling.
Authorized licensed use limited to: KU Leuven Libraries. Downloaded on December 29,2023 at 11:25:06 UTC from IEEE Xplore. Restrictions apply.
IEEE INTERNET OF THINGS JOURNAL, VOL. 3, NO. 6, DECEMBER 2016
Besides, the CCSA focuses on pervasive networks, includ-
ing general requirements, applications, networking, sensing,
and related short range radio frequency connectivity. The
OMA M2M device manageability defines the requirements
for the M2M gateway managed object [8]. The IETF is
working on communication protocols for embedded Web ser-
vices of M2M applications based on constrained application
protocol [16]. Moreover, low power wide area (LPWA) tech-
nologies [17] that can reduce the energy consumption of
wide coverage for M2M systems, are attracting massive atten-
tions. Specifically, narrowband (NB)-IoT and direct sequence
spread spectrum modulation schemes have been proposed for
the physical layer of LPWA M2M systems thanks to their
excellent coverage performance. The NB-IoT is particularly
suitable for the refarming of underutilized resource blocks or
channels from legacy systems like global system for mobile
communications (GSMs) and LTE [18].
III. C ATEGORIZATION OF M2M S YSTEMS
In the previous section, we briefly review the general
architecture and communication solutions of M2M systems,
which serve as the foundations of M2M systems. Referring to
the general architecture and communication solutions, diverse
M2M systems for different purposes and scenarios are inves-
tigated, and solutions for different M2M systems are highly
segmented due to excessive number of “gray areas” in build-
ing an M2M system. Therefore, we endeavor to categorize
M2M systems into different types, as an effort of regulat-
ing the emerging M2M systems. With the guidance of the
proposed M2M system categorization, we further evaluate the-
oretical and practical M2M system solutions effectively. As
shown in Fig. 3, we categorize M2M systems according to
the type of the M2M context, M2M task, and M2M objective,
respectively, which is described in detail as follows.
A. Categorize M2M Systems by M2M Context
It is intuitive to design M2M systems for a specific context
(e.g., civil transportation, electric power grid, medical treat-
ment, and industrial automation). Solutions can be developed
with the consideration of unique characteristics of M2M con-
texts. According to the M2M context, M2M systems can be
divided into two types, namely, dynamic M2M systems and
static M2M systems. The major difference of these two types
is the topology of the M2M system. In dynamic M2M sys-
tems, some nodes (e.g., M2M devices and M2M gateways) are
moving, i.e., the topology is changing over time, which can
result in the change of wireless link quality, communication
mode switch, and dynamic resource allocation. In contrast,
the topology for static M2M systems keeps unchanged for
a relatively long time. For example, vehicular M2M system,
medical M2M system, and robotic M2M system are dynamic
M2M systems and energy M2M system, home M2M system,
and industrial M2M system are static M2M systems.
1) Dynamic M2M Systems:
a) Vehicular M2M system: Vehicular M2M system is a
representative dynamic M2M system with fast changing topol-
ogy (vehicles are moving fast) [19]. Most vehicular M2M
CAO et al.: SURVEY OF EMERGING M2M SYSTEMS: CONTEXT, TASK, AND OBJECTIVE
Fig. 3. Categorization of M2M systems.
applications can be categorized into one of the following:
safety and security, information and navigation, diagnostics,
and entertainment [20]. For example, information and nav-
igation services provide access for the vehicle occupants
to a variety of location sensitive information and contents.
Diagnostic services enable the occupants and/or vehicle main-
tenance/repair centers to collect data from a multitude of
sensors located throughout the vehicles in order to make
maintenance and/or repair recommendations.
As an example shown in Fig. 4, automatic crash notifi-
cation (ACN) service enables various crash sensors on the
vehicles to report the location and extents of damages to the
vehicles in the event of a crash [21]. ACN also facilitates
reports of the crash to the vehicle data management system
via nearby vehicles [vehicle-to-vehicle (V2V)] or via roadside
units (vehicle-to-infrastructure). According to the ETSI M2M
standard, the roadside unit can be viewed as the M2M gate-
way, the sensor data aggregator in the vehicle can be viewed
as the M2M device, and the M2M server is in the vehicle data
management system. The roadside unit acts as a proxy for the
access network toward the sensor data aggregators.
Obviously, a vehicular M2M system requires a combina-
tion of short-range, low-power, low-throughput wireless access
(e.g., WiFi and Bluetooth) for sensing processes and local con-
nectivity (in-vehicle or V2V communications), and long-range,
low-latency, high-throughput wireless access such as LTE for
status reporting or Internet access [22]. The topology change
of the vehicular M2M system is not difficult to be predicted
since vehicles are drove along the road/lane in only two direc-
tions with the speed data that can be collected by the global
positioning system.
b) Medical M2M system: Another example of dynamic
M2M systems is the medical M2M system for e-health, which
aims at improving the quality of health management for all
people and reducing the healthcare costs and risks for patients.
Key applications enabled by medical M2M system include
telemedicine and electronic health management. Specifically,
telemedicine improves patient care via more accurate and
Authorized licensed use limited to: KU Leuven Libraries. Downloaded on December 29,2023 at 11:25:06 UTC from IEEE Xplore. Restrictions apply.
1249
faster reporting of changes in the physical condition of a
patient. Moreover, automated connectivity of medical devices
to the hospital network and remote management of these
devices are enabled with telemedicine [23]. Electronic health
management can be achieved via wearable devices that com-
municate with smart phones, with the purpose of recording
the physical condition changes and providing health-related
suggestions to their users. In the case of remote health mon-
itoring and care, bio-sensors worn by a patient form a body
area network to record health and fitness indicators, e.g., blood
pressure, body temperature, heart rate, and number of paces.
The sensors forward the collected data to an M2M device (e.g.,
smart phone) that acts as an information aggregator. Then, the
aggregator forwards the data to the M2M server via commu-
nication networks. The M2M server responds to the collected
data by sending alerts and appropriate medical records to med-
ical providers. In emergency situations, an M2M device can
directly provide the medical status of a patient and route to the
hospital, allowing physicians, and medical devices to prepare
for treatment in advance of the arrival of the patient [24]. The
mobility of medical users results in the change of topology of
M2M systems, and how to accurately predict the user behavior
is still an open question.
c) Robotic M2M system: Robotic M2M system is an
emerging dynamic M2M system, which enables multiple auto-
matic robots to complete complex tasks, e.g., proving taxi
services in a city, managing goods in a warehouse, and clean-
ing guest rooms in a hotel [25]. As an illustrative example,
robotic taxis powered by batteries swarm together in an area
to provide taxi service in a timely and efficient manner. The
robotic taxis respond to real-time traffic conditions of the city,
and cooperate to reduce congestion at bottlenecks in the city
and to service pick-up areas that are most frequently used.
When they are not in use, the taxis head for charging sta-
tions where actuators set off recharging batteries, perform
simple maintenance tasks, and clean the cars. The charging
stations communicate with each other to ensure no over or
under utilization. Different from vehicles driven by humans
1250
Fig. 4. Illustration of the vehicular M2M system.
and devices carried by e-health users, the mobility of robotic
taxis is much easier to be predicted because the M2M system
operator knows the purpose of the robotic network and can
directly control it.
2) Static M2M Systems:
a) Energy M2M system: Energy M2M system is a repre-
sentative static M2M system, which is an important component
in the smart grid [26]. By integrating communication capabil-
ities with electric power generation, delivery and distribution
infrastructures, energy M2M system can achieve automatic
monitoring and control. Significant savings in resource con-
sumption is possible when power supply is dynamically
matched with demand. Key smart grid applications are smart
metering, distribution network automation, demand response,
equipment diagnostics, as well as the wide area monitoring
and control [27]–[29].
As shown in Fig. 5, an illustrative example of the energy
M2M system is described. First, M2M-enabled smart meters
collect utility usage information from appliances via home area
networks (HANs) or business area networks (BANs). Then,
smart meters send the information to the M2M server through
the access network and the Internet. Finally, the meter data
management system can utilize the information. Because the
devices in the smart grid is usually in fixed locations, the topol-
ogy of the energy M2M system is static, and the HAN/BAN
can be wired Ethernet or short-range wireless networks such
as WiFi and ZigBee, while the access network can be wired
network or LTE. Recently, some M2M system solutions were
proposed for the smart grid. For example, based on the case
that smart meters communicate with M2M servers via M2M
gateways, a data aggregation scheme for an M2M gateway is
proposed to intelligently collect data from a large number of
smart meters in an area and sends the aggregated information
to the M2M server [30].
Authorized licensed use limited to: KU Leuven Libraries. Downloaded on December 29,2023 at 11:25:06 UTC from IEEE Xplore. Restrictions apply.
IEEE INTERNET OF THINGS JOURNAL, VOL. 3, NO. 6, DECEMBER 2016
b) Home M2M system: Home M2M system is another
static M2M system that deploys sensors and actuators in
houses/offices to make the indoor environment more comfort-
able [31]. Specifically, room climate control can be adapted
to human preferences and to weather, the room lighting can
change according to the time of the day, domestic incidents
can be avoided with appropriate monitoring and alarm sys-
tems, and energy can be saved by automatically switching
off the electrical equipments when not needed. Considering
the static topology of M2M devices in a home M2M system,
low-cost communication methods like WiFi and ZigBee can
be sufficient.
c) Industrial M2M system: Industrial M2M systems help
in improving the automation in industrial plants (termed
“Industrie 4.0” or “Industrial Internet”) with a massive deploy-
ment of sensors, actuators, and radio frequency identification
(RFID) tags associated to the production parts [32]. Industrial
M2M system is a static M2M system because the locations
of industrial M2M devices (sensors and actuators) are rarely
changed. For example, as production parts reach the process-
ing point, the tag is read by the RFID reader. An event is
generated by the reader with all the necessary data, such
as the RFID number, and stored on the database of the
M2M system. The actuators get notified by this event (as
it has subscribed to the service) and picks up the produc-
tion part. By matching data from the database and the RFID
tag, it knows how to further process the part. For an indus-
trial M2M system to prevail, communication technologies
will need to efficiently connect machinery over varying dis-
tances in a flexible manner with high security, robustness, and
availability at a low cost. Industrial M2M system requires reli-
able communication capabilities among M2M devices and the
M2M system operator to ensure the quality of the production
process.
CAO et al.: SURVEY OF EMERGING M2M SYSTEMS: CONTEXT, TASK, AND OBJECTIVE
Fig. 5. Illustration of the energy M2M system.
Remarks: It is intuitive to provide M2M system solutions
according to the requirements and features of the M2M con-
text. To tailor M2M system solutions to specific context, three
critical questions should be answered: 1) what are the unique
requirements of this context?; 2) what are the unique features
that can be utilized in this context?; and 3) what are the unique
challenges to provide M2M system solutions for this context?
By categorizing M2M systems into dynamic M2M systems
and static M2M systems, one can focus on the network topol-
ogy of M2M systems instead of an enormous number of
different contexts. However, M2M context only offers one
perspective of categorizing M2M systems, which is not com-
prehensive. To further understand the essential characteristics
of M2M systems, we need to examine and categorize them by
other perspective, i.e., M2M task and M2M objective, which
will be given in the following part of this section.
B. Categorize M2M Systems by M2M Task
Here, we define an M2M task as an operation supported by
an M2M system, which has a specific purpose. For example,
fault detection is an M2M task supported by the energy M2M
system in smart grid, while vehicle status monitoring is an
M2M task supported by a vehicular M2M system. Specifically,
an M2M task can be device-centric or system-centric accord-
ing to who provokes or executes the task. For example, a
device-centric task is provoked or executed by an M2M device,
while a system-centric task is provoked or executed by an
M2M system operator. Furthermore, a device-centric M2M
system is defined as an M2M system whose feature task is
device-centric, while a system-centric M2M system is defined
as an M2M system whose feature task is system-centric.
Note that device-centric tasks and system-centric tasks can
be operated at a single M2M system, however, such case can
Authorized licensed use limited to: KU Leuven Libraries. Downloaded on December 29,2023 at 11:25:06 UTC from IEEE Xplore. Restrictions apply.
1251
be a trivial extension from device-centric M2M systems and
system-centric M2M systems.
In what follows, we will introduce some representative
M2M systems categorized by the M2M task. Specifically,
monitoring and sensing system and device cooperation sys-
tem are device-centric M2M systems, while decision making
and control system and data analysis system are system-centric
M2M systems.
1) Device-Centric M2M Systems:
a) Monitoring and sensing system: One common device-
centric M2M system is the monitoring and sensing system,
where M2M devices including sensors, cameras, and meters
capture raw data from objectives, and then send the raw data or
processed data to the M2M system operator [33]. For example,
remote patient monitoring system, real-time remote security
and surveillance system, and vehicle status monitoring sys-
tem [6] are typical monitoring and sensing M2M systems from
different contexts.
Characteristics of M2M devices in a device-centric M2M
system can be exploited to develop applicable solutions. For
a monitoring and sensing system, an M2M device with lim-
ited computing capability cannot process the raw data, thus
the raw data is needed to be sent to a stronger M2M device
(termed “surrogate” or “hub”) or to the M2M system opera-
tor. When an M2M device has sufficient computing capability,
it can process its own raw data or the raw data from other
devices. Therefore, only a local decision is needed to be sent
to the M2M system operator, which results in a significant
reduction of the communication bandwidth.
b) Device cooperation system: In a device cooperation
system, a cluster of M2M devices (vehicles or robots) take
actions to achieve the same goal. One example is that robotic
vehicles patrol an area via a cooperative manner, that is, there
is a leader vehicle, whose trajectory needs to be copied by
1252
follower vehicles in real time [34]. Another example is the
smart alarm clock, that is, the alarm clock that can communi-
cate with its owner’s vehicle, refrigerator, and so on, to set the
best wake-up time. Clearly, device cooperation system can be
bandwidth demanding due to frequent real-time interactivities
among multiple devices, which is more challenging when the
number of cooperative devices is huge and the covering area
is large.
2) System-Centric M2M Systems:
a) Decision making and control system: Decision mak-
ing system is a representative system-centric M2M system.
For example, the operator of an electric power management
system (a specific M2M system) needs to decide the opti-
mal defense plan to reduce security risks with limited costs.
To make timely decisions, data from an M2M device is
periodically collected by an M2M gateway via short-range
communications, and the retrieved information at the M2M
gateway will be sent to the M2M server through wide-area
communication networks [30]. It is worth noting that data
integrity is quite important for the decision making system,
however, it is possible that the collected data is uncompleted
or even incorrect due to channel fading or other factors. One
challenging issue is how to make decisions quickly and prop-
erly according to a large volume of incomplete or incorrect
data [35].
Once the decision is made by the M2M system operator,
actions take effect, that is, the behaviors of M2M devices are
directly controlled via M2M servers to alarm people, cut off
power, and stop the current operation [36]. Such device con-
trol requires reliable and low-latency downstream connections
between M2M servers and M2M devices. Since M2M devices
are usually automatic, it is critical to protect M2M devices
from malicious control of attackers [37], [38]. For example, a
computer hacker can hijack the controller of a self-driving-car
to make a car accident.
b) Data analysis system: Another system-centric M2M
system is the data analysis system. Generally, data analy-
sis can help to make predictions based on collected data
from many M2M devices during a specific period of time.
A wide range of problems such as traffic congestion, power
distribution anomalies (e.g., voltage level spikes), malicious
attacks, and so on may fall into the scope of prediction [39].
The data mining technologies can be applied to analyze col-
lected data from M2M devices. When the volume of data,
the variety of data types, and the velocity of data generated
are massive, it is quite challenging to make analysis for the
“big data” [40], [41].
Remarks: By categorizing M2M systems into device-centric
M2M systems and system-centric M2M systems, one can
focus on the characteristics of M2M devices or M2M system
operators according to the types of M2M tasks. Moreover, it
is crucial to develop schemes that achieve the goals of M2M
tasks while taking into account limited system resources (e.g.,
radio resources and computing capabilities).
C. Categorize M2M Systems by M2M Objective
Here, we define M2M objective as the design criterion of
an M2M system solution. For example, an M2M system can
Authorized licensed use limited to: KU Leuven Libraries. Downloaded on December 29,2023 at 11:25:06 UTC from IEEE Xplore. Restrictions apply.
IEEE INTERNET OF THINGS JOURNAL, VOL. 3, NO. 6, DECEMBER 2016
be expected to achieve high performance or low consumption
for different purposes. Specifically, we divide M2M systems
into performance-oriented M2M systems and consumption-
oriented M2M systems according to the M2M objective, e.g.,
QoS-aware M2M system, secure M2M system, and reliable
M2M system are performance-oriented M2M systems, while
energy-efficient M2M system and cost-effective M2M system
are consumption-oriented M2M systems.
1) Performance-Oriented M2M Systems:
a) QoS-aware M2M system: To realize a QoS-aware
M2M system, one big challenge is that the communication
infrastructure needs to accommodate large volumes of report-
ing messages simultaneously [37]. We use smart meter data
collection system (a specific energy M2M system) as an exam-
ple. As the control center of the utility establishes a contract
with the generators periodically (e.g., every hour) to buy
power supply, historical data of power demand collected from
smart meters can be used in this contract [30]. Given the
estimated aggregated power demand (estimated demand) of
all electric appliances in a service area, the power price is
determined by the power generator for a periodic contract
(e.g., locational margin price [42]). However, the historical
data of power demand can be incomplete and outdated due
to the loss and delay of data transmitted from smart meters.
Therefore, actual aggregate power demand (actual demand)
can be different from estimated demand. It is obvious that
the data loss can incur additional costs for the distribution
network and consumers. Define a non-negative information
loss rate (ILR) as a QoS metric for reporting data transmis-
sions as ILR = max{R − r, 0}, where R denotes the minimum
average data rate in order to complete reporting data trans-
mission of a smart meter during a reporting duration [43].
r denotes the average achievable data rate of the reporting
data transmission of a smart meter during a reporting dura-
tion. Clearly, how to minimize the overall ILR (ensuring
QoS) with a large number of reporting smart meters is quite
challenging.
b) Secure M2M system: Security (authentication and data
integrity) is important for human-life related M2M systems
such as energy/industrial M2M systems and life-supporting
M2M systems [6]. Since most of the M2M communications
are wireless, eavesdropping becomes extremely simple [44].
Moreover, most of the M2M devices are characterized by low
capabilities in terms of both energy and computing resources
and thus, they cannot implement complex schemes supporting
security.
Meanwhile, privacy is more and more important for peo-
ple in the information era. For example, the data generated
by a smart meter can reflect the behavior of the residents.
Once such data is revealed, the risk of pilferage increases. In
fact, the methods of data collection, mining, and provision-
ing in the M2M system are completely different from those
in traditional systems and there will be an surprising number
of occasions for personal data to be collected. Therefore, for
human individuals it will be impossible to personally control
the disclosure of their personal information. To realize secure
M2M systems, novel schemes are in urgent need to utilize the
features of M2M systems [45], [46].
CAO et al.: SURVEY OF EMERGING M2M SYSTEMS: CONTEXT, TASK, AND OBJECTIVE
c) Reliable M2M system: A reliable M2M system can
operate robustly in occasion of component faults, unreliable
transmissions and so on. When a failure occurs, a single M2M
node may not be sufficient to accurately sense and process
monitoring data. Therefore, redundancy (increasing the num-
ber of M2M nodes and transmission times) is an effective
solution to enhance the reliability [6]. Moreover, reliable M2M
systems are required to provide ultrastable performance, e.g.,
fixed transmission rate [47]. To achieve the ultrastable per-
formance, interference-aware resource scheduling methods are
expected to manage the interference among multiple M2M
communication links according to real-time information on
the wireless environment.
2) Consumption-Oriented M2M Systems:
a) Energy-efficient M2M system: Mass-produced M2M
devices (e.g., sensors and tags) are usually powered by the
battery with limited capacity, thus an energy-efficient M2M
system is recommended. Energy saving can be achieved by
optimizing the sensing, processing, and transmissions of M2M
devices to prolong the battery lifetime. Since a mass of M2M
devices can be deployed in the M2M systems, it is benefi-
cial to cut down on the total energy consumption through
improving the energy efficiency for each M2M device. In
addition, since the cellular base station is also a power-
consuming component in M2M communications, it is valuable
to improve the energy efficiency of the base station to achieve
environmentally friendly, green M2M communications [6].
b) Cost-effective M2M system: To make M2M systems
more affordable by the public, it is important to cut down
on the cost [48]. Such cost can be divided into two main
parts: 1) the implementation cost of M2M devices/M2M
servers/M2M communication networks and 2) the maintenance
cost of M2M devices/M2M servers/M2M communication
networks. The implementation cost includes the cost of pur-
chasing M2M devices and related value-adding services from
the M2M device vendors, the cost of renting M2M servers
from Internet data centers and the cost of setting up M2M
communication networks. The maintenance cost includes the
electricity cost of powering M2M devices/M2M servers/M2M
communication infrastructures and the cost of using network
bandwidth/spectrum. To realize a cost-effective M2M system,
it is important to choose proper communication and computing
solutions for M2M systems and make the best use of network
bandwidth/spectrum and energy resources.
Remarks: Please note that, by categorizing M2M systems
into performance-oriented M2M systems and consumption-
oriented M2M systems, one can focus on the design criterion
of the M2M system solution. Generally, different types of
M2M objectives place different requirements for determining
communication solutions (e.g., low date rate Bluetooth/ZigBee
or high data rate WiFi/LTE) and computing solutions (e.g.,
microcontroller or high-performance multicore processors).
D. Summary of Three Aspects for M2M
System Categorization
The above three aspects, namely, M2M context, M2M
task, and M2M objective, are highlighted in the proposed
Authorized licensed use limited to: KU Leuven Libraries. Downloaded on December 29,2023 at 11:25:06 UTC from IEEE Xplore. Restrictions apply.
1253
M2M system categorization. Clearly speaking, M2M con-
text emphasizes the topology of the M2M system (dynamic
or static), M2M task emphasizes the characteristic of M2M
system (device-centric or system-centric), and the M2M objec-
tive emphasizes the design criterion of the M2M system
(performance-oriented or consumption-oriented).
Note that, an aspect can affect one other aspect and
vice versa. For example, the solution for a dynamic M2M
system should strike a balance between mobility and perfor-
mance if its M2M objective is performance-oriented, since
the M2M device with a high mobility usually has constrained
resources [49]. Another example is that the solution for a
consumption-oriented M2M system can utilize the fixed topol-
ogy to lower the cost if its M2M context is static, i.e.,
M2M context provides an opportunity (fixed topology) to
M2M objective. In summary, it can be quite beneficial to
propose solutions for M2M systems by jointly considering
three aspects and exploiting opportunities brought by different
aspects.
IV. S URVEY OF R ECENT S OLUTIONS FOR M2M S YSTEMS
In Section III, we propose an M2M system categoriza-
tion and summarize the three aspects of the categorization.
Based on the proposed M2M system categorization, we not
only investigate the literature of improving the performance
of M2M systems but also practical M2M platforms from
industrial enterprises in this section. Since an M2M sys-
tem can be jointly characterized by the M2M context, M2M
task, and M2M objective, we use a context-task-objective
format to represent the subcategories. For example, we use
“static-system-performance” to label a static, system-centric
and performance-oriented M2M system.
Recently, some notable solutions have been proposed by
academia and industry for M2M systems. As summarized
in Table I, solutions surveyed in this section are categorized
according to the proposed M2M system categorization, where
practical solutions are marked with *. In the following part,
the survey of solutions is separated into two parts: the first
part is theoretical solutions for M2M systems and the second
part is practical solutions for M2M systems.
A. Theoretical Solutions for M2M Systems
1) Static-System-Performance M2M Systems: For a static
M2M system, e.g., energy M2M system, M2M devices are
in fixed locations. Fadlullah et al. [39] studied the energy
M2M system for smart grid with a system-centric M2M task,
i.e., fault prediction. Specifically, they modeled the malicious
and/or abnormal events, which can compromise the security
and privacy of smart grid users, as a Gaussian process. Based
on this model, a novel early warning scheme was proposed
for anticipating malicious events. With the warning scheme,
the control center can forecast such malicious events, and thus
enable smart grid to react before hand and mitigate the possible
impact of malicious activity.
2) Static-Device-Performance M2M Systems: Other M2M
solutions focus on static M2M systems with device-centric
M2M tasks, e.g., monitoring and sensing. Fouda et al. [37]
1254
TABLE I
S URVEY OF M2M S OLUTIONS BASED ON THE P ROPOSED M2M S YSTEM C ATEGORIZATION
proposed a light-weight message authentication scheme for
secure communication amongst smart meters in the power dis-
tribution network. Ko et al. [53] proposed a random access
scheme for the 3GPP M2M system, which makes use of
fixed timing alignment information at a large number of fixed-
location machine devices to reduce collision probability and
lower average access delay, and thus the QoS is ensured.
3) Static-Device-Consumption M2M Systems: Clearly,
above solutions [37], [39], [53] are performance-oriented
M2M systems. By contrast, the following solutions are for
consumption-oriented M2M systems with static topology and
device-centric tasks. When a large share of smart meters trans-
mit data via the cellular network at configured points in time,
e.g., at midnight or every hour, high periodic traffic peaks
can be formed. Paiva et al. [54] proposed a method to reduce
the peak load on the access grant channel, which is based
on the property of the M2M traffic that many M2M devices
are stationary and connect at a regular rhythm, and thus the
total cost of using the spectrum can be reduced. Similarly,
Wang and Wong [55] exploited timing advance information
as well as the access class barring (ACB) to reduce random
access overload in M2M systems. By optimizing the ACB
parameter, the expected number of served M2M devices in
each random access slot can be maximized.
Another method of reducing the cost is to use previ-
ous generations of cellular technologies. Generally, with the
progress of LTE/LTE-Advanced networks, previous gener-
ations of cellular technologies like GSM, general packet
radio service (GPRS), and wideband code division multiple
access (WCDMA) are less attractive to high end applica-
tions like video streaming. However, previous generations of
cellular technologies can still be valuable to support cost-
effective M2M systems, that is, many M2M devices (that
only require low throughput) can be sufficiently served by
legacy technologies such as GPRS. Ming et al. [56] investi-
gated the efficiency of three main GSM/GPRS bearers used
for MTCs. Krnic and Krco [57] presented the initial results
of a WCDMA radio access network capacity calculation and
analysis as affected by a few selected wireless sensor network
(WSN)-based applications in M2M systems.
Authorized licensed use limited to: KU Leuven Libraries. Downloaded on December 29,2023 at 11:25:06 UTC from IEEE Xplore. Restrictions apply.
IEEE INTERNET OF THINGS JOURNAL, VOL. 3, NO. 6, DECEMBER 2016
How to embed low-cost WSN to M2M systems is also
important to achieve cost effectiveness. Zhang et al. [35] stated
that M2M and WSN are evolving from heterogeneous to con-
verged, in order to meet the increasing requirement for M2M
communications with relatively low costs. Considering the
fact that traditional cellular spectrum is expensive and over-
crowded, it is necessary to utilize other spectrum band to
support low-cost M2M communications. Zhang et al. [58] first
surveyed the use of cognitive radio technology [74] in M2M
systems from the aspects of applications, industry support, and
standardization. Then, the proposed cognitive M2M network
architecture and cognitive machine model in television white
spaces were presented.
4) Dynamic-Device-Performance M2M Systems: For a
dynamic M2M system like vehicular M2M system, M2M
devices are in variable locations. Ma et al. [60] studied V2V
communication problems in the vehicular M2M system for
a device-centric M2M task, i.e., cooperative safety message
broadcast (device cooperation). Specifically, a novel protocol
for reliable and fast delivery of safety-related messages using
the control channel of dedicated short range communications
was proposed to enhance the reliability of delay-sensitive data
transmissions.
For general dynamic M2M systems, cellular networks can
serve as access networks for long range wireless communi-
cations among a large amount of mobile devices and Internet
M2M servers. Clearly, latency problem introduced by schedul-
ing overhead is quite challenging. Nikaein and Krea [61]
proposed an M2M system based on LTE/LTE-advanced cel-
lular network, and the delays associated with each part of the
system were highlighted. Three real-time M2M tasks were
analyzed and the main latency bottlenecks were identified.
Brown and Khan [62] proposed a predictive resource allo-
cation algorithm for LTE-based M2M systems to exploit the
correlation of the traffic pattern of M2M devices in a group.
Therefore, the latency is reduced compared to waiting for each
device to request resources.
To reduce the access delay of M2M systems, there are
some solutions focused on M2M random access channels in
3GPP cellular networks. A recent report deeply presented the
CAO et al.: SURVEY OF EMERGING M2M SYSTEMS: CONTEXT, TASK, AND OBJECTIVE
fact that trillions of machine devices in 3GPP cellular net-
works may suffer from severe collisions at the random access
channels which are used for network entry/reentry, handover,
scheduling request, and uplink timing synchronization [75].
Lien et al. [63] proposed a cooperative ACB scheme for global
stabilization and access load sharing to eliminate substantial
defects in the ordinary ACB, and thus access delays in random
access channels are significantly reduced in M2M systems.
Similarly, Lin and Wei [64] proposed an auction-based scheme
to balance the traffic load of random access channel. The suc-
cess probability can be increased due to lower access delay.
Kim et al. [65] proposed to reuse the preamble resources based
on a spatial grouping during the random access procedure, so
as to cut down on the access delay while accommodating a
significantly large number of M2M devices.
Some other solutions considered requirement diversity of
M2M tasks. Lien and Chen [66] proposed a solution to
the critical issue on the air interface of M2M communica-
tions to effectively perform M2M task for a large amount
of M2M devices with enormously diverse characteristics.
Zheng et al. [67] developed scheduling schemes based on
requirements of different M2M tasks.
5) Dynamic-Device-Consumption M2M Systems: Since
battery lifetime is an essential constraint for mobile M2M
devices, energy efficiency is attracting extensive attentions.
Some energy efficient solutions utilize dynamic M2M sys-
tem features such as multihop transmissions, burst traffic, and
high mobility. Andreev et al. [68] stated that it is not nec-
essary to explicitly schedule the data packet transmissions
when all the data packets from M2M devices are suffi-
ciently short to be transmitted within their random access
requests. Predojev et al. [69] evaluated the achievable energy
saving-delay tradeoff in embedded multihop wireless net-
works with the focus on the channel coding. Fu et al. [70]
proposed intelligent reporting mechanisms to deal with the
energy saving-validity tradeoff of real-time monitoring tasks
for mobile M2M devices.
Some other solutions have been proposed as energy effi-
ciency solutions via access control and grouping schemes for
M2M devices. Rost et al. [71] proposed and evaluated a novel
asymmetric user assignment scheme, which enables operators
to save network energy, while satisfying the capacity demands.
Tu et al. [72] proposed a scheme to improve the energy effi-
ciency by implementing grouping and coordinator selection,
in which grouping limits the amount of access from mobile
device to the base station and the coordinator selection guar-
antees low energy consumption. To minimize the total energy
consumption of the M2M system, Ho and Huang [73] pro-
posed joint massive access control and resource allocation
scheme, which first performs machine grouping, coordinator
selection, and resource allocation, and then determines the
number of groups.
B. Practical Solutions for M2M Systems
1) Static-System-Performance M2M Systems:
a) Silver spring smart street lights [50]: This M2M
system features remote management and real-time outage
Authorized licensed use limited to: KU Leuven Libraries. Downloaded on December 29,2023 at 11:25:06 UTC from IEEE Xplore. Restrictions apply.
1255
response for street lighting. Networking gives operators remote
access and advanced functionality, including the ability to dim
street lights and control their runtime by scheduling them
to switch on/off as conditions (such as shorter/longer days)
warrant. Moreover, networked street lights provide continu-
ous, accurate status information to operators (QoS guarantee),
enabling them to identify outages immediately.
2) Static-System-Consumption M2M Systems:
a) eyeQ [51]: eyeQ harnesses in-store shopper data to
combine the personalization and product information of online
shopping with the physical interaction and instant gratification
of traditional retail. Gleaning insights from analytics of in-
store cameras and touch-screen kiosks, eyeQ helps individual
shoppers find what they need and want, and helps retailers
sell it to them. Based on IBM’s Open Cloud Architecture and
Cloud Foundry, the M2M system can be efficient and cost-
effective.
b) SIEMENS integrated smart parking solution [52]:
This M2M system enables users to effortlessly form a clear
picture of where available parking spaces can be found and
how long each space has been occupied for. The Siemens
Integrated Smart Parking Solution is a modular, infrastructure-
based sensor system that goes beyond the possibilities of
ground sensors. By obtaining statistical and real-time informa-
tion on parking vacancies, intelligent data collection simplifies
sensible decisions on parking prices and regulations to meet
strategic targets, while allowing for detection of parking
violations in a cost-effective manner.
3) Dynamic-System-Performance M2M Systems:
a) AGT and Cisco traffic incident management
solution [59]: This M2M system aims at solving road
safety and congestion issues and improving the traffic safety
and efficiency. It helps traffic departments accurately detect
more incidents before they become more serious, and enables
quicker response by monitoring and analyzing traffic flow
data. The solution stands out with two important innovations:
the fusing of data (e.g., average vehicle speed, traffic volume,
and lane occupancy) from multiple sensors and cameras to
identify real incidents and reduce false alarms (reliability
guarantee). Moreover, it analyzes the distribution of traffic
incidents and reports and helps predict traffic flow patterns.
Summary: Obviously, a majority of solutions are about
static/dynamic-device-consumption and dynamic-device-
performance M2M systems. Device-centric M2M tasks bring
the challenge of intense data transmissions, and consumption
reduction is always the most important issue for resource
constrained M2M systems. Dynamic topology can result
in unstable wireless conditions, thus it is more difficult to
ensure QoS and reliable communications. Existing solutions
have not well investigated massive concurrent data transmis-
sions with QoS guarantee, configuration and management
overheads/costs, etc.
V. C HALLENGES IN D EVELOPING
M2M S YSTEM S OLUTIONS
In the previous section, we review the existing solutions for
M2M systems according to the proposed categorization, and
1256
Fig. 6. Underlaying D2D communications for M2M systems.
summarize the main challenges that are not well investigated
by existing solutions. In this section, we discuss several main
challenges in developing solutions for M2M systems from the
following aspects.
A. Enabling Massive Concurrent Data Transmissions for
M2M Systems
When a massive number of M2M devices in a dynamic
M2M system access the base station simultaneously, it is quite
challenging to handle access requests over control channels.
As surveyed in the previous section, some literatures proposed
schemes to reduce signaling overheads over control channels,
and some other literatures proposed grouping schemes to form
multiple M2M devices into a group and select a coordinator for
such group, and thus the number of M2M devices that directly
connect to the base station is reduced. However, how to handle
massive concurrent data transmissions over data channels is
still an open problem.
Device-to-device (D2D) communication [76], [77] is a pow-
erful mechanism of cutting down on the demand for more
spectrum resources, which enables direct communications
between two devices in proximity by reusing the cellular
spectrum, without traversing the base station and the core
network. As depicted in Fig. 6, network-assisted underlay-
ing D2D communications enable the cooperation between a
number of M2M devices and a standard cellular device (as an
M2M gateway) by reusing the data channels of other cellu-
lar devices, thereby meeting the QoS requirements of M2M
communications without affecting the transmissions of cel-
lular devices [78]. The interference management [79], [80]
of underlaying D2D communications for M2M systems is an
interesting aspect, which has not been well investigated yet.
B. Configuration and Management of M2M Systems
The configuration and management of M2M systems can
be quite challenging due to: 1) enormous overheads of signal-
ing and resource scheduling and 2) extremely heterogeneous
M2M devices (e.g., Siemens sensor on a Verizon data plan
may work with an ABB actuator on an ATT data plan).
Current resource scheduling solutions mostly adopt central-
ized scheduling methods with high complexity, which may
not work well in practical systems with a large number of
devices.
Authorized licensed use limited to: KU Leuven Libraries. Downloaded on December 29,2023 at 11:25:06 UTC from IEEE Xplore. Restrictions apply.
IEEE INTERNET OF THINGS JOURNAL, VOL. 3, NO. 6, DECEMBER 2016
It can be beneficial to develop distributed and low-
complexity scheduling methods for large-scale M2M systems.
Game theory [77] can be one of the useful tools for distributed
scheduling, especially for the M2M systems [81]. However,
reducing the computational complexity of scheduling schemes
while providing a theoretical performance guarantee is quite
challenging. Another trend is to offload configuration and man-
agement tasks of M2M systems to clouds/data centers, thus the
scheduling overhead can be reduced [82].
C. Coexistence of M2M Systems and Legacy Systems
The coexistence of M2M systems and legacy systems [e.g.,
human-to-human (H2H)] has not been fully investigated yet.
It is possible that M2M systems and legacy systems share
the same access network. For example, the operator of an
LTE network should handle control signals and data pack-
ets from M2M devices and legacy devices at the same time in
M2M/H2H coexistence scenarios, which is a key challenging
issue [83].
The share of communication and networking infrastructures
can greatly reduce the cost of deploying and managing M2M
systems. However, effective radio resource allocation schemes
are in an urgent need for M2M/H2H coexistence scenarios
to achieve QoS guarantee and energy/cost efficiency. It is
also interesting to study the spectrum sharing and coopera-
tion between M2M devices and legacy devices via cognitive
radio technologies to achieve a win-win situation [84].
D. User Satisfaction for M2M Systems
Some design criterions of M2M systems would conflict
with each other, e.g., QoS guarantee and low cost. Thus, it
is challenging to find a proper tradeoff among multiple con-
flicting design criterions. Studies that jointly consider multiple
criterions are attracting more and more attentions [49], [69].
One of the possible trends is to consider the user sat-
isfaction of M2M systems. For example, a proper tradeoff
between the performance and the consumption would improve
the user satisfaction of the M2M system, thus more clients
would accept M2M system services. Generally, considering
user satisfaction means not to pursuit QoS solely. User satis-
faction can characterize the match between user expectations
and delivered M2M services. Inappropriately high QoS may
waste system resources compared to a lower but adequate QoS.
Thus, given a specific M2M system, higher QoS does not
always make for higher user satisfaction. Moreover, the opti-
mization of the user satisfaction with respect to QoS is still
recognized as NP-complete, i.e., computationally impractical.
How to improve the user satisfaction of M2M systems is an
interesting and challenging problem.
VI. C ONCLUSION
M2M systems are attracting persistent attention recently
and many solutions have been proposed for M2M systems. In
this survey paper, we have categorized M2M systems accord-
ing to the types of M2M context, M2M task, and M2M
objective. Based on the proposed M2M system categorization,
recent solutions for M2M systems from both academia and
CAO et al.: SURVEY OF EMERGING M2M SYSTEMS: CONTEXT, TASK, AND OBJECTIVE
industry have been surveyed. We have summarized that M2M
systems can be jointly characterized by the types of M2M
context, M2M task, and M2M objective, and it is crucial to
propose effective solutions considering the type of an M2M
system. Furthermore, we have summarized main challenges of
developing M2M system solutions, namely, enabling massive
concurrent data transmissions for M2M systems, configura-
tion and management of M2M systems, coexistence of M2M
systems and legacy systems, and user satisfaction for M2M
systems.
R EFERENCES
[1] J. Kim, J. Lee, J. Kim, and J. Yun, “M2M service platforms: Survey,
issues, and enabling technologies,” IEEE Commun. Surveys Tuts.,
vol. 16, no. 1, pp. 61–76, 1st Quart. 2014.
[2] G. Lawton, “Machine-to-machine technology gears up for growth,”
Computer, vol. 37, no. 9, pp. 12–15, Sep. 2004.
[3] D. Evans, “The Internet of Things: How the next evolution of the Internet
is changing everything,” CISCO White Paper, 2011.
[4] M. Weyrich, J.-P. Schmidt, and C. Ebert, “Machine-to-machine commu-
nication,” IEEE Softw., vol. 31, no. 4, pp. 19–23, Jul./Aug. 2014.
[5] Y. Yan, Y. Qian, H. Sharif, and D. Tipper, “A survey on smart grid com-
munication infrastructures: Motivations, requirements and challenges,”
IEEE Commun. Surveys Tuts., vol. 15, no. 1, pp. 5–20, 1st Quart. 2013.
[6] R. Lu, X. Li, X. Liang, X. Shen, and X. Lin, “GRS: The green, reliability,
and security of emerging machine to machine communications,” IEEE
Commun. Mag., vol. 49, no. 4, pp. 28–35, Apr. 2011.
[7] T. Taleb and A. Kunz, “Machine type communications in 3GPP net-
works: Potential, challenges, and solutions,” IEEE Commun. Mag.,
vol. 50, no. 3, pp. 178–184, Mar. 2012.
[8] G. Wu, S. Talwar, K. Johnsson, N. Himayat, and K. D. Johnson, “M2M:
From mobile to embedded Internet,” IEEE Commun. Mag., vol. 49, no. 4,
pp. 36–43, Apr. 2011.
[9] Z. Fan, R. J. Haines, and P. Kulkarni, “M2M communications for
e-health and smart grid: An industry and standard perspective,” IEEE
Wireless Commun., vol. 21, no. 1, pp. 62–69, Feb. 2014.
[10] M. Chen, J. Wan, S. Gonzalez, X. Liao, and V. C. M. Leung, “A sur-
vey of recent developments in home M2M networks,” IEEE Commun.
Surveys Tuts., vol. 16, no. 1, pp. 98–114, 1st Quart. 2014.
[11] K. Chang, A. Soong, M. Tseng, and Z. Xiang, “Global wireless machine-
to-machine standardization,” IEEE Internet Comput., vol. 15, no. 2,
pp. 64–69, Mar./Apr. 2011.
[12] OneM2M. (2015). OneM2M: The Interoperability Enabler for the
Entire M2M Ecosystem. [Online]. Available: http://www.onem2m.org/
images/files/oneM2M-whitepaper-January-2015.pdf
[13] ETSI, “Machine-to-machine communications (M2M); functional archi-
tecture,” Eur. Telecommun. Standards Inst., Sophia Antipolis, France,
TS 102.690 v1.1.1, 2011.
[14] 3GPP, “Service requirements for machine-type communications,” TS
22.368 v12.0.0, 2012.
[15] 3GPP, “System improvements for machine-type communications,” TR
23.888 v11.0.0, 2012.
[16] B. C. Villaverde et al., “Service discovery protocols for constrained
machine-to-machine communications,” IEEE Commun. Surveys Tuts.,
vol. 16, no. 1, pp. 41–60, 1st Quart. 2014.
[17] X. Xiong, K. Zheng, R. Xu, W. Xiang, and P. Chatzimisios, “Low power
wide area machine-to-machine networks: Key techniques and prototype,”
IEEE Commun. Mag., vol. 53, no. 9, pp. 64–71, Sep. 2015.
[18] J. Gozalvez, “New 3GPP standard for IoT [mobile radio],” IEEE Veh.
Technol. Mag., vol. 11, no. 1, pp. 14–20, Mar. 2016.
[19] G. Karagiannis et al., “Vehicular networking: A survey and tutorial on
requirements, architectures, challenges, standards and solutions,” IEEE
Commun. Surveys Tuts., vol. 13, no. 4, pp. 584–616, 4th Quart. 2011.
[20] I. Lequerica, M. G. Longaron, and P. M. Ruiz, “Drive and share: Efficient
provisioning of social networks in vehicular scenarios,” IEEE Commun.
Mag., vol. 48, no. 11, pp. 90–97, Nov. 2010.
[21] H. T. Cheng, H. Shan, and W. Zhuang, “Infotainment and road safety
service support in vehicular networking: From a communication per-
spective,” Mech. Syst. Signal Process., vol. 25, no. 6, pp. 2020–2038,
2011.
Authorized licensed use limited to: KU Leuven Libraries. Downloaded on December 29,2023 at 11:25:06 UTC from IEEE Xplore. Restrictions apply.
1257
[22] H. Nishiyama, T. Ngo, S. Oiyama, and N. Kato, “Relay by smart device:
Innovative communications for efficient information sharing among
vehicles and pedestrians,” IEEE Veh. Technol. Mag., vol. 10, no. 4,
pp. 54–62, Dec. 2015.
[23] P. Phunchongharn, E. Hossain, D. Niyato, and S. Camorlinga, “A cog-
nitive radio system for e-health applications in a hospital environment,”
IEEE Wireless Commun., vol. 17, no. 1, pp. 20–28, Feb. 2010.
[24] S. Adibi, “Link technologies and BlackBerry mobile health (mHealth)
solutions: A review,” IEEE Trans. Inf. Technol. Biomed., vol. 16, no. 4,
pp. 586–597, Jul. 2012.
[25] L. Atzori, A. Iera, and G. Morabito, “The Internet of Things: A survey,”
Comput. Netw., vol. 54, no. 15, pp. 2787–2805, 2010.
[26] T. Jiang, Y. Cao, L. Yu, and Z. Wang, “Load shaping strategy based
on energy storage and dynamic pricing in smart grid,” IEEE Trans.
Smart Grid, vol. 5, no. 6, pp. 2868–2876, Nov. 2014.
[27] A. Q. Huang, M. L. Crow, G. T. Heydt, J. P. Zheng, and S. J. Dale, “The
future renewable electric energy delivery and management (FREEDM)
system: The energy Internet,” Proc. IEEE, vol. 99, no. 1, pp. 133–148,
Jan. 2011.
[28] Y. Cao, T. Jiang, and Q. Zhang, “Reducing electricity cost of smart appli-
ances via energy buffering framework in smart grid,” IEEE Trans.
Parallel Distrib. Syst., vol. 23, no. 9, pp. 1572–1582, Sep. 2012.
[29] Z. M. Fadlullah et al., “Toward intelligent machine-to-machine commu-
nications in smart grid,” IEEE Commun. Mag., vol. 49, no. 4, pp. 60–65,
Apr. 2011.
[30] D. Niyato, L. Xiao, and P. Wang, “Machine-to-machine communications
for home energy management system in smart grid,” IEEE Commun.
Mag., vol. 49, no. 4, pp. 53–59, Apr. 2011.
[31] R. Fantacci, T. Pecorella, R. Viti, and C. Carlini, “A network archi-
tecture solution for efficient IoT WSN backhauling: Challenges and
opportunities,” IEEE Wireless Commun., vol. 21, no. 4, pp. 113–119,
Aug. 2014.
[32] R. Drath and A. Horch, “Industrie 4.0: Hit or hype? [industry forum],”
IEEE Ind. Electron. Mag., vol. 8, no. 2, pp. 56–58, Jun. 2014.
[33] S. He, D.-H. Shin, J. Zhang, J. Chen, and Y. Sun, “Full-view
area coverage in camera sensor networks: Dimension reduction and
near-optimal solutions,” IEEE Trans. Veh. Technol., to be published,
doi: 10.1109/TVT.2015.2498281.
[34] S. Bosch, M. Marin-Perianu, R. Marin-Perianu, H. Scholten, and
P. Havinga, “Follow me! mobile team coordination in wireless sensor
and actuator networks,” in Proc. IEEE Int. Conf. Pervasive Comput.
Commun. (PerCom), Galveston, TX, USA, Mar. 2009, pp. 1–11.
[35] J. Zhang, L. Shan, H. Hu, and Y. Yang, “Mobile cellular networks and
wireless sensor networks: Toward convergence,” IEEE Commun. Mag.,
vol. 50, no. 3, pp. 164–169, Mar. 2012.
[36] W. Luan, D. Sharp, and S. Lancashire, “Smart grid communication net-
work capacity planning for power utilities,” in Proc. IEEE PES Transm.
Distrib. Conf. Expo., New Orleans, LA, USA, Apr. 2010, pp. 1–4.
[37] M. M. Fouda, Z. M. Fadlullah, N. Kato, R. Lu, and X. S. Shen,
“A lightweight message authentication scheme for smart grid com-
munications,” IEEE Trans. Smart Grid, vol. 2, no. 4, pp. 675–685,
Dec. 2011.
[38] H. Zhang, P. Cheng, L. Shi, and J. Chen, “Optimal DoS attack schedul-
ing in wireless networked control system,” IEEE Trans. Control Syst.
Technol., vol. 24, no. 3, pp. 843–852, May 2016.
[39] Z. M. Fadlullah, M. M. Fouda, N. Kato, X. Shen, and Y. Nozaki,
“An early warning system against malicious activities for smart grid
communications,” IEEE Netw., vol. 25, no. 5, pp. 50–55, Sep./Oct. 2011.
[40] C.-W. Tsai, C.-F. Lai, M.-C. Chiang, and L. T. Yang, “Data mining for
Internet of Things: A survey,” IEEE Commun. Surveys Tuts., vol. 16,
no. 1, pp. 77–97, 1st Quart. 2014.
[41] X. Wu, X. Zhu, G.-Q. Wu, and W. Ding, “Data mining with big data,”
IEEE Trans. Knowl. Data Eng., vol. 26, no. 1, pp. 97–107, Jan. 2014.
[42] R. Bo and F. Li, “Probabilistic LMP forecasting considering load
uncertainty,” IEEE Trans. Power Syst., vol. 24, no. 3, pp. 1279–1289,
Aug. 2009.
[43] Y. Cao, T. Jiang, M. He, and J. Zhang, “Device-to-device communica-
tions for energy management: A smart grid case,” IEEE J. Sel. Areas
Commun., vol. 34, no. 1, pp. 190–201, Jan. 2016.
[44] Y. Zou, J. Zhu, X. Wang, and L. Hanzo, “A survey on wireless security:
Technical challenges, recent advances, and future trends,” Proc. IEEE,
to be published, doi: 10.1109/JPROC.2016.2558521.
[45] J. Sun, R. Zhang, and Y. Zhang, “Privacy-preserving spatiotemporal
matching,” in Proc. IEEE Int. Conf. Comput. Commun. (INFOCOM),
Turin, Italy, Apr. 2013, pp. 800–808.
1258
[46] J. Sun, R. Zhang, and Y. Zhang, “Privacy-preserving spatiotemporal
matching for secure device-to-device communications,” IEEE Internet
Things J., to be published, doi: 10.1109/JIOT.2016.2549046.
[47] N. K. Pratas and P. Popovski, “Underlay of low-rate machine-type
D2D links on downlink cellular links,” in Proc. IEEE Int. Conf.
Commun. (ICC) Workshops, Sydney, NSW, Australia, Jun. 2014,
pp. 423–428.
[48] S. He, J. Chen, X. Li, X. Shen, and Y. Sun, “Mobility and intruder prior
information improving the barrier coverage of sparse sensor networks,”
IEEE Trans. Mobile Comput., vol. 13, no. 6, pp. 1268–1282, Jun. 2014.
[49] J. Sun, R. Zhang, X. Jin, and Y. Zhang, “SecureFind: Secure and
privacy-preserving object finding via mobile crowdsourcing,” IEEE
Trans. Wireless Commun., vol. 15, no. 3, pp. 1716–1728, Mar. 2016.
[50] S. Spring. (2016). Smart Street Lighting Providing
Maximum Value for Utilities and Cities. [Online]. Available:
http://www.silverspringnet.com/solutions-technology/street-lights
[51] EyeQ. (2016). The New Power of Personalization. [Online]. Available:
https://www.eyeqinsights.com
[52] SIEMENS. (2016). Smart City Vision: Movement Through Innovation.
[Online]. Available: http://www.mobility.siemens.com/mobility/global/
en/urban-mobility/road-solutions/integrated-smart-parking-solution
[53] K. S. Ko et al., “A novel random access for fixed-location machine-to-
machine communications in OFDMA based systems,” IEEE Commun.
Lett., vol. 16, no. 9, pp. 1428–1431, Sep. 2012.
[54] R. C. D. Paiva, H. Wilhelm, M. Saily, D. Navratil, and M. Taponen,
“Overload control method for synchronized MTC traffic in GERAN,”
in Proc. IEEE Veh. Technol. Conf. (VTC Fall), San Francisco, CA, USA,
Sep. 2011, pp. 1–5.
[55] Z. Wang and V. W. S. Wong, “Optimal access class barring for stationary
machine type communication devices with timing advance informa-
tion,” IEEE Trans. Wireless Commun., vol. 14, no. 10, pp. 5374–5387,
Oct. 2015.
[56] F. Ming, X. Zhu, M. Torres, L. Anaya, and L. Patanapongpibul,
“GSM/GPRS bearers efficiency analysis for machine type communi-
cations,” in Proc. IEEE Veh. Technol. Conf. (VTC Spring), Yokohama,
Japan, May 2012, pp. 1–5.
[57] J. Krnic and S. Krco, “Impact of WSN applications’ generated traffic
on WCDMA access networks,” in Proc. IEEE Int. Symp. Pers. Indoor
Mobile Radio Commun. (PIMRC), Cannes, France, Sep. 2008, pp. 1–5.
[58] Y. Zhang et al., “Cognitive machine-to-machine communications:
Visions and potentials for the smart grid,” IEEE Netw., vol. 26, no. 3,
pp. 6–13, May/Jun. 2012.
[59] Cisco. (2014). AGT and Cisco Traffic Incident Management
Solution: Improving Traffic Safety and Efficiency. [Online]. Available:
http://www.cisco.com/web/strategy/docs/agt-s-cc-city-tim-aag.pdf
[60] X. Ma, J. Zhang, X. Yin, and K. S. Trivedi, “Design and analysis of
a robust broadcast scheme for VANET safety-related services,” IEEE
Trans. Veh. Technol., vol. 61, no. 1, pp. 46–61, Jan. 2012.
[61] N. Nikaein and S. Krea, “Latency for real-time machine-to-machine
communication in LTE-based system architecture,” in Proc. Eur.
Wireless, Vienna, Austria, Apr. 2011, pp. 1–6.
[62] J. Brown and J. Y. Khan, “A predictive resource allocation algorithm in
the LTE uplink for event based M2M applications,” IEEE Trans. Mobile
Comput., vol. 14, no. 12, pp. 2433–2446, Dec. 2015.
[63] S.-Y. Lien, T.-H. Liau, C.-Y. Kao, and K.-C. Chen, “Cooperative access
class barring for machine-to-machine communications,” IEEE Trans.
Wireless Commun., vol. 11, no. 1, pp. 27–32, Jan. 2012.
[64] G.-Y. Lin and H.-Y. Wei, “Auction-based random access load control for
time-dependent machine-to-machine communications,” IEEE Internet
Things J., to be published, doi: 10.1109/JIOT.2015.2480070.
[65] T. Kim, H. S. Jang, and D. K. Sung, “An enhanced random access
scheme with spatial group based reusable preamble allocation in cellular
M2M networks,” IEEE Commun. Lett., vol. 19, no. 10, pp. 1714–1717,
Oct. 2015.
[66] S.-Y. Lien and K.-C. Chen, “Massive access management for QoS guar-
antees in 3GPP machine-to-machine communications,” IEEE Commun.
Lett., vol. 15, no. 3, pp. 311–313, Mar. 2011.
[67] K. Zheng, F. Hu, W. Wang, W. Xiang, and M. Dohler, “Radio
resource allocation in LTE-advanced cellular networks with M2M
communications,” IEEE Commun. Mag., vol. 50, no. 7, pp. 184–192,
Jul. 2012.
[68] S. Andreev, O. Galinina, and Y. Koucheryavy, “Energy-efficient client
relay scheme for machine-to-machine communication,” in Proc. IEEE
Glob. Telecommun. Conf. (GLOBECOM), Houston, TX, USA, 2011,
pp. 1–5.
Authorized licensed use limited to: KU Leuven Libraries. Downloaded on December 29,2023 at 11:25:06 UTC from IEEE Xplore. Restrictions apply.
IEEE INTERNET OF THINGS JOURNAL, VOL. 3, NO. 6, DECEMBER 2016
[69] T. Predojev, J. Alonso-Zarate, and M. Dohler, “Energy-delay trade-
off analysis in embedded M2M networks with channel coding,” in
Proc. IEEE Int. Symp. Pers. Indoor Mobile Radio Commun. (PIMRC),
Istanbul, Turkey, 2010, pp. 2733–2738.
[70] H.-L. Fu, H.-C. Chen, P. Lin, and Y. Fang, “Energy-efficient report-
ing mechanisms for multi-type real-time monitoring in machine-to-
machine communications networks,” in Proc. IEEE Int. Conf. Comput.
Commun. (INFOCOM), Orlando, FL, USA, Mar. 2012, pp. 136–144.
[71] P. Rost, A. Maeder, and X. Perez-Costa, “Asymmetric uplink-downlink
assignment for energy-efficient mobile communication systems,” in
Proc. IEEE VTC Spring, Yokohama, Japan, May 2012, pp. 1–5.
[72] C.-Y. Tu, C.-Y. Ho, and C.-Y. Huang, “Energy-efficient algorithms
and evaluations for massive access management in cellular based
machine to machine communications,” in Proc. IEEE Veh. Technol.
Conf. (VTC Fall), San Francisco, CA, USA, Sep. 2011, pp. 1–5.
[73] C. Y. Ho and C.-Y. Huang, “Energy-saving massive access con-
trol and resource allocation schemes for M2M communications in
OFDMA cellular networks,” IEEE Wireless Commun. Lett., vol. 1, no. 3,
pp. 209–212, Jun. 2012.
[74] Y. Cao, D. Qu, and T. Jiang, “Throughput maximization in cognitive
radio system with transmission probability scheduling and traffic pattern
prediction,” Mobile Netw. Appl., vol. 17, no. 5, pp. 604–617, 2012.
[75] ZTE, “R2-104662: MTC simulation results with specific solutions,”
3GPP TSG RAN WG2 Meeting 71, 2011.
[76] P. Phunchongharn, E. Hossain, and D. I. Kim, “Resource allocation
for device-to-device communications underlaying LTE-Advanced net-
works,” IEEE Wireless Commun., vol. 20, no. 4, pp. 91–100, Aug. 2013.
[77] Y. Cao, T. Jiang, X. Chen, and J. Zhang, “Social-aware video
multicast based on device-to-device communications,” IEEE Trans.
Mobile Comput., vol. 15, no. 6, pp. 1528–1539, Jun. 2016.
[78] Y. Cao, T. Jiang, and C. Wang, “Cooperative device-to-device commu-
nications in cellular networks,” IEEE Wireless Commun., vol. 22, no. 3,
pp. 124–129, Jun. 2015.
[79] N. Saquib, E. Hossain, and D. I. Kim, “Fractional frequency reuse
for interference management in LTE-Advanced hetnets,” IEEE Wireless
Commun., vol. 20, no. 2, pp. 113–122, Apr. 2013.
[80] N. Saquib, E. Hossain, L. B. Le, and D. I. Kim, “Interference man-
agement in OFDMA femtocell networks: Issues and approaches,” IEEE
Wireless Commun., vol. 19, no. 3, pp. 86–95, Jun. 2012.
[81] S. Bayat, Y. Li, Z. Han, M. Dohler, and B. Vucetic, “Distributed
data aggregation in machine-to-machine communication networks
based on coalitional game,” in Proc. IEEE Wireless Commun. Netw.
Conf. (WCNC), Istanbul, Turkey, Apr. 2014, pp. 2026–2031.
[82] Y. Cao, T. Jiang, and C. Wang, “Optimal radio resource allocation for
mobile task offloading in cellular networks,” IEEE Netw., vol. 28, no. 5,
pp. 68–73, Sep./Oct. 2014.
[83] A. Aijaz, M. Tshangini, M. R. Nakhai, X. Chu, and A.-H. Aghvami,
“Energy-efficient uplink resource allocation in LTE networks with
M2M/H2H co-existence under statistical QoS guarantees,” IEEE Trans.
Commun., vol. 62, no. 7, pp. 2353–2365, Jul. 2014.
[84] Y. Cao, T. Jiang, C. Wang, and L. Zhang, “CRAC: Cognitive radio
assisted cooperation for downlink transmissions in OFDMA-based
cellular networks,” IEEE J. Sel. Areas Commun., vol. 30, no. 9,
pp. 1614–1622, Oct. 2012.
Yang Cao (S’09–M’14), photograph and biography not available at the time
of publication.
Tao Jiang (M’06–SM’10), photograph and biography not available at the time
of publication.
Zhu Han (S’01–M’03–SM’09–F’14), photograph and biography not available
at the time of publication.