An Overview of DLMS/COSEM

and G3-PLC for Smart metering

Rahma Gloulou Med Yassine Ghzaiel Fares Cherni

Abstract - In This article we describe the Application layer and a product to work with other systems or products without
physical layer for a smart meter to have a proper communication special effort on the part of the customer. Specifically, it
over smart grid. The Application layer describes the
points to an environment where any meter can be read by any
DLMS/COSEM protocol based on IEC specifications. The
physical layer describes the OFDM based PLC technique using system, and which includes no special involvement of
G3 PLC. vendors. There is a need to define universal data protocols so
that all the partners in the communication chain can reliably
and successfully receive and send messages, and to
I. INTRODUCTION unambiguously excerpt and interpret useful data. There is a
need for communication standards. DLMS/COSEM [1] is the
A. Smart Grid common language such that the partners can understand each
other.
Dependence on electricity is now increasing in a high
rate. So, production is increasing. But the power grid
delivering the power is now aging and under stress. So, the II. SMART METER MODEL
power grid needed to be upgraded. “Smart grid” is a term that
refers to the modernization of the electric system through the DLMS/COSEM is promoted by DLMS UA, through
integration of new information-age technologies, new the two essential documents Green Book [1] and Blue Book
strategic public policies, and allows for new uses of the [1]. All meter data had to be mapped to COSEM objects
electric grid, both in operations and through new customer according to the Blue Book, and all data messaging according
side applications. Modern Grid must transfer energy from to the Green Book.
where it is made to where it is necessary. For that it must
Every parameter in a meter is represented by a
satisfy two- w a y communications. So, every device
COSEM object. The Meter functionality is modelled using
connected to it can listen and talk to every other one. Smart
objects. Examples of some objects are: Real time clock, date
Grid includes smart meters, smart sensors, and smart controls.
and time, Cumulative energy, Frequency, Line voltage, Line
Its benefits include higher reliability, higher security, lower
current, billing counter, Timestamp values etc. The COSEM
cost, and increase dependence on green power sources.
objects [2] are instances of a class (interface class). All objects
of the same class have the same attributes and methods.
B. smart metering
Within a specific class, the common characteristics (attributes
The conventional functionality of Automatic Meter and methods) are described once for all objects. The
Reading (AMR) is changing in the direction of smart multi- instantiations of an interface class are called COSEM interface
metering or multi-functional Advanced Metering objects. These are shown in fig 1
Infrastructure (AMI). The term AMI refers to systems that
measure, collect, and analyze energy usage data, from
advanced devices such as electricity meters, gas, heat and/or
water meters, through various communication media on
request or on a predefined schedule. Advanced functionalities
provided by an AMI system, like customers relation
management, demand side management, local power quality
monitoring, energy supply / billing, load management,
network condition and health monitoring, meter condition and
health monitoring, energy services.
.
C. The Need for DLMS/COSEM

For the implementation of an efficient AMI system Fig.1. Interface classes and objects
the devices have to be interoperable. This means the system or

Authorized licensed use limited to: UNIVERSITE DE SOUSSE. Downloaded on November 28,2022 at 10:30:38 UTC from IEEE Xplore.
Restrictions apply.
TABLE.1. HDLC FRAME FORMAT
Flag Frame Destination Source Control HCS Info FCS flag
Format Address Address

For example, meter parameters like active energy in connection oriented, HDLC based communication protocol,
first tariff can be instances of the register or extended register the lower HDLC address is used for addressing the physical
class, whereas maximum demand can be of the class register, device and the upper HDLC address is used for addressing the
extended register, or profile generic. Register objects are various logical devices within the same physical device.
modeled through logical name, value and scaler-unit Application associations are established between a Client and
attributes, while extended register object attributes: status and a Logical Device in the Server. The associations are identified
capture time. For e.g., suppose the vendor wants to know how by the pair of HDLC addresses (Client HDLC address –
much energy is consumed by a house. then first the vendor Server lower HDLC address-Server upper HDLC address).
must identify that particular house. After the connection The client address is always 1 byte and server address can be
establishment is done, assume that vendor must read 1, 2, or 4 bytes. [6][5].
accumulated energy out of the meter.
The objects that are supported in a meter is known
For this purpose, the DLMS provides GET requests to the client via the association objects. There are Mandatory
to acquire all needed object attribute data. First, all the objects objects [2] that need to be implemented in every meter.
were defined, enumerated and OBIS [4] code was appended to The Mandatory elements allow us to get the necessary
each object. OBIS uses six value groups. For example, an information about the content of a physical device. Each
abject having an OBIS code= “1.1.1.8.2.255”, represents total physical device has a management logical device, at
positive active energy in second tariff. This object is of class address 1(0x0001). A management logical device contains a
Register, with the OBIS code as its first attribute or logical list of all the available logical devices in the physical
name. The second attribute-value is expressed with unsigned device. This list is the second attribute of the object of class
32 data type while the third attribute scaler unit {-3, 30} SAP Assignment with the predefined name 0.0.41.0.0.255.
means the value should be scaled down by 1000 and the unit Each logical device contains a list of all its available objects.
is the data name is provided in a standard way, such that This list is the second attribute of the object of class
everybody (meters and energy providers) can understand what Association, with the predefined name 0.0.40.0.0.255(current
is carried. Secondly, routines for packing object data into association object) [2][7]. The current association object’s
DLMS messages is done using 3-layer HDLC protocol. We object list attribute keeps track of all the objects implemented
assume that our meter implements the 3-layer, connection- in the meter. After a client establishes the application
oriented, HDLC based profile. HDLC frame format type 3[5] association connection with a server it can send a get
is used to transmit packets through wires or optically. The request to acquire the object list. Finally, before the actual
HDLC frame format is shown in Table 1. metering information can be exchanged an association has to
be build up, which is initiated by the client. The DLMS client
The Server is modeled as a physical device can then access the interface object model inside the server.
containing one or more logical devices, each logical device Once an association exists the DLMS server is also
supporting one or more application associations with the able to send notifications to the client without an explicit
Client(s) and two or more COSEM objects. In the three-layer, request.

Fig.2. Smart meter coupled to power line

Authorized licensed use limited to: UNIVERSITE DE SOUSSE. Downloaded on November 28,2022 at 10:30:38 UTC from IEEE Xplore.
Restrictions apply.
 band width required will be reduced to almost 50% than
III. POWER LINE COMMUNICATION (PLC)
Frequency Division Multiplexing. Even though the carriers
Fig.2 shows a smart meter coupled to power line.
Every building and home are already equipped with the power
line and all devices are connected to power line and connected
to the power grid. So, building a home network using the
existing AC electrical wiring will have advantages. The
regulations are described in CENELEC standard EN 50065
entitled “Signaling on low-voltage electrical installations in
the frequency range 3 kHz to 148.5 kHz”, the allowed
frequency band and output voltage for communications over
the Residential Power Circuit (RPC) are as shown in Table
2[8].

TABLE 2 CENELAC A band voltages and frequency

Band Frequency Amplitude Usage
range
A 9-95 10 Power supply
utilities
B 95-125 1.2 For private
usage
C 125-140 1.2 For private
usage Fig.3. OFDM flow diagram

are made to overlap the message signal can be recovered by

While considering PLC various factors like Access selecting subcarriers as orthogonal sinusoids which means at
Impedance (AI), Reflections, Power line noises has to be the maximum of one carrier, all other carriers will be zero
considered. AI shunts the power amplifier and its coupling Fig.3. shows the steps for OFDM. CTOTAL the total
circuit and hence it determines the injected signal power. number of bits to be modulated. Digital modulation can be
Every load or appliance connected to the mains contributes to PSK or QAM. NOFDM will vary depend on modulation
this access impedance. But these loads are not fixed but technique. IDFT converts the frequency domain samples to
continuously changing in random patterns with respect to time time domain samples. Cyclic prefix is added to avoid the Inter
and the active loads connected. Along its path form Symbol Interference (ISI). Number of samples for cyclic
prefix should be greater than or equal to channel impulse
Transmitter to Receiver, the signal encounters several response. Cyclic prefix makes the output circular convolution
impedance discontinuities. It may travel across different cable between input and channel response, so at the receiver input
types and may encounter branching or even couple across can be separated by dividing with the channel response . By
phases. Such discontinuities introduce impedance mismatches choosing OFDM with cyclic prefix as modulation technique
and dramatically degrade the signal strength by dividing its ISI due to multipath reflections and frequency dependent
power among branches. A transfer function of PLC channel channel attenuation can be eliminated .
suggested by Zimmerman, Dostert is given in equation 1
[9]. A. OFDM frequency spectrum for G3 PLC
Every connected load running from the mains injects
The PLC G3 is one of the standards defined for the
physical layer for smart grid, which is based on OFDM. It
supports the portion between 35.9 kHz to 90.6 kHz of the
(1) CELENEC A band. There is potential to use this standard to
support communication in frequencies up to 180 kHz. So,
the sampling frequency at the transmitter and receiver is
periodic and aperiodic load specific noise signature which selected to be 0.4MHz in order to provide some margin
contributes to power line noise [10]. The issues can be above the Nyquist frequency for signal filtering in the
recovered by using various modulation techniques and transmitter and at the receiver. According to various field
channel encoding techniques. trials best result is achieved when, an IFFT size of 256 is
used. This results in a frequency spacing between the
IV. OFDM FOR PLC OFDM carriers equal to 0.64ms. After considering various
imperfections, the number of carriers was chosen to 36.
OFDM is a multicarrier modulation where the
subcarriers are made to overlap with each other , therefore

Authorized licensed use limited to: UNIVERSITE DE SOUSSE. Downloaded on November 28,2022 at 10:30:38 UTC from IEEE Xplore.
Restrictions apply.
 TABLE 3 Sub Carrier Frequency
No of carriers First carrier Last carrier
36 35.938 90.625

The carriers were chosen to be from 23 to 58 which results

in frequency from 35.938 to 90.625. Last thirty-time domain
samples are added as cyclic prefix [14][15].

REFERENCES
[1] Companion Specification for Energy Metering.
COSEM Identification System and Interface Classes-
Blue book 8th edition. DLMS/COSEM Architecture
Fig.4. G3 PLC carrier modulation and Protocols-Green book 7th edition.

[2] INTERNATIONAL STANDARD IEC 62056-62

OFDM spectrum simulated as per G3 PLC is shown in Interface classes. Electricity metering- Data
Fig.5 exchange for meter reading, tariff and load control.
[3] INTERNATIONAL STANDARD IEC 62056-53
COSEM application layer. Electricity metering -Data
exchange for meter reading, tariff and load control
[4] INTERNATIONAL STANDARD IEC 62056-61
Object identification system (OBIS).Electricity
metering -Data exchange for meter reading, tariff and
load control.
[5] INTERNATIONAL STANDARD IEC 62056-46
[6] “Practical Issues of Power Line Communication for
Automatic Meter Reading Systems” Yasser Fathi,
MIEEE, Tamer A. Kawady, MIEEE, Ahmed Husein,
and Mohamed El-Geziry
[7] John. G. Proakis, Masoud Salehi. “Communications
Systems Engineering”, Prentice Hall, 1994
[8] “Orthogonal Frequency Division Multiplexing
Fig.5 OFDM spectrum simulated as per G3 PLC OFDM”, fred harris, San Diego State University.
[9] “G3-PLC Specification for Powerline
V. Conclusion
Communication:Overview, System Simulation and
Field Trial Results” Kaveh Razazian, Maher Umari,
Amir Kamalizad, Victor Loginov, Michael Navid
After covering a summary on smart metering systems and Signal Processing and Conversion Business Unit
different wireless and wireline technologies employed in various Maxim Integrated Products Irvine,
sections of the AMI network, it was discussed that the PLC is the
preferred technology for smart metering applications in dense [10] PLC G3 physical layer specification, ERDF
urban areas. Among all PLC-based standards deployed in AMI
networks, the G3-PLC is the most widely used standard and has
the fastest growth. This is a result of various communication
techniques considered in the G3-PLC design, which has been
made it a suitable choice for a variety of applications. However,
the increase in G3-PLC penetration into more devices and
applications has brought up new challenges and research
opportunities to overcome its existing limitations. Therefore, this
paper first provided a holistic view of G3-PLC in smart metering
systems, including its interaction with its data exchange standard,
the DLMS/COSEM. This was followed by the still-existing
challenges in different layers of the G3-PLC and recent advances
that have been made to alleviate such challenges. Finally, it was
concluded that the future AMI system would benefit from a
hybrid of wireless and PLC-based technologies to efficiently
exploit rare spectral resources.

Authorized licensed use limited to: UNIVERSITE DE SOUSSE. Downloaded on November 28,2022 at 10:30:38 UTC from IEEE Xplore.
Restrictions apply.
Authorized licensed use limited to: UNIVERSITE DE SOUSSE. Downloaded on November 28,2022 at 10:30:38 UTC from IEEE Xplore.
Restrictions apply.
After covering a summary on smart metering systems and different wireless and wireline technologies employed in various sections of
the AMI network, it was discussed that the PLC is the preferred technology for smart metering applications in dense urban areas.
Among all PLC-based standards deployed in AMI networks, the G3-PLC is the most widely used standard and has the fastest growth.
This is a result of various communication techniques considered in the G3-PLC design, which has been made it a suitable choice for a
variety of applications. However, the increase in G3-PLC penetration into more devices and applications has brought up new challenges
and research opportunities to overcome its existing limitations. Therefore, this paper first provided a holistic view of G3-PLC in smart
metering systems, including its interaction with its data exchange standard, the DLMS/COSEM. This was followed by the still-existing
challenges in different layers of the G3-PLC and recent advances that have been made to alleviate such challenges. Finally, it was
concluded that the future AMI system would benefit from a hybrid of wireless and PLC-based technologies to efficiently exploit rare
spectral resources.

Authorized licensed use limited to: UNIVERSITE DE SOUSSE. Downloaded on November 28,2022 at 10:30:38 UTC from IEEE Xplore.
Restrictions apply.