Upload
others
View
3
Download
1
Embed Size (px)
Citation preview
Communications and Networking
for Smart Grid Systems
Dusit Niyato
Nanyang Technological University (NTU), Singapore
Rose Qingyang Hu
Utah State University
Ekram Hossain
University of Manitoba, MB, Canada
Yi Qian
University of Nebraska-Lincoln
1
IEEE GLOBECOM 2011, Houston, USA December 9, 2011
Tutorial Outline
1. Introduction, Background, and Overview of Smart Grid Systems
2. Data Communication Requirements in Smart Grid
3. Communication Architectures, Area Networks, and Components for Smart Grid
4. Data Communications and Networking in Smart Grid
5. Cyber Security and Privacy in Smart Grid Communications Infrastructure
6. Field Trials and Case Studies
7. Open Issues and Future Research Directions
8. Summary
2 IEEE GLOBECOM'11
Introduction • What is smart grid?
– Smart grids – add communication capabilities and intelligence to
traditional grids
• What enables smart grids?
– Intelligent sensors and actuators
– Extended data management system
– Expanded two way communications between power generation,
distribution, and customers
– Network security
– etc.
IEEE GLOBECOM'11
3
Smart Grid: The “Energy Internet”
2-way flow of electricity and information
Standards Provide a Critical Foundation 4
Motivations
Smart Grid Enables:
• Higher Penetration of
Renewables
• Smart Charging of
Electric Vehicles
• Consumers to Control
Energy Bills
• Efficient Grid
Operations &
Reduced Losses
• Reduced Distribution
Outages
• Improved System
Reliability & Security
IEEE GLOBECOM'11
6
Primary objectives
• National integration
• Self healing and adaptive –Improve distribution and
transmission system operation
• Allow customers freedom to purchase power based on
dynamic pricing
• Improved quality of power-less wastage
• Integration of large variety of generation options
IEEE GLOBECOM'11
7
Economic and social benefits
• Provide Customer Benefits
• Reduce Peak Demand
• Increase Energy Conservation & Efficiency
• Reduce Operating Expenses
• Increase Utility Worker Safety
• Improve Grid Resiliency and Reliability
• Reduce Greenhouse Gas Emissions
• Promote Energy Independence
• Promote Economic Growth & Productivity
IEEE GLOBECOM'11
8
Distributed Generation
• Hybrid Energy Resource
– Fossil-Fuel
– Wind
– Solar
– Bio-Mass
– Batteries
– Capacitors
– Flywheel
– Etc.
IEEE GLOBECOM'11
10
Smart Metering
(b) Microsoft Hohm (a) Google PowerMeter
• Automatic Metering
– Automatic Meter Reading (AMR)
– Automated Metering Management (AMM)
– Advanced Metering Infrastructure (AMI)
Example smart metering systems:
IEEE GLOBECOM'11
11
Intelligent electronic devices (IEDs)
• Protection relay
• Auxiliary relay
• Cheap contractors
• Remote terminal units
• Circuit breaker monitor
• Revenue meters
• Solar flare detectors
• Power quality monitors
• Phasor measurement units
• Communication processors
• Communication alarm
• Etc.
GE CFD Intel 4004
12
Monitoring and Controlling
– Supervisory Control And Data Acquisition (SCADA)
– Energy management system (EMS)
– Information and Communications Technology (ICT)
IEEE GLOBECOM'11
13
Telecontrol
• Different protocols for different operations
– Proprietary protocols (more than 100)
– Standards
• SCADA
• Modbus
• DNP
• IEC61850
IEEE GLOBECOM'11
14
SCADA Protocols
• Siemens quad 4 meter
• CONITEL 2000
• CONITEL 2100
• CONITEL 3000
• CONITEL 300
• HARRIS 5000
• HARRIS 5600
• HARRIS 6000
• UCA 2.0 or MMS
• PG & E 2179
• MODBUS
• DNP3
• ICCP
• IEC 61850
IEEE GLOBECOM'11
15
General Protocols
• MODBUS -Primitive without security and not very
extensible
• DNP3 –Advanced SCADA protocol
• DNP1 and 2 are proprietary protocols
• IEC 61850 the most used protocol for new implementations
• ICCP
IEEE GLOBECOM'11
16
Groups working on smart grids
• UCA International user group www.ucaiug.org
• International electrochemical commission www.iec.ch
• Electric power research institute www.epri.com
• Intelligrid consortium and architecture www.intelligrid.epri.com
• IEEE smart grid www.smartgrid.ieee.org
• NIST csrc.nist.gov
IEEE GLOBECOM'11
17
Communication Media
• Urge for new FCC allocation for smart grids
• PLC –Power line carriers
• Ethernet
• WLAN
• ZigBee
• Bluetooth
• Optical fiber
• Microwave
• Etc.
IEEE GLOBECOM'11
20
Data Communications in Smart Grid Optimal Network(s)
• Broadcast data (Demand Response, price signals, emergency events, etc.)
– Low volume, infrequent
– Can use currently available communication infrastructure (cellular, broadband, WiFi, Pager) with standard internet security measures
• Real-time Consumption Data (high volume, frequent)
– Useful primarily for real-time control & usage information to consumer
– We favor meter premises where displays & controllers can locally act upon this data along with pricing information
• Minimizes risk (privacy & network stability) and maximizes benefit from real time info.
• Raw Billing Data (reading when price changes)
– Utility operations
• Aggregate Data
– Comparison over time & among neighbors, best practices, consumption pattern recognition, suggest corrective actions, etc.
• Utility or third party cloud-based applications operating on anonymous summary data
• Little risk for privacy or network stability in case of breach of security
• Can use standard internet communication with standard security measures
• T&D: relatively few points (substations); mission-critical, but already connected
Broadband
Cellular
WiFi, etc.
Direct meter
to HAN
AMI
Internet portal
Existing
connectivity
IEEE GLOBECOM'11
21
Match Info To The Communications Medium
Information
category
Smart Grid Signals Detailed Consumption Data
Examples ToU pricing, critical peak pricing,
reliability, carbon content, etc.
Periodic meter readings (e.g., once
a minute)
Location of
information
Utility servers connected to Internet Embedded meter hardware
Evolution potential High, as new applications arise (e.g.,
PHEVs, micro-grids)
Very low
Optimal approach
AMI-centric approach
Communication
medium
General telecom infrastructure
(Internet): broadband, cellular,
municipal WiFi, etc.
Specialized embedded hardware
(short-range radio, power-line
carrier, etc.)
IEEE GLOBECOM'11
22
Requirements
• Latency
• Bandwidth
• Interoperability
• Scalability
• Security
• Standardization
IEEE GLOBECOM'11
25
Latency
• The real-time operational data communications in smart grid include online
sensor/meter reading and power system control signals.
• The communication is characterized by the fact that most of interactions must
take place in real time, with hard time bound.
• The communication requirements define the design of the technical solutions.
• For real-time sensing/metering purposes, reading messages should be
transmitted within a very short time frame.
– For instance, the maximum allowed time is in the range of 12-20 ms, depending on
the type of protection scheme which origins from the fact that the disconnection of
fault current should within approximately 100 ms.
• Power System Control signals mainly include supervisory control of the power
process on secondary or higher levels. These systems are of the kind
SCADA/EMS.
– Measured values must not be older than 15 seconds, when arriving at the control
center. Breaking information shall arrive no later than 2 seconds after the emergency
event has occurred
IEEE GLOBECOM'11
26
Bandwidth
• As more and more interconnected intelligent elements are added to the
electricity network with the evolution of the smart grid, the
communication infrastructure should be able to transport more and more
messages simultaneously without severe effect on latency.
• The network bandwidth must increase faster than the demand of these
interconnected intelligent elements in the network.
• An Example: (A. Aggarwal, S. Kunta, P. K. Verma, “A proposed communications
infrastructure for the smart grid,” in Innovative Smart Grid Technologies (ISGT), 2010,
pp. 1-5.) – Model the communication bandwidth requirements for a moderate size electricity distribution system. In this
model, a distribution substation is connected to 10,000 feeders and each feeder connects to 10 customers.
– Assuming that every electric meter generates a message every second to the distribution substation, the total is
100,000 messages per second. The feeders themselves will generate messages to each other and to the distribution
substation.
– The authors in this paper modeled the messages in the smart grid arriving at servers located at the control center
as M/M/1 traffic. Then, the transmission line bandwidth is evaluated over 100 Mbps through the M/M/1 queuing
model. It can be observed that this situation results in a very poor bandwidth utilization of the transmission
facilities as well.
– Unfortunately, a higher level of utilization will not permit meeting the assumed latency constraint.
IEEE GLOBECOM'11
27
Interoperability
• The ability of 2 or more networks, systems, devices, applications, or
components to communicate & operate together effectively, securely, &
without significant user intervention
– Communication requires agreement on a physical interface &
communication protocols
– Exchanging meaningful & actionable information requires common
definitions of terms & agreed upon responses
– Consistent performance requires standards for the reliability, integrity,
and security of communications
– Interoperability may include:
• “Plug and play”: connect them & they work together
• Interchangeability: Ability to readily substitute components
IEEE GLOBECOM'11
28
Standards • EISA 2007 Directs National Institute of Standards & Technology (NIST) to:
– Coordinate the development of model standards for interoperability of smart grid devices and systems
• Create flexible, uniform, and technology neutral standards
• Enable traditional resources, distributed resources, renewables, storage, efficiency, and demand
response to contribute to an efficient, reliable grid
• EISA Directs FERC, when sufficient consensus, to:
– Adopt standards necessary to insure smart-grid functionality and interoperability in the interstate transmission of electric power, and regional and wholesale electricity markets
– EISA did not expand FERC‟s Federal Power Act authority to enforce standards
• State Commissions:
– May adopt standards by regulation, separately or in parallel with FERC
– May consider standards when approving utility investments
• Considerations for Regulators:
– Ensuring interoperability & security, without impeding innovation
– Consistent action will influence the vendor community
– Vendors often will follow standards that are not legally mandated
– SGIP standards reflect efforts to build broad stakeholder consensus
IEEE GLOBECOM'11
31
Standardization (cont‟d)
• IEEE
– IEEE P2030
• Power Engineering Technology
• Information Technology
• Communications Technology
• IEC – IEC 61968 - Distribution Management
– IEC 61970 - Common Information Model
– IEC 60870 - Inter-control Center Communication Protocol
– IEC 62210 - Data and Communication Security
– IEC 62357 - Reference Architecture
– IEC 61850 - Standard for Design of Substation Automation
• IEC 61850-7-420 - Integration of Distributed Energy Resources
• IEC 61850-7-410 - Integration of Hydro Resources
– IEC 61400 - Integration of Wind Farms to Utility Communication Network
– IEC 62056 - Communication
IEEE GLOBECOM'11
32
Security
• DISA Security Technical Implementation Guides (STIGs)
• FIPS 201
• North American Electrical Reliability Corporation-Critical
Infrastructure Protection (NERC CIP)
• National Infrastructure Protection Plan (NIPP)
• IEEE 1402
• International Society of Automation(ISA)
• ISO 17799
• NIST GWAC
– DEWGs
• Home-to-Grid (H2G)
• Building-to-Grid (B2G)
• Industrial-to-Grid (I2G)
• Transmission and Distribution (T&D)
• Business and Policy (B&P)
IEEE GLOBECOM'11
33
Summary
Application Security Bandwidth Reliability Coverge Latency Back-up Power
Advanced Metering Infrastructure High 14-100 kbps per node 99.0-99.99% 20-100 % 2000 ms 0-4 hours
AMI Network Management High 56-100 kbps 99.00% 20-100% 1000-2000 ms 0-4 hours
Automated Feeder Switching High 9.6-56 kbps 99.0-99.99% 20-100% 300-2000 ms 8-24 hours
Capacitor Bank Control Medium 9.6-100 kbps 96.0-99.00% 20-90% 500-2000 ms 0 hours
Charging Plug-In Electric Vehicles Medium 9.6-56 kbps 99.0-99.90% 20-100% 2000 ms - 5 min. 0 hours
Demand Response High 56 kbps 99.00% 100% 2000 ms 0 hours
Direct Load Control High 14-100 kbps per node 99.0-99.99% 20-100 % 2000 ms 0-4 hours
Distributed Generation High 9.6-56 kbps 99.0-99.99% 90-100% 300-2000 ms 0-1 hour
Distribution Asset Management High 56 kbps 99.00% 100% 2000 ms 0 hours
Emergency Response Medium 45-250 kbps 99.99% 95% 500 ms 72 hours
Fault Current Indicator Medium 9.6 kbps 99.00-99.999% 20-90% 500-2000 ms 0 hours
In-home Displays High 9.6-56 kbps 99.0-99.99% 20-100% 300 -2000 ms 0-1 hour
Meter Data Management High 56 kbps 99.00% 100% 2000 ms 0 hours
Network Protection Monitoring Medium - High 56-100 kbps 99.00-99.999% 100% 2000-5000 ms 0 hours
Outage Management High 56 kbps 99.00% 100% 2000 ms 0 hours
Price Signaling Medium 9.6-56 kbps 99.0-99.90% 20-100% 2000 ms - 5 min. 0 hours
Real-time Pricing High 14-100 kbps per node 99.0-99.99% 20-100 % 2000 ms 0-4 hours
Remote Connect/Disconnect High 56-100 kbps 99.00% 20-100 % 2000-5000 ms 0 hours
Routine Dispatch Medium 9.6-64 kbps 99.99% 95% 500 ms 72 hours
Transformer Monitoring Medium 56 kbps 99.00-99.999% 20-90% 500-2000 ms 0 hours
Voltage and Current Monitoring Medium 56-100 kbps 99.00-99.999% 100% 2000-5000 ms 0 hours
Workforce Automation Medium 256-300 kbps 99.90% 90% 500 ms 8 hours
CURRENT FUNCTIONAL REQUIREMENTS
National Broadband Plan: RFI Communications Requirements
Comments of Utilities Telecom Council, July 12, 2010
IEEE GLOBECOM'11
34
Challenges for Smart Grid Communication Infrastructure
• Complexity
• Efficiency
• Reliability
• Security
IEEE GLOBECOM'11
35
Complexity
• Need to support multi-physics approach
• Need to support multidisciplinary approach
• Need to support dynamic and reconfigurable model
level definition
• Need to provide visualization to support system
analysis
• Need to provide support for uncertainty propagation
IEEE GLOBECOM'11
36
Efficiency
• Better Telemetry
• Faster Controls
• More Robust Controls
• Embedded Intelligent Devices Communication
• Integrated And Secure Communications
• Enhanced Computing Capabilities
• Internet Technology
37
Reliability
• Renewable Resources
• Demand Response
• Load Management
• Storage Devices
IEEE GLOBECOM'11
38
Security
• Information security domains
– Public, supplier, maintainer domain
– Power plant domain
– Substation domain
– Telecommunication domain
– Real-time operation domain
– Corporate IT domain
• SCADA
– De-coupling between operational SCADA/EMS and admin IT
– Governmental coordination on SCADA security
• Threats to
– AMI (similar to WSN)
– SCADA
IEEE GLOBECOM'11
39
Communication Architectures
• Communication Architecture and Model for Distribution
Network
• Home-Area Networks (HANs)
• Neighborhood-Area Networks (NANs)
• Wide-Area Networks (WANs)
• Sensor and Actuator Networks (SANETs)
40 IEEE GLOBECOM'11
Communication Architectures
Communication Architecture and Model for Distribution
Network
41 IEEE GLOBECOM'11
DAU/
NAN GW
Generation
MDMS
Tansmission and Distribution Customer
Premises
Transmission
Substation
Solar
EnergyWind
Turbines
Smart Meter/
HAN GW
Communication Core Network
E.g., TCP/IP Network, WiMax, Cellular (GSM or CDMA), Ethernet
Control CenterControl Center Control Center
HAN
NANWAN
Ele
ctri
cal
Infr
ast
ruct
ure
Co
mm
un
ica
tio
n
Infr
ast
ruct
ure
Customer Networks
E.g., ZigBee, WiFi,
PLC
Legends: DAU=Data Aggregator Unit, MDMS=Meter Data Management System, HAN=Home Area Network,
NAN=Neighborhood Area Network, WAN=Wide Area Network , GW = Gateway
Electric FlowInformation Flow
Distribution
Substation
Distribution
Feeder
Sensor Network
Advanced Metering Infrastructure (AMI)
Last Mile Connection
Transmission
Feeder
Communication Architectures
Communication Architecture and Model for
Distribution Network
• Smart grid follows the same electrical architecture
• Electricity is delivered from the generation to consumers
through transmission and distribution substations
• Transmission substation delivers electricity from power
generation plant over a high voltage transmission line (over
230kV) to the distribution substation
• Distribution substation converts the electric power to
medium voltage level
• Distribution feeder then converts the medium voltage to
lower level for distributing to the consumer‟s end
42 IEEE GLOBECOM'11
Communication Architectures
Customer Premise and Customer Network
43 IEEE GLOBECOM'11
NAN1
HAN3HAN1HAN2
Smart Meter
(HAN Gateway)
DAU/NAN GW
Bluetooth/ZigBee/
WiFi
BACnet, KNX,
PLC protocol
Smart Devices
(e.g., AC) With
Sensors
Display
HAN3HAN1
HAN2
NAN2
DAU/NAN GW
MDMS
Control
Center
HAN3
Control
Center
MDMS
Communication Architectures
Customer Premise and Customer Network
• Data aggregator unit (DAU) also referred to as NAN GW
acts as a data sink to collect and relay the information from
the consumer side to meter data management system
(MDMS)
• MDMS will provide storage, management, and processing
of meter data for proper usage by other power system
applications and services
44 IEEE GLOBECOM'11
Communication Architectures
Home-Area Networks (HANs)
• HAN (sometimes referred to as Premise Area Network
(PAN) or a Building Area Network (BAN)) is the smallest
subsystem in the hierarchical chain of smart grid
• HAN provides a dedicated demand side management
(DSM), including energy efficiency management, and
demand response by proactive involvement of power users
and consumers
• HAN consists of smart meter, smart devices with sensors
and actuators, and in-home display for energy management
system (EMS)
– EMS will provide means of reducing energy consumption by
monitoring and controlling different electrical appliances
45 IEEE GLOBECOM'11
Home-Area Networks (HANs)
General Structure
46 IEEE GLOBECOM'11
Electric supply from
Transmission
HAN
Gateway
In-Home
Display
DAU/NAN
Sensors
Light
Temperature
Voltage
Wired/Wireless
Connection
(e.g., Zigbee,
BACnet
Smart
Devices
Smart
Devices
Smart
Devices
Actuators
ActuatorsActuators
(Smart Meter
or
Dedicated
in-home
Gateway)
Enabling Communications Technologies
• Short Range Wireless Technologies
– Wi-Fi, Bluetooth, ZigBee, Z-Wave
• Z-Wave:
– Proprietary wireless standard designed for home control automation,
specifically to remote control applications in residential homes
– Z-Wave was originally developed by Zensys A/S and is being
marketed by Z-Wave Alliance
– Z-Wave wireless protocol provides reliable and low-latency
communication of small data packets within HANs
– Z-Wave also uses a mesh networking approach with source routing
47 IEEE GLOBECOM'11
Home-Area Networks (HANs)
Enabling Communications Technologies
• Z-Wave:
– Bandwidth: 9,600 bit/s or 40 kbit/s
– Modulation: GFSK
– Range: Approximately 100 feet (or 30 meters)
– Frequency band: The Z-Wave Radio uses the 900 MHz ISM band
• 908.42 MHz (United States)
• 868.42 MHz (Europe)
• 919.82 MHz (Hong Kong)
• 921.42 MHz (Australia/New Zealand)
48 IEEE GLOBECOM'11
Home-Area Networks (HANs)
Enabling Communications Technologies: Wireless Technologies
49 IEEE GLOBECOM'11
Key Criteria WiFi Bluetooth ZigBee Z-Wave
Feature - Designed for providing wireless
connection for accessing Internet
and is direct replacement to
traditional Ethernet
network
-Designed for consumer electronics to
provide short-range wireless
communication
to connect a wide range of devices
easily and quickly
- Designed specifically for industrial and
home automation for connecting
sensors,
monitors and control devices
- Designed for home automation,
specifically to remote control
applications in residential home such as
light, entertainment systems, etc
Frequency Band - 2.4/5 GHz - 2.4 GHz - 2.4 GHz, 915MHz and 868MHz - 900 MHz
Standards - International Standard (IEEE
802.11 a/b/g/n)
- Open Standard
- International Standard (IEEE 802.15.1)
- Open Standard
- International Standard (IEEE 802.15.4)
- Open Standard
- Proprietary Standard (Z-Wave
Alliance and Zensys)
- Closed standard
Speed - 54 Mbps ( 802.11. b/g)
- 150 Mbps (802.11 n)
- 2.1 Mbps (V 2.0)
- 20 Mbps (V 3.0, recently released)
- 250 Kbps - 9600 bits/s
Range - 70m (indoor) to 250m (outdoor) - 10m - 70m (indoor) to 400m (outdoor) - 30m (indoor) to 100m (outdoor)
Power Consumption - High - Lower than WiFi - Lower than WiFi and Bluetooth - Almost same as ZigBee
Maximum Nodes - 2007 - 8 - > 64000 - 232
Home-Area Networks (HANs)
Enabling Communications Technologies: Wireless Technologies
50 IEEE GLOBECOM'11
Key Criteria WiFi Bluetooth ZigBee Z-Wave
Security - WEP (Wired Equivalent privacy)
-WPA (Wi-Fi Protected access)
- WPA2
- E0 stream cipher
- More Secure than WiFi
- 128 AES (Advanced
Encryption Standard )
keys
- 3 DES(Triple Data Encryption
Standard)
Strength - Easy to deploy,
equipment costs
dropping rapidly
- Supports mesh topology
- Most popular protocol for transferring
data and wireless alternative to RS-
232 data cables
- Supports ring topology
- Low power requirements and
implementation costs
- Particularly designed for use in
industrial and home automation or
security applications
- Scalable and flexible
- Supports mesh topology
- Low power, low latency, and low cost
- Less interference due to use of sub-
GHz frequency
- Higher propagation range 2.5 times
the 2.4 GHz signal
- Supports mesh topology
Concern - High power consumption
- Higher data latency
- Additional security layer should be
implied to use WiFi within HAN
-Lack of proper installation in consumer
portal context such as fire alarm, security
sensors, etc
- Does not support mesh networking
- Limited range and low data rates
- Interference due to overlapping with
WiFi standard
- Low data rates
- Requires to add devices into network
manually
- Slightly higher installation cost than
ZigBee
- Offers less flexibility due to close
nature
Home-Area Networks (HANs)
Home-Area Networks (HANs)
Enabling Communications Technologies: Wired Technologies
51 IEEE GLOBECOM'11
Key
Criteria
X10 HomePlug GP BACnet KNX
Feature - Simple and popular protocol designed for
providing simple automation functionality such as
on and off.
-Designed specifically for smart grid to
provide lower power consumption
- A data communication protocol that
attempts to unifies all the proprietary
communication protocol into single
communication language
- Global standard protocol designed
basically for home automation and
control
Wireless
Support
- Yes
- 310 MHz U.S. 433 MHz European
- Yes
- Recent ZigBee/HomePlug initiatives
- - Yes
- KNX RF (868.3 MHz)
Standards - De facto Standard
- Open Standard
- International Standard (IEEE 1901) - ANSI/ASHRAE 135-2008
- ISO 16484-5
- Open Standard
- CENELEC EN 50090 and CEN EN
13321-1
- ISO/IEC 14543-3
- GB/Z 20965, ANSI/ASHRAE 135
Speed - 20 bits/s - 4 to 10 Mbps - Depends on choice of LAN technology
used
- wired 9.6 kbps
- wireless 16.4 kbps
Maximum
Nodes
- 256 - 253 (theoretically)
- 10 (Practically)
- No limit - 57600 network nodes for wired
connection
Home-Area Networks (HANs)
Enabling Communications Technologies: Wired Technologies
52 IEEE GLOBECOM'11
Key
Criteria
X10 HomePlug GP BACnet KNX
Security - lack of encryption - AES pro 128
Security (128-bit triple AES encryption
and a time lock)
- Assume that all devices are sitting
behind a firewall
- EIBsec
Strength - Commonly used with variety of equipment
available in the market
- No installation cost as uses power line
-Ubiquitous reach throughout the home
environment
-Interoperability with consumer home
networking
-Low-cost and low-power network
interfaces
-Cross-compatibility between wired and
wireless Smart Grid applications
- Well established as an enabler for
commercial building automation
technologies
- Already has the needed functionality
for energy management and load control
- Independent of current LAN or WAN
technologies
-Scalable
- Interoperable with other KNX
products
- Hardware/software independent
- Well established promoter which
provide any application for home
control
- Compatible with any buildings
Concern - Extreme low data rate
- Lack of standard and security
- Limited functionality
- Prone to interference from neighbors using the
same X10 device addresses
- Limited connection (10) when
transferring data simultaneously
- Susceptible to power line interference
and old wiring in home.
- client-server system might create
bottleneck when fully deployed to
consumer premise
- Security concerns
- Object model is limited to low-level
types
- Low data rates
Neighborhood-Area Networks (NANs)
General Structure
53 IEEE GLOBECOM'11
Utility Network Back Bone
Wide Area Network
(WAN)
Neighborhood Area
Network (NAN)
HAN = Home Area Network
Cel
lula
r,IP
net
wor
k,
BPL
, WiM
AX
PLC,
ANSI C12
Neighborhood Area
Network (NAN)
Neighborhood Area
Network (NAN)
Neighborhood Area
Network (NAN)
MDMS
MDMS
DAU
HAN
DAU
HAN
DAU
HAN
DAU
HAN
Cellu
lar,IP n
etwork
,
BP
L, W
iMA
X
Neighborhood-Area Networks (NANs)
Neighborhood-Area Networks (NANs)
• NAN connects multiple HANs together
• Wired Technologies
– Power Line Communication (PLC):
• Ultra narrow band (UNB) operates in 0.3-3 kHz bands
• Narrow band (NB) PLC operates in 3-500 KHz bands
• Broadband (BB) PLC or BPL operates in 1.8-250 MHz bands
– Internet Protocol (IP)-Based Networks
– Internet Based Virtual Private Networks (Internet VPN)
• Internet VPN technology can provide reliable, secure, and robust alternative to
ensure security and QoS requirement
• Wireless Technologies
– 3G and LTE cellular Networks
– WiMAX Technology
54 IEEE GLOBECOM'11
Wide-Area Networks (WANs)
Core Communication Network and Last Mile Connectivity
55 IEEE GLOBECOM'11
Enabling Technolgies Scope Strength Concern
1. Power Line Communication - Communication Core
Network and Last Mile
Connection
- Complete control over the communication path
with extensive coverage that is solely controlled
by the utility industry
- Provides low cost solution to overlay the
communication network over already available
power lines
- Provides direct route between controllers and
other subsystem to ensure low latency
- Mature technology with many variants available
commercially
- The power line are connected to various equipments such as
motor, power supplies, which can act as noise sources that
eventually degrades the performance of PLC
- The load impedance fluctuation, and electromagnetic
interference causes signal attenuation and distortion, which can
result to failure of communication link
- Lack of standard status and government regulation due to
industry fragmentation result in high interference from other
PLC technology deployed at close range
- Cost of PLC modem are still high
- Coexistence issue from many commercial technologies
2. Internet Protocol (IP)-Based
Networks
- Communication Core
Network and Last Mile
Connection
- IP-based networks have rich convergence
capabilities which can help to connect the overall
systems and subsystems in smart grid
-Can provide QoS and reliable connection using
technologies such as DiffServ and MPLS
Security can be enhanced using technologies
(IPSec)
- In case of master/slave configuration, transmitting IP packets
from slave is not possible, which might increase the data
latency for those applications which requires fast response as in
case of smart grid
- Unless private IP-based network (e.g., Internet VPN) is used,
security remains crucial issue
Wide-Area Networks (WANs)
Core Communication Network and Last Mile Connectivity
56 IEEE GLOBECOM'11
Enabling Technolgies Scope Strength Concern
3. Wireless Communication - Communication Core Network and
Last Mile Connection
- Huge coverage area, potential for low cost
- Packet-Switched Cellular Data has lower cost and
much higher data rates
- WiMAX can support mesh networks for higher
reliability
- Utilities have to depend on these technologies
without any control over them
- Packet switch technologies are not available in all
deployed cellular structures
- Requires connection to network before transmitting
the data and might be problem in case of outage and
emergency
4. Communication and Networking
Middleware
- Communication Core Network - Hybrid network can provide better needs to
specific smart grid application
- Improves Interoperability
- Requires more research to combine technologies to
form network middleware
Communication Architectures
Standard Activities: Standard Developing Organization (SDO)
• ANSI - American National Standards Institute (www.ansi.org)
• IEC - International Electrotechnical Commission (www.iec.ch)
• IEEE - Institute of Electrical and Electronics Engineers (www.ieee.org)
• ISO - International Organization for Standardization (www.iso.org)
• ITU - International Telecommunication Union (www.itu.int)
57 IEEE GLOBECOM'11
Communication Architectures
Standard Activities (1)
58 IEEE GLOBECOM'11
Standards Application Strength Concern
ANSI C12 Suite
ANSI C12.19/IEEE 1377
ANSI C12.22
- Defines utility industry end device data
tables for representing the data produced by
revenue meters.
- Standard protocol for network
communication
- Defines format of data for meter
- Provides transport independent application
level protocol for data exchange with low
overheads between nodes.
- Supports transport of C12.9 table data
- Provide authentication & encrypting the C12.9
data.
- Does not specify protocol to transport it
-Lacks full interoperability due to
specialized local profile
- Requires complexity in implementation in
clients
ANSI/ASHRAE 135/ISO 16484-5
BACnet
- Defines information model & messages as
objects for providing common language for
different proprietary protocols
- Open, mature standard with interoperability
testing developed and maintained by SDOs
- Serves as customer side communication
protocol with relevancy in price, DR/DER &
energy usage
- Object model might be limited to low
level protocols
- Requires structural view & specific profile
to address consumer portals
ANSI /EIA/CEA 709 & CEA 8521
Protocol Suite LONworks
ANSI/CEA 709.1-B
ANSI/CEA 709.2
ANSI/CEA 709.3
ANSI/CEA 709.4
- General purpose LAN protocol for
providing communication over with home &
building automation
- The Control Network
- Power Line Carrier Physical Layer
- Twisted Pair Physical Layer
- Fiber Optic Physical Layer
- Widely used matured protocol
- Specify as one of the data link & physical
layer option for BACnet
- de facto standard controlled by Echelon
with limited support in power industry
- Lack of complex object model to support
function
Communication Architectures
Standard Activities (2)
59 IEEE GLOBECOM'11
Standards Application Strength Concern
ZigBee/ HomePlug Smart Energy Profile Strategic alliance of ZigBee & HomePlug to
provide communication & information model
in HAN
- Interoperable between two distinct HAN technology
- Technology independent
IEC 62056 Device Language Message
Specification (DLMS) &
Companion Specification for
Energy Metering (COSEM)
- Standard representation of metering data
used for accessing and exchanging structured
data models
- Supports object modeling of application data as object
identification system (OBIS) and the Open Systems
Interconnection (OSI) model
- Matured and internationally recognized standard
- Supports variety of media such as PSTN, GSM network,
PLC and recently ZigBee protocols
IEEE 1901 - Broadband communications over
Powerline medium access control
(MAC) and physical layer (PHY)
Protocols for HAN and also access
application
- High speed (>100 Mbps) communication for devices
using frequency below 100 MHz.
- Uses inter-system protocol (ISP), which allow device to
coexist with devices based on ITU-T G.hn standard
- Initiate harmonization and coexistence of PLC with
other technologies
- Has backward compatibility with HomePlug standard
- Short range due to higher
attenuation of the medium as a
result of using broadcast channels
above 80 MHz
Communication Architectures
Standard Activities (3)
60 IEEE GLOBECOM'11
Standards Application Strength Concern
ITU-T G.hn/G.9960 Home Networking
Standard
- In-home networking over power lines,
phone lines, and coaxial
cables
- Designed especially for HAN
- Use single fast Fourier
transform (FFT) OFDM modulation and low-
density parity-check code (LDPC) forward
error correction (FEC) code
- Does not address PLC access application
- Does not support HomePlug standard
ISO/IEC 15045, A residential gateway
model for Home electronic
system
- Defines specification for residential
gateway (RG) that connects HAN to network
domain outside the home basically last mile
connection
- Defines functional requirement &
architecture for RG
- Defines security requirements for connecting
to WANs
- Still under consideration by independent
organization
ISO/IEC 15067-3, Model for an energy
management system for Home
electronic system
- Defines a model for energy management
system that accommodates a range of load
control strategies
- Specifies methods for demand response that
may be implemented by an electric utility or
by a third-party supplier of energy
management services
- Supports various smart appliances
Communication Architectures
Cognitive Radio [Yu_2011] (1)
• Cognitive radio based communications architecture is
presented for the smart grid
• Cognitive radio allows unlicensed (secondary) user to
access spectrum licensed to licensed (primary) user
– Improve spectrum utilization
– Improve spectrum efficiency
• The proposed architecture is motivated by
– Explosive data volume
– Diverse data traffic
– Need for QoS support
61 IEEE GLOBECOM'11
Communication Architectures
Cognitive Radio [Yu_2011] (2)
• Proposed Network Architecture
62 IEEE GLOBECOM'11
Communication Architectures
Cognitive Radio [Yu_2011] (3)
63 IEEE GLOBECOM'11
Cognitive area network
Home area network (HAN) Neighborhood area network (NAN)
Wide area network (WAN)
Spectrum band Unlicensed band Licensed band Licensed band
Network topology Centralized/decentralized Centralized Centralized
Network users Smart meters/sensors/acuators HGW
HGWs, NGWs spectrum broker
Featured strategy Cross-layer spectrum sharing Hybrid dynamic spectrum access
Optimal spectrum leasing
Key techniques Access control, power coordination
Guard channel, spectrum handoff
Join spectrum management
Communication Architectures
Cognitive Radio [Yu_2011] (4)
• Dynamic Spectrum Sharing in Cognitive HAN
– HGW will connect to the HAN, which in turn will connect to
external networks (e.g., Internet and NAN)
– Within a HAN, the HAN cognitive gateway (HGW) manages the
license-free spectrum bands to provide optimal data rate with low
interference
– HGW enables other devices and sensors to join the network, assigns
channel and network addresses to each device, and coordinates the
communications between the devices within the HAN
64 IEEE GLOBECOM'11
Communication Architectures
Cognitive Radio [Yu_2011] (5)
• Cognitive Communications in Neighborhood Area Network
(NAN)
– NAN Cognitive gateway (NGW) connects several HGWs from
multiple HANs together
– Hybrid dynamic spectrum access (H-DSA) is proposed
– Some licensed spectrum bands are leased/bought from a
telecommunication operator, and these bands are used as licensed
access for the HGWs to ensure the QoS of data communications
– The NGW distributes these licensed bands to the HGWs according
to the transmission demand
– However, if licensed spectrum bands are not enough to meet the
demand, unlicensed access is also needed for the HGWs to improve
the capacity and throughput of the NAN
– In unlicensed access, the HGWs and NGW could be considered
secondary users 65
IEEE GLOBECOM'11
Communication Architectures
Cognitive Radio [Yu_2011] (6)
• Cognitive Communications in Wide Area Network (WAN)
– In WAN, each NGW is a cognitive node with the capability to
communicate with the control center through frequency space
unused by a licensed primary user
– Control center is connected with cognitive radio base stations
– Spectrum broker controls sharing the spectrum resources among
different NANs to enable coexistence of multiple NANs
– Joint WAN/NAN spectrum management is proposed by minimizing
the maximum dropping probability of data connection in NAN
66 IEEE GLOBECOM'11
Sensor and Actuator Networks (SANETs)
Applications of Data Sensing in Smart Grid
• Power Generation
– WSN called WiMMS unit is deployed in the wind turbine structure [Wang_2007] to
provide information about dynamic behavior of wind turbine and response to loading
– For energy storage, lead-acid batteries will be used, and sensor network can be used
to monitor temperature, voltage, and current
• Power Transmission and Distribution
– Data sensing can be used to monitor substations, transformers, underground lines,
and overhead lines
• Power Consumption
– Smart meter acts as a sensor node and records the electricity consumption (kilo watt
hour [kWh]) and time of use (TOU)
67 IEEE GLOBECOM'11
Sensor and Actuator Networks (SANETs)
Requirements for Data Sensing and Communication
• Sensor and Actuator Requirements
– Longer life span
– Reliability and energy-efficiency
– Cost-effectiveness and secured operation
• Data Collection Requirements
– Machine readable format
– Contain the temporal information including the time-stamp
– Identification of location
• Requirements for Communication Networks
– Distributed operation
– Interoperability
– Scalability
– Security
68 IEEE GLOBECOM'11
Sensor and Actuator Networks (SANETs)
SANET in Transmission Line Monitoring [Hung_2010]
• The linear sensor network for transmission line is analyzed
• Accelerometer (inclination and cable position and tilt), magnetic field
sensor (current and power quality), strain sensor, and temperature sensor
are considered
69 IEEE GLOBECOM'11
Sensor and Actuator Networks (SANETs)
Approaches for Data Sensing
• Phasor Measurement Units
– Phasor measurement units (PMUs) (also referred as synchrophasors) measure the
electrical waves, using a common time source for synchronization
– IEEE Standard C37.118-2005 deals with issues concerning the use of PMUs in
electric power systems
• Compressive Sensing
– Compressive sensing (CS) is proposed which links data acquisition, compression,
dimensionality reduction, and optimization together
– CS senses less and computes more to obtain the useful data
• Decentralized and Cooperative Sensing
– Distributed information processing and control are needed in power system
operations
– For example, distributed state estimation methods have been considered for decades
with the goal of reducing the computational burden at the central control by
distributing the tasks across the system.
70 IEEE GLOBECOM'11
Sensor and Actuator Networks (SANETs)
Approaches for Data Communication
• Cooperative Communications
– Cooperative communications refer to the techniques in which multiple nodes help
each other (e.g., in wireless mesh, ad hoc, and sensor networks) to relay or forward
data packets to their destinations
– Cooperative wireless sensor network (IEEE 802.15.4 ZigBee) is used to provide data
transmission in urban-scale smart grid environment [Ullo_2010]
– Secure and reliable collaborative communication scheme for advanced metering
infrastructure (AMI) is introduced [Yan_2011]
– Multihop wireless network is used to connect smart meters with AMI to transfer
meter data to a local collector
• Cognitive Radio
– CR-based wireless sensor network using the 802.15.4 ZigBee standard is proposed in
[Sreesha_2011]
– In the design, a coordinator is used to provide the synchronization and control of data
transmission, while a spectrum sensor is used to support frequency agility so that the
transmission can be adapted based on the wireless channel condition
71 IEEE GLOBECOM'11
Data Communications and Networking in
Smart Grid
• Demand Response Management (DRM)
• Home Energy Management System (HEMS)
• Advanced Metering Infrastructure (AMI)
• Wide-Area Measurement Systems (WAMSs)
72 IEEE GLOBECOM'11
Demand Response Management (DRM)
• DRM is the programs implemented by utility companies to
control the energy consumption at customer side
73 IEEE GLOBECOM'11
Permanent Days Seconds Time
Optimality
Optimized
infrastructure
Optimized
schedule
Temporary
adjustment
Energy
efficiency
TOU
Market
DR
Physical
DR
Spinning
reserve
Demand Response Management (DRM)
• Energy efficiency focuses on users and behavioral changes
to achieve more efficient energy usage
– Users buy appliance with energy reduction feature
74 IEEE GLOBECOM'11
Demand Response Management (DRM)
• Smart pricing or time of use (TOU)
– Customers (re)arrange their energy consumption to minimize costs
• Market demand response
– Direct load control (DLC): utility or grid operator control energy
consumption of consumers
– Interruptible/curtailable rates: customers has a contract with limited
sheds feature from utility
– Emergency demand response programs: customers voluntarily adjust
energy consumption based on emergency signals (e.g., blackout)
– Demand bidding programs: customers can bid for curtailing at
attractive price
75 IEEE GLOBECOM'11
Demand Response Management (DRM)
• Physical demand response
– Grid management and emergency signals (on the utility side)
– Signal if the grid (power lines, transformers, and substations) are in
a reduced performance due to maintenance or failure
• Spinning Reserves (SR)
– Generators are online, synchronized to the grid, that can increase
output immediately in response to a major outage and can reach full
capacity [Hirst_1998]
76 IEEE GLOBECOM'11
Demand Response Management (DRM)
• Energy efficiency vs. demand response
77 IEEE GLOBECOM'11
Original load
Time
Energy consumption
Energy efficiency
Demand response
without rebound Demand response
with rebound
Demand Response Management (DRM)
Residential load management [Mohsenian-Rad_2010]
• Residential load management programs usually are to
reducing consumption and shifting consumption
• In direct load control (DLC), utility company sets up an
agreement with its customers
• Utility company can manage and control remotely the
operations and energy consumption of certain household
appliances
– Lighting
– Thermal and cooling system
– Refrigerators
– Pumps
78 IEEE GLOBECOM'11
Demand Response Management (DRM)
Smart Pricing [Mohsenian-Rad_2010]
• With smart pricing, energy consumers are encouraged to
individually and voluntarily manage their loads
– Reducing their consumption at peak hours
• Critical-peak pricing (CPP), time-of-use pricing (ToUP),
and real-time pricing (RTP) can be used
• For example, in RTP, the price of electricity varies at
different hours of the day
– Prices are usually higher during the afternoon, on hot days in the
summer, and on cold days in the winter
79 IEEE GLOBECOM'11
Home Energy Management System (HEMS)
• HEMS acts as the subset of energy management system
(EMS) and together with smart meter provides a necessary
interface to the HAN for better energy management
80 IEEE GLOBECOM'11
Heating and cooling 49%
Water heater 13%
Refrigerator 5%
Dishwasher 2%
Clothes washer & Dryer 6%
Lighting 10%
Electronics 7% Other 8%
http://www.energystar.gov/
Home Energy Management System (HEMS)
• HEMS (or EMS) sets a certain user limit threshold based on
the information about real-time price-responsive load
management and consumption history (i.e., collected from
smart meter) to control the energy usage of appliances
• HEMS is generally integrated into HAN to offer a channel
for the consumers to interact with the electrical power grid
• HEMS may reside in the smart meter or in an independent
gateway such as residential gateway and network adapters
• HAN contains many electrical appliances (e.g., routers, TV,
AC, computers, etc) which provide different services, e.g.,
wireless access, VoIP calls, ambient temperature control
• These services can be controlled by using different power
control elements (PCEs) such as Ethernet switch, PSTN,
and DSL modem 81
IEEE GLOBECOM'11
Home Energy Management System (HEMS)
• Example: GE Demand Reduction Approach
82 IEEE GLOBECOM'11
Price Event Signal to
Smart Appliance
Smart Appliance will indicate
to consumer
Price Event has occurred
Smart Appliance will
recommend to delay
start
Over
Ride?
Over
Ride?
Run Normal operating mode
Consumer Choice
Initiate delayed start function
Initiate peak reduction mode
Data communications Electric supply from
Transmission
HAN
Gateway
In-Home
Display
DAU/NAN
Sensors
Light
Temperature
Voltage
Wired/Wireless
Connection
(e.g., Zigbee,
BACnet
Smart
Devices
Smart
Devices
Smart
Devices
Actuators
ActuatorsActuators
(Smart Meter
or
Dedicated
in-home
Gateway)
Home Energy Management System (HEMS)
Machine-to-Machine Communications [Niyato2011]
• Network design issue of M2M communications for a home
energy management system (HEMS) is considered
• The network architecture for HEMS to collect status and
power consumption demand from home appliances is
introduced
• Optimal HEMS traffic concentration is presented and
formulated as the optimal cluster formation
83 IEEE GLOBECOM'11
Home Energy Management System (HEMS)
Machine-to-Machine Communications
• Network model
84 IEEE GLOBECOM'11
Concentrator
Base stationInternet
backhaul
Control
center
Service area with wide area network (WAN)
Neighborhood area network (NAN)
Home area
network (HAN)
Smart meter
Home Energy Management System (HEMS)
Machine-to-Machine Communications
• Optimal cluster is determined
• The average cost per node under different packet generation
rates is shown
85 IEEE GLOBECOM'11
0.4 0.6 0.8 1 1.2 1.4 1.6 1.80
2
4
6
8
10
12
14
16
18
20
Packet generation rate (packets/minute)
Avera
ge c
ost
per
node
Optimal formation
Fixed formation
Cluster size = 10
Cluster size = 5
Cluster size = 1
Cluster size = 2
Cluster size = 4
Cluster size = 3
Advanced Metering Infrastructure
• AMI acts as the gateway for access enabling the
bidirectional flow of information and power in support of
distributed energy resource (DER) management or
distributed generation (DG) and consumer participation
• AMI will provide near real-time consumption data including
fault and outage to the utility control center
• AMI supports time-based and dynamic tariffs such as Time
of Use (TOU), Real-Time Pricing (RTP), and Critical Peak
Pricing (CPP)
• AMI consists of several different components
– Smart meters and data aggregator units (DAUs))
– Hierarchical area networks (e.g., home-area networks (HANs) and
neighborhood-area networks (NANs), and wide-area networks
(WANs))
86 IEEE GLOBECOM'11
Advanced Metering Infrastructure
• Comparison
87 IEEE GLOBECOM'11
Manual/Automatic Meter Reading (AMR)
AMI
Pricing Fixed price and measure total consumption only
Total consumption Time-of-use Critical peak pricing Real-time pricing
Other demand response None Load control Demand bidding Demand reserves Critical peak rebates
Customer feedback Monthly bill Monthly bill Monthly detailed report Web display In-home display
Customer bill savings Turn off appliances manually Turn off appliances Shift appliances off peak Manual or automatic control
Outages Customer phone calls Automatic detection Verification of restoration at individual home level
Distribution operations Use engineering models Dynamic, real-time operations
Advanced Metering Infrastructure
Benefit of AMI [Liu_2010]
• Fault Location, Isolation and Service Restoration (FLISR)
– AMI will be able to automatically report loss of power, and the information can be
used to assist locating the fault location
• Emergency Load Shedding
– AMI helps to shed large amounts of load very quickly (within seconds) to avoid
power system instability and loss of system integrity (e.g., during bulk power grid
emergencies)
• Distribution System Planning and Analysis
– AMI provides accurately metered data for all customers on the feeder from billing
records, and this information will enable the system to prepare much more accurate
short term load forecast
• Continuous Condition Monitoring
• Equipment and System Performance Forecasting
• Automated “Triggering” for Maintenance and Work Assignments
• Substation and Line Monitoring
88 IEEE GLOBECOM'11
Advanced Metering Infrastructure
Wireless Broadband Architecture [Mao_011] and Key Design
Issues
• Address Depletion
– For AMI, a very large number of new subscriber devices, i.e. smart meters, will need
address for communications
• Traffic Scheduling
– Critical alarm indication data should be reported immediately and not be queued until
the next scheduled connected period
• Congestion Control
– A very large numbers of SM give rise to potential “traffic burst” scenarios which
arise when large numbers of devices are simultaneously (or near simultaneous)
reporting or reacting to a common event
89 IEEE GLOBECOM'11
Advanced Metering Infrastructure
Service-Oriented AMI [Chen_2010]
• Service-oriented approach to AMI aiming at solving the
intercommunication problem and meanwhile providing a trust and
secure environment for smart grids
– System integration and cooperation are done through service composition.
– Generic service interfacing method is designed to develop standardized
– services for heterogeneous power systems
– Role-based access control mechanism is used to guarantee secure access
90 IEEE GLOBECOM'11
Advanced Metering Infrastructure
Reliability Analysis
• Reliability analysis of the wireless communications system
in the smart grid can be performed
• Availability performance can be obtained given the random
failure of the system devices
• Availability measure can be used to calculate the cost of
power-demand estimation error and damage of power
distribution equipment if its failure cannot be reported
• Redundancy design approaches can be developed to
minimize the cost of failure as well as the cost of
deployment of the wireless communications system in the
smart grid
91 IEEE GLOBECOM'11
Advanced Metering Infrastructure
Reliability Analysis
92 IEEE GLOBECOM'11
Neighborhood area
network (NAN)
Home area
network (HAN)
HAN gateway and
smart meter
NAN gateway Data aggregator
unit (DAU)
Meter data-management system (MDMS)
NAN with gateway
redundancy
Power distribution
equipment
Advanced Metering Infrastructure
Reliability Analysis: Operation of a power system
93 IEEE GLOBECOM'11
Smart meter estimates power
demand in the next period
(e.g., using power scheduling)
HAN gateway sends power
demand collected from smart meter
to the corresponding NAN gateway to
forward to DAU and subsequently MDMS
MDMS buys additional
power supply in
economic dispatch stage Power demand is added into
amount of power to be supplied
Power demand of each house
is received by MDMS?
MDMS uses mean power
consumption of that house to
compute amount of
power to be supplied
MDMS buys power supply
in unit commitment stage
MDMS checks if power
supply is enough or not?
No
Yes
Yes
No
Advanced Metering Infrastructure
Reliability Analysis: Operation of a power system • If the power demand of any house is not received by the MDMS (e.g.,
due to failure of the HAN gateway, the NAN gateway, or the DAU), the
MDMS uses historical data to compute the aggregated power demand
• x% of mean power-consumption1 of those houses is used as the
estimated demand
94 IEEE GLOBECOM'11
Power consumption (kWh)
Pro
bab
ility
dis
trib
ution
Estimated power demand (i.e., reserved
power from unit commitment stage) for
x=100% of mean
Cost of under-reservation
Cost of over-reservation
0
Advanced Metering Infrastructure
Reliability Analysis: Availability • Availability of a component/device/system is the probability that the
component/device/system has not failed or repaired and it can operate
normally
• Uptime is also known as the mean time between failure (MTBF)
• Downtime is known as the mean time between repair (MTBR)
• Failure rate can be obtained a 1-Availability
95 IEEE GLOBECOM'11
Advanced Metering Infrastructure
Reliability Analysis: Availability • Dependence diagram (DD) determines the contribution of each
component to the availability of the system
• The components can be connected in parallel and/or series
96 IEEE GLOBECOM'11
Radio interface Single board computer Adaptor
Power
Software
Power
Control unit Radio interface Power Metering engine
Dependence diagram of smart meter and home area network gateway
Dependence diagram of neighborhood area network gateway
Node B Radio network
controller (RNC)
Service gateway
support node (SGSN)
GPRS gateway support
node (GGSN)
Dependence diagram of UMTS network
Advanced Metering Infrastructure
Reliability Analysis: Availability • HAN gateway and a smart meter can be integrated into a single device.
The availability of a HAN gateway is computed from AHAN =
availability of metering engine × availability of control unit ×
availability of power module × availability of radio interface
• Availability of a NAN gateway is computed from ANAN = availability of
radio interface × availability of single board computer × availability of
adaptor × availability of software × (1 − (1−availability of power
module)2)
• 3G cellular base station is assumed to have the DAU functionality
whose availability is computed from: ADAU = availability of node B ×
availability of radio network controller (RNC) × availability of service
gateway support node (SGSN) × availability of GPRS gateway support
node (GGSN)
97 IEEE GLOBECOM'11
Advanced Metering Infrastructure
Reliability Analysis: Cost of Network Unavailability • Cost of demand-estimation error of individual house i whose connection
to the MDMS is unavailable can be obtained from
• Ei = x/100 × Meani is the power supply reserved in the unit
commitment stage
• Meani is the mean power-consumption of house I
• Maxi is the maximum power-consumption
• fA(i)(a) is the PDF of actual power demand a
• puc and ped denote the power prices in the unit commitment and in the
economic dispatch stages, respectively
98 IEEE GLOBECOM'11
Advanced Metering Infrastructure
Reliability Analysis: Cost of Network Unavailability
• Number of houses
• Number of redundant NAN gateways
99 IEEE GLOBECOM'11
20 40 60 80 100 120 140 160 180 2000
10
20
30
40
50
60
70
Number of houses in NAN
Cost
of
dem
and e
stim
ation e
rror
per
month
($)
Failure rate of HAN gateway = 2 days in 1 years
Failure rate of HAN gateway = 2 days in 2 years
Failure rate of HAN gateway = 2 days in 3 years
Failure rate of HAN gateway = 2 days in 4 years
0 1 2 3 4 5 6 7 8 9 10160
180
200
220
240
260
280
300
Number of redundant NAN gateways
Avera
ge t
ota
l cost
per
month
($)
Failure rate of NAN gateway = 2 day in 2 years
Failure rate of NAN gateway = 2 day in 3 years
Failure rate of NAN gateway = 2 day in 4 years
Wide-Area Measurement Systems (WAMSs)
• WAMS is used to conduct real time monitoring and control
in dynamic power system states
• WAMS uses a synchronized phasor measurement unit
(PMU) to guarantee for security and stability of power
systems
• WAMS is typically composed of PMUs, phasor data
concentrator (PDC), control center (CC), as well as the
high-speed data communication networks
100 IEEE GLOBECOM'11
Wide-Area Measurement Systems (WAMSs)
Applications [Naduvathuparambil_2002]
• State estimation: PMUs can measure and relay information
on a continuous basis to the control centers, and control
center will generate a state vector of system dynamics
• Instability prediction: Synchronized phasor measurements
can enable real-time stability analysis and instability
prediction
• Improved control of power systems: Controllers (e.g.,
variable series capacitors [VSC], universal power flow
controllers [UPFCs] and power system stabilizers) can
receive feedback from control center to regulate the grid
101 IEEE GLOBECOM'11
Wide-Area Measurement Systems (WAMSs)
Data Communication
• Telephone lines
– Easy to set up and economical to use, but low speed
• Fiber-optic cables
– Immunity to RF & atmospheric interference
– Large bandwidth
• Satellites: low-earth orbiting (LEO)
– Large coverage area,
– High cost, narrow bandwidth, and large delays
• Power lines
– Uses the medium and low voltage electric supply grid for transmission of
data and voice
• Microwave links
– Easy to set up and are highly reliable
– Signal fading and multipath propagation
102 IEEE GLOBECOM'11
Wide-Area Measurement Systems (WAMSs)
Centralized WAMS [Shahraeini_2011]
• All data resources send data to control center (CC)
• After processing the received data, appropriate decisions are
made and related commands are sent back to controllable
devices
103 IEEE GLOBECOM'11
Wide-Area Measurement Systems (WAMSs)
Decentralized WAMS [Shahraeini_2011]
• System is divided into multiple areas
• Each area has its own are control center area (ACC)
• In each area, ACC processes the acquired data and perform
control
• For the control of a system, ACCs share information among
each other through communication systems
104 IEEE GLOBECOM'11
Reliability Analysis
• [Bruce_1998], [Xie_2002], [Wang_2010]
– Synchronized phasor measurement unit (PMU)
– Phasor data concentrator (PDC)
– Ring interface unit (RIU)
– Control center (CC)
105 IEEE GLOBECOM'11
Fault tree analysis of WAMS
Wide-Area Measurement Systems (WAMSs)
Reliability Analysis
• Availability is calculated from
• Ai is the availability of the th PMUs-PDC working group
• AijPMU is the availability of PMU j in PMUs-PDC working group i
• Mi is the number of PMUs in group I
• AiRN is availability of regional communication network
• AiPDC is availability of PDC device
106 IEEE GLOBECOM'11
Wide-Area Measurement Systems (WAMSs)
Cyber Security for Smart Grid
• Introduction
• Why do we need cyber security
• Adversaries
• Threats
• Impacts
• How to achieve cyber security
• Survey some solutions
IEEE GLOBECOM'11
107
Transmission
TOP1 – Operational Information
Distribution
DIST1 - Operational Information
DISTx – Operational Information
Customers Generation
GEN1 - Operational Information
GENx - Operational Information
Current Electric Grid – Islands of Technology
TOPx – Operational Information
IEEE GLOBECOM'11
108
Convergence of Enterprise & Operations IT
Enterprise Systems
Web Applications
Control Systems
Protection Systems
Information Technology Operations Technology
AMI
DSM
OMS
GIS
Smart Grid Technology
Integration counters key security principals of isolation and segregation
Cyber Secure
Integration counters key security principals of isolation and segregation
Convergence of Enterprise & Operations IT
IEEE GLOBECOM'11
109
Smart Grid – Connectivity with Security
Transmission Distribution Customers Generation
System
Operators
Conservation
Authorities
End-to-End Communications, Intelligence, and Defense-in-Depth Security
AMI DSM
IEEE GLOBECOM'11
110
Why do we need cyber security ?
• Network security is a priority and not a add on for smart
grids
• Protecting control center alone - not enough
• Remote access to devices
• QoS requirement from security system
• Safety (line worker public and equipment)
• Reliability and availability
111
Drivers
Increasing Number
Of Systems and
Size of Code Base
Control Systems
Not Designed with
Security in Mind
Increasing Use of
COTS Hardware
and Software
New Customer
Touch Points into
Utilities
New 2-Way
Systems
(e.g. AMI, DSM)
Increasing
Interconnection
and Integration
Increased Attack Surface
Increased Risk to Operations
112
Threats-I
Example from 2006 SANS SCADA Security Summit, INL
1. Hacker sends an e-mail with malware
2. E-mail recipient opens the e-mail and the
malware gets installed quietly
3. Using the information that malware gets,
hacker is able to take control of the e-mail
recipient’s PC!
4. Hacker performs an ARP (Address Resolution
Protocol) Scan
5. Once the Slave Database is found, hacker sends
an SQL EXEC command
6. Performs another ARP Scan
7. Takes control of RTU
Internet
Admin
Acct
Opens
Email with
Malware
Admin
Send e-mail
with malware
Slave
Database
Operator
Operator
Master
DB
RTU
Perform
ARP Scan
SQL
EXEC
Perform
ARP Scan
113
Example from AMRA
Webinar, Nov ’06
“The Active Attacker”
Threats-II
U N I V E R S I T YU N I V E R S I T Y
AMI WAN AMI WAN AMI WAN
Communications
Network
(WAN)
Communications
Network
(WAN)
Data Management
Systems
(MDM/R)
Retailers
3rd Parties
AMCC
(Advanced Metering
Control Computer)
Attacker
Cyber
Penetration
Attacker
Controls the
Head End
Attacker
Performs
Remote
Disconnect
114
Impacts-I
Meter
EMS
AMI
Network
HAN
Energy
Service
Provider
Wide Effect, High Impact on
the Grid, Attacker may be
Remote
Local Effect, Narrow Impact,
Attacker Needs to be Local
Energy Consumption Data
Demand Response Trigger
Utility
Back Office
Direct Energy Information Access from Meter and Local Control in
Customer Premises has Lowest Risk
The Impact of a Security Breach*
* Does not represent the difficulty or ease of executing the breach.
115
Impact-II
Threat Attacker
Location
Impact
Spread
Impact
Effect
AMI Network
Compromised Remote Wide Network
Stability
DR Manipulated in
―Cloud‖*
Remote Wide Network
Stability
Customer Privacy
Breached in ―Cloud‖* Remote Wide Loss of Privacy
HAN Compromised Local Narrow
Local
Nuisance
* “Cloud” refers to both a Utility Back Office and Energy Service Provider
IEEE GLOBECOM'11
116
Cost of Power
Disturbances:
$25 - $188 billion
per year
~$6 billion lost
due to 8/14/03
blackout
Northeast Blackout – August 14, 2003
• Affected 55 million
people
• $6 billion lost
• Per year $135
billions lost for
power interruption
http://en.wikipedia.org/wiki/Northeast_Blackout_of_2003
IEEE GLOBECOM'11
117
Adversaries
• Hostile States
• Hackers
• Terrorist /Cyber terrorists
• Organized crime
• Other criminal elements
• Industrial competitors
• Disgruntled employees
• Careless and poorly trained employees
IEEE GLOBECOM'11
118
COMPONENT BASED ATTACK -STUXNET
• Specifically programmed to attack SCADA and could reprogram
PLC‟s
• Zero day attack
• Highly complex
• 0.5 Mb file transferred able to multiply
• Targets- Iran nuclear plants ,Process plants in Germany and ISRO
India
Source: wikipedia
IEEE GLOBECOM'11
120
COMPONENT BASED ATTACK - SCADA attacks
• Internal attacks Employee
Contractor
• External attacks Non specific- malware , hackers
Targeted Special knowledge – former insider
No special knowledge –hacker terrorist
Natural disaster
Manmade disasters
IEEE GLOBECOM'11
121
SCADA – vulnerability points
• Unused telephone line – war dialing
• Use of removable media – stuxnet
• Infected Bluetooth enabled devices
• Wi-Fi enabled computer that has Ethernet connection to scada
system
• Insufficiently secure Wi-Fi
• Corporate LAN /WAN
• Corporate web server email servers internet gateways
IEEE GLOBECOM'11
122
SCADA-CYBER ATTACKS
• Web servers or SQL attacks
• Email attacks
• Zombie recruitment
• DDOS attacks
IEEE GLOBECOM'11
123
Protocol based attacks
• All protocols runs on top of IP protocol and IP protocol has its own set of
weakness
• DNP3 implements TLS and SSL encryption which is weak
• The protocol is vulnerable to out-of-order, unexpected or incorrectly formatted
packets
• A significant weakness for IEC 61850 is that it maps to MMS (Manufacturing
message specification)as the communications platform, which itself has a wide
range of potential vulnerabilities
IEEE GLOBECOM'11
124
Challenges
• The challenge is complex and continuously changing
• Legacy systems need to be protected
• Number and geographic location of end points
• Relationship to physical security
• Systems are 7x24 and critical
• The human element / social engineering
IEEE GLOBECOM'11
125
Challenges („cont.)
• Scale
• Legacy devices
• Field location
• Culture of security through obscurity
• Evolving standards and regulations
IEEE GLOBECOM'11
126
How to achieve cyber security?
• Security by obscurity
• Trust no one
• Layered security framework
• Efficient firewall
• Intrusion detection
• Self healing security system
IEEE GLOBECOM'11
127
Types of Cyber Security Solutions
• Reactive vs. Proactive – Reactive
o Incident response plan
o Applied for general purpose computers more
– Proactive Security for embedded computers
• High assurance boot
• Secure software validation
• Secure association termination if found
infected
• Device assentation
IEEE GLOBECOM'11
128
Solution - Incidence response plan
IEEE GLOBECOM'11
Attack
Prevention Services
Containment Services
Detection &
Notification Services
Recovery &
Restoration Services
129
Solution - Defense in Depth
• Perimeter Protection
– Firewall, IPS, VPN, AV
– Host IDS, Host AV
– DMZ
– Physical Security
• Interior Security
– Firewall, IDS, VPN, AV
– Host IDS, Host AV
– IEEE P1711 (Serial Connections)
– NAC
– Scanning
• Monitoring
• Management
• Processes
IDS Intrusion Detection System
IPS Intrusion Prevention System
DMZ DeMilitarized Zone
VPN Virtual Private Network (encrypted)
AV Anti-Virus (anti-malware)
NAC Network Admission Control
IEEE GLOBECOM'11
130
Solution –Control Network
Internet
Enterprise Network
Control Network
Field Site Field Site Field Site
Partner
Site
VPN
VPN
FW
FW
IPS
IDS
Scan
AV
FW IPS
P1711
FW
AV Host IPS Host AV Proxy
Host IDS Host AV
IDS Scan
NAC
NAC
• Defense in Depth
• Access Control
• Secure connections
• Link to Physical
• Security Management
• Apply same approach
to other Smart Grid
elements
Key Points:
131
Solution – Key management
• Issue of key management – Scale
• PKI with trusted computing elements- considerable
amount of security
• Embedded vs. general-purpose computing
IEEE GLOBECOM'11
132
Issues with PKI
• Updating the keys
• Parameter generation
• Key distribution
• Staffing for key management
IEEE GLOBECOM'11
134
Calculation of cyber security conditions (omega)
IEEE GLOBECOM'11
136
Rules for Conditions 1, 2, and 3
Conditions Rules
Condition 1 The system is free of intrusion attempt that
is concluded from the electronic evidences
in the system
Condition 2 At least one or more countermeasures are
implemented to protect an attack leaf.
Condition 3 At least one or more password policies are
enforced corresponding to each attack leaf.
Calculations of vulnerability index
• Leaf VI : max( total countermeasures implemented
/total countermeasures available x ω , ω x weighing
factor of password policy)
• Scenario vulnerability index : Product of its leaf
vulnerability indices
• System vulnerability index is the max of all
scenario vulnerabilities indices
IEEE GLOBECOM'11
138
State estimation attack - introduction
• State estimation is to determine the optimal estimate for the complex voltages at each bus based on real-time analog measurements. – The state typically refers to bus voltage magnitudes
and phase angles
• Bad data processing is to detect measurement errors, and identify and eliminate them if possible. – It is effective against random noises, but
– It lacks the ability to detect intentionally coordinated bad data
• That conforms to the network topology and physical laws
IEEE GLOBECOM'11
139
State estimation attack - 1
• Attack on state estimation [Giani_2011]
– By compromising some line meters, sending wrong information about voltage / current status
• Force the energy management system to make wrong balancing operations that causes outage
– Main characters of the attack
• Sparse attacks are common (unobservable attacks)
– Large number of coordinated attacks can be detected by a bad data detection algorithm
• [Giani_2011] A. Giani, E. Bitary, M. Garciay, M. McQueenz, P. Khargonekarx, and K. Poolla, “Smart
Grid Data Integrity Attacks: Characterizations and Countermeasures”, Proceedings of IEEE SmartGridComm 2011.
IEEE GLOBECOM'11
140
Main contributions
• An efficient detection algorithm for – Case I : the attackers compromise
• Two power injection meters coordinately
• Arbitrary number of line meters
• The algorithm require O(n2×m) flops – n is the number of buses , m is the number of line meters
– Case II • Limited number of coordinated meters for attack (i.e., 3, 4, or 5)
• All lines are metered
• The algorithm requires O(n2) flops
• Countermeasures – Using known-secure PMUs for counteracting the attacks
– Demonstrate that p+1 PMUs are enough to neutralize a collection of p cyberattacks
• The positions of PMUs need to be carefully chosen
IEEE GLOBECOM'11
141
State estimation attack - 2
• Study the vulnerability of the state estimator to attacks performed against the communication infrastructure [Vukovic_2011]
• Use the security metrics defined by them to show – how various network and application layer mitigation
strategies can be used • to decrease the vulnerability of the state estimator
• Background – An attacker that wants to change the measurement on one
substation might have to change several other measurements • To avoid a bad data detection (BDD) alarm
• [Vukovic_2011] O. Vukovic, K-C Sou, G. Dan, and H. Sandberg, “Network-layer Protection Schemes
against Stealth Attacks on State Estimators in Power Systems”, Proceedings of IEEE SmartGridComm 2011.
IEEE GLOBECOM'11
142
Main ideas
• Substation is the weak point – Measurement data are usually collected through
substations
– An attacker can access and modify all data that traverses a substation
– The authors proposed to assess the importance of each substation with respect to state estimation
• Security metrics – Substation attack impact
• The number of measurements on which an attack can perform a stealth attack
– Measurement of attack cost • Minimum number of substations that have to be attacked in
order to perform attack against the measurement
IEEE GLOBECOM'11
143
Main contributions
• Protective methods
– Network layer solutions
• Single-route routing vs. Multi-path routing
• Modify single-route path to decrease the vulnerability
of the system
• Multi-path routing could reduce the maximum attack
impact by 50%
– Application layer solutions
• Data authentication increases the attack cost
– The solutions are very realistic
IEEE GLOBECOM'11
144
State estimation attack - 3
• This paper introduced a procedure that aims to achieve network-wide optimal attack detection and state estimation [Tajer_2011]
• The procedure is distributed – Different controlling agents distributed across the
network carry out the attack detection and system recovery tasks through
• local processing and message passing, and
• An iterative process
– Distributed state estimation method can reduce the computational burden on the centralized control system
• Using a decompose-merge approach
• [Tajer_2011] A. Tajer, S. Kar, V. Poor, and S. Cui, “Distributed Joint Cyber Attack Detection and State
Recovery in Smart Grids”, Proceedings of IEEE Globecom 2011.
IEEE GLOBECOM'11
145
Main contributions
• Reliable detection + reliable estimate of the
false injected data
– Means that the system can still obtain relatively
accurate estimation of the data in spite of attacks
– Different from works that avoid data to be
compromised
– Used an information theoretic method
IEEE GLOBECOM'11
146
State estimation attack - 4
• This paper [Esmalifalak_2011] demonstrate an attack method that – Inject false data with low detectability
– Without knowledge of the network topology
– Makes the inference from the correlations of line measurements
• But assume that the attackers can break into the SCADA system
• Main contributions – Demonstrate that an attacker can estimate both the system
topology and power states just by observing the power flow measurements
– Independent component analysis (ICA) is used • to infer the linear structure of the power flow measurements
• [Esmalifalak_2011] M. Esmalifalak, H. Nguyen, R. Zheng, and Z. Han, “Stealth False Data Injection using Independent Component
Analysis in Smart Grid”, Proceedings of IEEE SmartGridComm 2011.
IEEE GLOBECOM'11
147
Performance evaluation
• The authors demonstrated that
– The ICA based attack is almost unobserserable
– The random attack is easy to be detected • Real – no attack
• Estimated – ICA based attack
IEEE GLOBECOM'11
148
Message authentication code aggregation
• Message Authentication Code (MAC) is used to authenticate each message [Kolesnikov_2011] – To prevent en route accidental and malicious data
corruption
– Aggregate MAC is often used • Since the communication channel capacity is often small,
and
• The data size is short compared to the MAC code
– The aggregate MAC is not resilient to denial-of-service (DOS) attacks
• [Kolesnikov_2011] V. Kolesnikov, W. Lee, and J. Hong, “MAC Aggregation Resilient to DoS
Attacks”, Proceedings of IEEE SmartGridComm 2011.
IEEE GLOBECOM'11
150
Main contributions
• The authors proposed a new authentication mechanism for the wireless sensor data – Securely combine authentication tags computed by
sensors • So that the aggregate tag is much shorter than the
concatenation of the constituent tags, but
• Provides same strong security guarantees
– Resilient to denial-of-service (DOS) attacks • A DoS attacker will only be able to disrupt a portion of the
data – Only the data he relays
• His point of insertion can be estimated based on which part of aggregate MAC is corrupted.
IEEE GLOBECOM'11
151
Secure energy routing
• The authors of [Zhu_2011] developed a novel secure energy routing mechanism – for securely and optimally sharing renewable
energy in smart microgrids
– It can detects most internal attacks by using message redundancy
• Spoofed route signaling
• Fabricated routing messages
• [Zhu_2011] T. Zhu, S. Xiao, Y. Ping, D. Towsley, and W. Gong, “A Secure
Energy Routing Mechanism for Sharing Renewable Energy in Smart Microgrid”, Proceedings of IEEE SmartGridComm 2011.
IEEE GLOBECOM'11
152
Intrusion detection systems for home area
networks • This paper [Jokar _2011] presents a layered specification-based
intrusion detection system (IDS) – Designed to target ZigBee technology
– Addressed the physical and MAC layer • Normal behavior of the network is defined through selected specifications
extracted from the IEEE 802.15.4 standard
• Deviations from the defined normal behavior is viewed as a sign of malicious activities
• The performance analysis demonstrated that the designed IDS provides a good detection capability against known attacks – The same is expected for unknown attacks
• Since the design of the IDS is based on anomalous event detection
• [Jokar _2011] P. Jokar, H. Nicanfar, V. Leung, “Specification-based Intrusion Detection for Home
Area Networks in Smart Grids”, Proceedings of IEEE SmartGridComm 2011.
IEEE GLOBECOM'11
153
Privacy-preserving authentication
• Privacy requirement: to preserve the privacy of the consumers, the electric usage information is hidden from the substations [Chim_2011] – But it should be known by the control center
• Pseudo identity is used
• Authentication requirement on each smart meter – To ensure requests are sent from valid users
• The authentication process is made very efficient by means of Hash-based Message Authentication Code (HMAC) – The overhead is only 20 bytes per request message
• Under attack, the substation allows 6 times more valid messages to reach the control center – when compared to the case without any verification
• [Chim_2011] T. Chim, S. Yiu, L. Hui, and V. Li, “PASS: Privacy-preserving Authentication Scheme
for Smart Grid Network”, Proceedings of IEEE SmartGridComm 2011.
IEEE GLOBECOM'11
155
Privacy-utility tradeoff
• Existing privacy preservation solutions for user‟s electricity usage data have also not quantified the loss of benefit (utility) of data dissemination [Rajagopalan_2011]
• Using tools from information theory, a new framework is presented that abstracts both the privacy and the utility requirements of smart meter data.
• For a stationary Gaussian Markov model of the electricity load, it is shown that the optimal utility-and-privacy preserving solution requires filtering out frequency components that are low in power – this approach encompass most of the proposed privacy approaches
• [Rajagopalan_2011] S. Rajagopalan, L. Sankar, S. Mohajer, and V. Poor, “Smart Meter
Privacy: A Utility-Privacy Framework”, Proceedings of IEEE SmartGridComm 2011.
IEEE GLOBECOM'11
156
Cooperative state estimation for preserving privacy
• This paper [Kim_2011] presents a cooperative state estimation technique that protects the privacy of users‟ daily activities. – By exploiting the kernel of an electric grid configuration
matrix
– Obfuscate the privacy-prone data without compromising the performance of state estimation
• The power consumption measurement is well obfuscated such that the consumers do not fully disclose their private behavioral information in the first place, and
• the obfuscated data retain the necessary information such that the state vector can be accurately estimated from the perturbed measurement
• [Kim_2011] Y. Kim, E. Ngai, and M. Srivastava, “Cooperative State Estimation for Preserving Privacy of User Behaviors in Smart Grid”, Proceedings of IEEE SmartGridComm 2011.
IEEE GLOBECOM'11
157
Summary on Cyber Security for Smart Grid
• Different security constraints that makes securing smart grids a difficult
problem
• Several highly efficient adversaries
• Use existing protocols like IP with known vulnerabilities and work
around to using new protocols with unknown vulnerabilities
• Use of layered security architecture and attack tree‟s for efficient
security and risk assessment
IEEE GLOBECOM'11
158
Field Trials and Case Studies for Smart Grid
Communication Infrastructures Smart Power Grid
• SDC
Smart Renewable
• W2B
Smart Electricity Service
• S&C‟s CES
Smart Transportation
• PHEV/EV
Smart Consumer
• MDM
• MYPOWER
IEEE GLOBECOM'11
159
SmartGridCityTM – Boulder, Colorado
“The fundamental component for making the smart grid work will be a robust and dynamic communications network; providing the utility the ability for real-time, two-way communications throughout the grid and enabling interaction with each component from fuel source to end use” (Xcel Smart Grid White Paper)
Collaborating to Build the Next Generation Utility
160
Status of SGC
City - City of Boulder - 100,000 people, 50,000
homes
Smart Meters - 14,398 as of 1/28/09
Premises - 16,616 BPL enabled homes as of
1/28/09
Telecom Fiber - 120 miles planned by June 2009
Delivery Dates - build out complete by 6/30/2009
Systems - plug and play demand and generation
response (in process)
161
Demand Management
• Reduce spinning reserves
• Generation following (not demand response)
• Availability-based pricing
• Automated generation dispatch
Renewables Management
• Align demand to availability
• Manage intermittency
• Opt for type of energy use
• Supply-based pricing
Asset Management
• Improve field efficiency
• Real-time asset status & control
• Expanded reliability
• Extended asset life
Premise Management
• Automated device response control
• Real-time pricing (device-level)
• New services and products
• Enable customer choice
SMARTGRIDCITY – Key Values
162
SmartGridCity-Objectives
Xcel Objective Measurement CURRENT Smart Grid Impact
Improving Customer Satisfaction by
reducing customer minutes out of service
Reduce SAIDI by 10% Distribution Automation
Analysis & Reporting of:
Incipient transformer failure
Secondary neutral failure
Voltage exceptions
Transformer Overload
Underground remote fault detection
Outage notification & restoration
Empowering Customers to Reduce
Electricity Usage
Decrease usage by 2.5% 2-way thermostat control
Demand response portals
Meter consumption reporting
Reduce Service and Billing Expense,
Increase Revenue Assurance
Up to 50% annually Call center meter pings
Automated meter reading
Proactive maintenance (reduced O&M)
Decrease System Losses Reduce CO2 emissions up to 500,000 tons
annually
System Optimization
Conservation voltage reduction
Volt/Var Control
Phase Load Balancing
Asset Optimization Reduce capital investment and
distribution/substation maintenance up to $32
mil annually
Substation monitoring
Targeted asset replacement (system
reports)
Develop a Smart Grid City Consortium
Framework
Seamless integration of applications and
business process
Open GridTM Platform
Develop a Regulatory Framework to
Recover Smart Grid Investment
TBD Smart Grid Value Model
163
Smart Grid Operational Impact
Trees
8%
Xformer
25%
Xformer
Lead/Connection
10%
Secondary Brkr
Tripped
30%
Secondary
2%Secondary/Xformer
9%
Arrestor
1%
Xformer Tap
1%
Capacitor
2%
Sub LV Bus
Voltage
6%
Secondary Neutral
Connection
6%
Examples of items detected by a Smart Grid:
Smart Grid Solutions:
● 24x7 real-time distribution
network monitoring in use
● Dispatching work crews to repair
problems detected by CURRENT
Smart GridTM
● Underground fault detection
installed
● Successful distribution
automation switching trial
94% of the incidents detected avoided customer complaints
54% of the incidents detected avoided outages
SmartGridCityTM
Consortium
164
Smart Renewable Grid Balancing
Renewable Integration
Outage Support
Capital Cost Avoidance
Emissions Savings
Transmission Support
Firm Renewable Power Pricing
Graph from John P.
Benner, Manager, PV
Industry Partnerships,
National Renewable
Energy Laboratory,
303-384-6496
165
Wind 2 Battery (W2B) Project Description
• 1 MW NaS Battery
System • Can deliver 1 MW for 7 hrs
• Power Conditioning Equipment
• Wind farm/grid interconnection
• Local and remote data and
communication equipment
• Two Phases of Study • Understand how system could
optimize wind farm economies
• Understand how system could
optimize utility integration of
wind resources
166
COMMUNITY ENERGY STORAGE (CES)
Growth of Customer-Owned
DG (solar)
• Availability?
• Reliability?
• Safety?
• Dispatch?
“Net Zero” or “Near Zero”
Customers and Areas
• Own their generation (solar or
wind)
• Grid-Independent (with storage)
• Third-party storage service
could take them off the utility
grid
168
S&C‟S CES PROJECT-HARDWARE OVERVIEW
CES is a small distributed energy storage unit connected to the secondary of
transformers serving a few houses or small commercial loads
Key Parameters Value
Power (active and reactive) 25 kVA
Energy 25-75 kWh
Voltage - Secondary 240 / 120V
Battery - PHEV Li-Ion
Round Trip AC Energy
Efficiency
> 85%
25 KVA
169
S&C‟s CES Project-A “Virtual” Substation Battery
Communication and Control Layout for CES
CES Control Hub
Power Lines Communication and Control Links
CES CES CES CES
CES is Operated as a Fleet providing Multi-MW, Multi-hour Storage
Grid Benefits:
• Load Leveling at substation
• Power Factor Correction
• Ancillary services
Local Benefits:
• Backup power
• Voltage correction
• Renewable Integration
Integration
Platform
Utility Dispatch
Center/ SCADA
Substation
170
SpeedNet™ Radios:
A Leading Solution for Self-Healing Applications
Features Benefits
Self healing—peer-to-peer
mesh network
Reliable performance even if a
communication point is lost
Multi-level security Improved performance, less susceptible
to interference
Low latency High speed communications—shorter
restoration times
Assignable messaging
priority
Effectively serves both AMI backhaul
and DA applications
172
Smart Transportation
Electric Drive &
Electronic Components
Vehicle Stability
Control
Functional Safety
& durability of the FEV Communication
Architecture for Energy,
Communication &
thermal management,
Energy / Power Storage
Systems
Vehicle 2 Grid
Interface
Integration of the FEV in
cooperative transport
Infrastructure
173
Electric Drive Vehicles
• Until now, base growth of 1% per year for USA system – At 25% of US vehicle fleet is “only” 2% of total MW*hr (but
billions of $ in generation and distribution costs)
– On distribution a car‟s 6 KW connection for an average home‟s
peak usage of 3 KW is +200% & is very significant
http://www.ornl.gov/info/ornlreview/v41_1_08/regional_phev_analysis.pdf
174
2007 Xcel Energy / NREL PHEV Study
Scenarios Production Cost Capacity Cost Avoided Gasoline Emissions Distribution Impacts
Do Nothing Good Worse* Good Better Worse*
Delay to 10pm Better Best Good Good Best
Optimized to Off-peak Best Best Good Worse Best
Opportunity Charging Worse Worse* Best Best Worse*
• For any utility:
Time of charging matters…
Coincident peak loading matters…
Tailpipe versus upstream emissions matter…
* Could be mitigated with control technology / incentives
For Xcel Energy with night time coal base
load:
Smart Charge after 10 PM avoids Capital
Costs and Green House Gasses 175
2008 Xcel Energy / NREL PHEV Study
• 6 Converted Ford Escapes (3 fleet, 3 personal use) and driven 40 miles per day (as do 85% of US commuters) at $7500 / car
• Results (yet not statistically significant)
Used only top 1/3 of 25 mile battery pack (parallel hybrid) Averaged over 6 months, 56.84 MPG in a SUV at $0.03 vs $0.11* Extremely consistent availability (except Sunday post 5:00 PM) Plugged In MORE often over time (from 50% to 80% over 6 months) Availability to utility at 60% - 85% with all factors considered Infrastructure is EVERYWHERE - “power to the curb” is there but what is the “tipping point”?
* at $2.00 / gal gas for 18 MPG for 12,000 per year at with $0.08 / kW*hr
* payoff at $7,500 cost to implement is 93,750 miles or 7.8 years while GM’s Volt is expected to have 140 MPG or 3.2 year payoff
Photo by ASC Designs 303-522-0066
176
Impacts from PHEVs & EVs
Without SmartCharging:
130 new power plants needed with 25% PHEV/EV penetration (source: ORNL), but still 40% less emissions when “filled” with coal based generation
With SmartCharging:
Theoretically ZERO new power plants needed (source: ORNL) until 73% of total fleet with generation “valley fill”
With SmartCharging:
Reduce to 85% fewer car emissions by reducing total number of power plants (source: NREL, and being studied by Xcel Energy)
177
Meter Data Management (MDM)
Multiple data
sources
Accurate and timely
data
Secure data
storage
Create and disseminate information
• AMI
• Manual
Readings
• SCADA
• OMS
• MWF
• Other
• Validating,
Editing and
Estimating (for
hourly data)
• Standards and
rules for service
order creation
• Proactive
assurance of
data availability
• Audit trail
• Securely manages
1,000 times more
data/meter than CIS
or AMI systems can.
• Tags for weather,
demographic and
other operational
characteristics
• Manage and access
non-traditional meter
data, e.g., PQ, volts,
etc.
• Interface to billing
systems
• Interface for Customer
Service Reps
• Create TOU billing
summaries
• Provide summary data
• Support operation &
planning needs
• Platform for customer
web presentment
180
Combined data flow
Data input,
validation and
warehouse
AMI/AMR
Systems:
• RF
• PLC
• Drive-by
Other inputs:
• Handhelds
• SCADA
• Manual data
• Weather data
• ???
CIS Systems:
• NISC
• SEDC
• Daffron
• Others
Analytics:
• Revenue
Protection
• System loss
analysis
• Planning
• Cost of Service
• Others
Operational
Support for AMI:
• Business rules
• Service order
interpretation
Web
Presentment:
• Meter data
• Customer and
billing data
• Demographic or
other data MDMS
G&T:
• Data for M&V of load control
• Class level data from each EMC
• Demographic data for planning
• Other
181
MDM Vendors
SIEMENS
182
myPower Pricing Pilot Overview
Control Group myPower Sense myPower Connection
Customers 450 Residential 379 Residential 319 Residential
Rate* RS TOU-CPP (RSP) TOU-CPP (RSP)
Equipment
Electric interval meter Electric interval meter Electric interval meter
Programmable thermostat
Two-way communications
infrastructure - PLC, RF, Hybrid
Customer Education
and Communication
N/A Mail
Telephone
Telephone
Signal to thermostat
Usage and Billing
Information
N/A Internet Internet
* RS = Residential Service, TOU-CPP = Time-of-Use, Critical Peak Pricing
183
myPower Time-of-Use – Critical Peak Pricing
(TOU-CPP)Summer 2007 Pricing Plan Weekdays
June - September
0
4
8
12
16
20
24
28
9 AM 1 PM
Time of Day
Pri
ce
in
ce
nts
pe
r kW
h
9 AM
8.7¢Medium
Price
(Base
Price)
23.7 ¢High
Price
(On-
Peak)
8.7¢Medium
Price
(Base
Price) 3.7¢
Low Price
(Night Discount)
6P M 10 PM
$1.46Critical
Price
Standard Residential Rate
WeekendsJune - September
0
4
8
12
16
20
24
28
9 AM
Time of Day
Pri
ce
in
ce
nts
pe
r k
Wh
9 AM
8.7¢Medium
Price
(Base Price)
3.7¢ Low Price
(Night Discount)
10 PM
Standard Residential Rate
184
myPower Sense Customers
Time-of-Use and Critical Peak Impacts
Source: myPower Pricing Pilot results based on 2006 and 2007 data through September 30, 2007
Customers who received no in-home technology were able to
reduce On-Peak period demand on critical peak days by up to
20%, even if they do not have Central AC.
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hour Ending
Average kW
per Customer
CPP
TOU
Baseline
Night Base On-Peak Base
With Central AC on Summer Peak Days
0.0
0.5
1.0
1.5
2.0
2.5
3.0
3.5
4.0
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24
Hour Ending
Average kW
per Customer
CPP
TOU
Baseline
Night Base On-Peak Base
Without Central AC on Summer Peak Days
185
myPower Connection and myPower Sense Customers
Summer Period Energy Savings Estimates
• Both the myPower participant and the Control Group customers showed increases in summer usage compared to prior years
• The increase in usage in the myPower participants‟ segments was significantly smaller than the Control Group.
• An overall energy savings estimate is developed by examining the difference between the Control Group‟s and participant groups‟ increase in energy use.
Source: myPower Pricing Pilot results based on 2006 and 2007 data through September 30, 2007
Customers who participated in myPower achieved summer period
energy savings in the range of 3-4%.
Variable
Control
Group
Change in
Use
Participant
Group
Change in
Use
Summer
Energy Savings
from TOU
(Percent)
Total Summer
Energy Savings
from TOU
(kWh per Cust)
myPower Connection 5.2% - 1.9% = 3.3% 139
myPower Sense with
Central AC 5.2% - 1.5% = 3.7% 144
myPower Sense without
Central AC 6.4% - 2.1% = 4.3% 127
186
Prototype on WSN for line monitoring
• Use a hierarchical communication topology [Casey_2011] – Avoid single point failure of sensors that is common in a multi-
hop sensor network
• Main characters of the implemented system – Sensors
• Self –Configurable
• Remote-controllable
• Able to adjust the data sampling frequency automatically – E.g. Increase the sampling frequency from 10 minutes to 5 seconds when a
fault is detected
– Gateway • Does not forward sensor packets until
– A full WLAN packet (about 18 sensor packets) has been accumulated, or
– Timeout happens
• [Casey_2011] P. Casey, N. Jaber, and K. Tepe, “Design and Implementation of a Cross-Platform Sensor Network
for Smart Grid Transmission Line Monitoring”, Proceedings IEEE SmartGridComm 2011.
187
Hardware and software implementation
• Hardware – Gateway
• Encompasses a ZigBee mote on Crossbow MIB510 programming board connected to a laptop
• ZigBee mote: Crossbow Micaz mote that utilizes the Chipcon CC2420 radio
• Linksys WUSB54GC as the WLAN interface
– Sensor node • A standalone ZigBee mote with a sensor board (Crossbow
MTS300CA)
• Software – TinyOS-2.x for the sensor
– Ubuntu 8.10 for the laptop (gateway)
188
Hierarchical communication topology
• Using ZigBee for communications between sensors and gateways
• Using 802.11 to build a mesh network among gateways
• Control center is the sink node
• The communication system for line monitoring is reliable since
– Both ZigBee and WLAN are reliable for this smart grid application
189
Field trial on PLC for smart meter applications
• PRIME (PoweRline Intelligent Metering Evolution) – A narrowband power line communications (PLC) technology
targeted for use in smart metering applications
– Use OFDM techniques and well-known forward error correction mechanisms, novel discovery and network-building MAC procedures
– Allow for cost-effective, seamless integration with recognized standard metering protocols such as DLMS/COSEM
– could become a globally recognized industry standard
• This paper [Berganza_2011] presents results obtained from real-field multi-vendor deployments with PRIME-compliant interoperable implementations at Iberdrola network in Spain.
• [Berganza_2011] I. Berganza, A. Sendin, A. Arzuaga, M. Sharma, and Badri Varadarajan, “PRIME on-field
deployment - First summary of results and discussion”, Proceedings IEEE SmartGridComm 2011.
190
Main lessons learned from the field trials
• Signal interference due to misconfigurations
– Two concentrators were deployed on the two
transformers in a same substation
• Beacons collide in the time domain
– Some service nodes are jumping between the two subnetworks
• Should only set one concentrator, and set others as
switches
• Unreliable communications when not all meters
governed by a substation are PRIME meters
– The signal-to-noise ratio might not be high enough
191
PLC communication for remote areas
• This paper [Kikkert_2011]describes an accurate SWER line model – Single Wire Earth Return (SWER) lines are used in Australia,
USA, South Africa and many other countries to provide power to remote communities
– The model demonstrate the severe signal channel degradation that can occur due to line branches and coupling networks
– The model is verified with measurements from two sites in Australia.
– Data rates are at 22.8 kbps on a 14 km SWER line • when the attenuator is set to less than or equal to 15 dB attenuation
– Predict that PLC communication systems using G3-PLC modems on SWER lines in excess of 2000 km are feasible
• [Kikkert_2011] C. Kikkert, “Effect of Couplers and Line Branches on PLC
Communication Channel Response”, Proceedings IEEE SmartGridComm 2011.
192
Device communications using SCADA systems
• Communications in traditional power grid are mainly enabled by a centralized supervisory control and data acquisition (SCADA) system
• In [Lu_2011], they establish a monitoring system for a Solid State Transformer (SST) in a micro smart grid - Green Hub – To verify that SCADA system can be used to support such an
application
– The one megawatt Green Hub system is a power electronics based power system in the FREEDM systems center at the North Carolina State University.
• It is established to demonstrate salient features and capabilities of the FREEDM system on renewable energy generation, distribution, storage and management
• [Lu_2011] X. Lu, W. Wang, A. Juneja, and A. Dean , “Talk to Transformers: An Empirical Study
of Device Communications for the FREEDM System”, Proceedings IEEE SmartGridComm 2011.
193
Implementation of SST monitoring system
• In the network domain, a control center is connected to the SST
controller
– via a Local Area Network (LAN)
• DNP3 is overlayed over TCP/IP in he implementation
– DNP3 (distributed network protocol 3.0) is a widely-adopted SCADA protocol
194
Conclusions and lessons
• Conclusion
– The DNP3 based SCADA system can be used in the smart grid
• for the device monitoring and control
• Lessons
– A careful optimization is crucial to reduce the total delay
• By optimizing every time-consuming part of every system component
• Delay is the primary concern for most smart grid applications
– The DNP3-based monitoring system is not suitable for more time stringent applications like relay protection
• The architecture is too complex and induce extra delay
195
Open Research Issues
• Cost-Aware Data Communication and Networking
Infrastructure
• Quality-of-Service (QoS) Framework
• Optimal Network Design
• Need of Secured Communication Network Infrastructure
• Plug-in Hybrid Electric Vehicle (PHEV)
196 IEEE GLOBECOM'11
Open Research Issues
Cost-Aware Data Communication and Networking
Infrastructure
• There is a cost in retrieving the real-time information (e.g.,
power pricing, metering data, and surveillance data), which
increases with the increase in frequency of inquiry
• However, the performances such as latency, bandwidth,
reliability must be met
• The cost optimization for data monitoring and transferred
must be performed
197 IEEE GLOBECOM'11
Open Research Issues
Quality-of-Service (QoS) Framework
• The QoS in smart grid can be defined by accuracy and
effectiveness with which different information such as
equipment‟s state, load information, and power pricing are
delivered timely to the respective parties
• QoS framework can be developed by identifying the
specific QoS requirements and priorities for specific
communication network in smart grid
198 IEEE GLOBECOM'11
Maximum Latency Communication Type ≤ 4 ms Protective relaying Sub-seconds Wide area situational awareness monitoring Seconds Substation and feeder supervisory control and data acquisition (SCADA) Minutes Monitoring noncritical equipment and marketing pricing information Hours Meter reading and longer-term pricing information Days/Weeks/Months Collecting long-term usage data
Open Research Issues
Optimal Network Design
• Dedicated network can be built to support the QoS- and
security-sensitive smart grid applications (e.g., status
monitoring and time-of-use report)
• Optimal network devices, their connections, and protocols
have to be chosen to avoid congestion and failure
• Shared network (e.g., cellular service) can be used to
support noncritical smart grid applications (e.g., billing)
• Integration of dedicated and shared network can be explored
199 IEEE GLOBECOM'11
Open Research Issues
Need of Secured Communication Network Infrastructure
• If smart grid is attacked, the hackers can penetrate the
network and alter critical system parameters which could
destabilize the grid in an unpredictable way causing
nationwide crisis
• Intrusion detection and prevention for smart grid (e.g., AMI
and WAMS)
• Public key infrastructure (PKI) for smart grid
200 IEEE GLOBECOM'11
Open Research Issues
Plug-in Hybrid Electric Vehicle (PHEV)
• With the use of electric power, PHEV has lower operational
cost and smaller emission of CO2
• PHEV requires electric charging from charging station
• To ensure stabilized load, electric power has to be supplied
according to the demand from PHEV
• Communications intrastructure for PHEV charging can be
proposed (e.g., [Erol-Kantarci 2011])
– Utility company communicates with substation control center (SCC)
using WiMAX and charging station using wireless mesh network
– SCC decides to accept or refuse the charging request from PHEV
201 IEEE GLOBECOM'11
Conclusion
• Smart grid will be a crucial technology to improve the
efficiency of the power grid
• There are many issues related to data communications and
networking
202 IEEE GLOBECOM'11
References (1) [Niyato_2011] D. Niyato, L. Xiao, and P. Wang, "Machine-to-machine communications for home energy management system in smart
grid," IEEE Communications Magazine, vol. 49, no. 4, pp. 53-59, April 2011.
[Bruce_1998] A. G. Bruce, “Reliability analysis of electric utility SCADA systems,” IEEE Transactions on Power Systems, vol. 13, no. 3,
pp. 844-849, August 1998.
[Xie_2002] Z. Xie, G. Manimaran, V. Vittal, A. G. Phadke, and V. Centeno, “An information architecture for future power systems and its
reliability analysis,” IEEE Transactions on Power Systems, vol. 17, no. 3, pp. 857- 863, August 2002.
[Wang_2010] Y. Wang, W. Li, and J. Lu, “Reliability analysis of wide-area measurement system,” IEEE Transactions on Power Delivery,
vol. 25, no. 3, pp. 1483-1491, July 2010.
[Erol-Kantarci 2011] M. Erol-Kantarci, J. H. Sarker, and H. T. Mouftah, "Communication-based Plug-In Hybrid Electrical Vehicle load
management in the smart grid," IEEE Symposium on Computers and Communications (ISCC), pp. 404-409, June 2011-July 2011.
[Liu_2010] E. Liu, M. L. Chan, C. W. Huang, N. C. Wang, and C. N. Lu, "Electricity grid operation and planning related benefits of
advanced metering infrastructure," International Conference on Critical Infrastructure (CRIS), pp.1-5, September 2010.
[Mao_011] R. Mao and V. Julka, "Wireless Broadband Architecture Supporting Advanced Metering Infrastructure," IEEE Vehicular
Technology Conference (VTC Spring), pp.1-13, May 2011.
[Chen_2010] S. Chen, J. Lukkien, and L. Zhang, "Service-oriented Advanced Metering Infrastructure for Smart Grids," Asia-Pacific
Power and Energy Engineering Conference (APPEEC), pp.1-4, March 2010.
[Yu_2011] R. Yu, Y. Zhang, S. Gjessing, C. Yuen, S. Xie, and M. Guizani, "Cognitive radio based hierarchical communications
infrastructure for smart grid," IEEE Network, vol. 25, no. 5, pp. 6-14, September-October 2011.
[Wang_2007] Y. Wang, J. P. Lynch, and K. H. Law, “A wireless structural health monitoring system with multithreaded sensing devices:
Design and validation,” in Structure and Infrastructure Engineering, vol. 3(2), pp. 103-120, 2007.
[Ullo_2010] S. Ullo, A. Vaccaro, and G. Velotto, “The role of pervasive and cooperative sensor networks in smart grids communication,”
in Proceedings of IEEE Mediterranean Electrotechnical Conference (MELECON), pp. 443-447, April 2010.
[Yan_2011] Y. Yan, Y. Qian, and H. Sharif, “A secure and reliable in-network collaborative communication scheme for advanced
metering infrastructure in smart grid,” in Proceedings of IEEE Wireless Communications and Networking Conference (WCNC), pp.
909-914, March 2011.
[Sreesha_2011] A. A. Sreesha, S. Somal, and I-Tai Lu, “Cognitive radio based wireless sensor network architecture for smart grid utility,”
in Proceedings of IEEE Long Island Systems, Applications and Technology Conference (LISAT), May 2011.
203 IEEE GLOBECOM'11
References (2) [Hung_2010] K. S. Hung, W. K. Lee, V. O. K. Li, K. S. Lui, P. W. T. Pong, K. K. Y. Wong, G. H. Yang, and J. Zhong, "On Wireless
Sensors Communication for Overhead Transmission Line Monitoring in Power Delivery Systems," IEEE International Conference
on Smart Grid Communications (SmartGridComm), pp. 309-314, October 2010.
[Naduvathuparambil_2002] B. Naduvathuparambil, M. C. Valenti, and A. Feliachi, "Communication delays in wide area measurement
systems," Proceedings of the Thirty-Fourth Southeastern Symposium on System Theory, pp. 118- 122, 2002.
[Shahraeini_2011] M. Shahraeini, M. H. Javidi, and M. S. Ghazizadeh, "Comparison Between Communication Infrastructures of
Centralized and Decentralized Wide Area Measurement Systems," IEEE Transactions on Smart Grid, vol. 2, no. 1, pp. 206-211,
March 2011.
[Mohsenian-Rad_2010] A. Mohsenian-Rad, V. W. S. Wong, J. Jatskevich, R. Schober, and A. Leon-Garcia, "Autonomous Demand-Side
Management Based on Game-Theoretic Energy Consumption Scheduling for the Future Smart Grid,” IEEE Transactions on Smart
Grid, vol. 1, no. 3, pp. 320-331, December 2010.
[Hirst_1998] E. Hirst and B. Kirby, “Unbundling Generation and Transmission Services for Competitive Electricity Markets: Ancillary
Services”, NRRI-98-05, National Regulatory Research Institute, Columbus, OH, Jan. 1998
[Giani_2011] A. Giani, E. Bitary, M. Garciay, M. McQueenz, P. Khargonekarx, and K. Poolla, “Smart Grid Data Integrity Attacks:
Characterizations and Countermeasures”, Proceedings of IEEE SmartGridComm 2011.
[Vukovic_2011] O. Vukovic, K-C Sou, G. Dan, and H. Sandberg, “Network-layer Protection Schemes against Stealth Attacks on State
Estimators in Power Systems”, Proceedings of IEEE SmartGridComm 2011.
[Tajer_2011] A. Tajer, S. Kar, V. Poor, and S. Cui, “Distributed Joint Cyber Attack Detection and State Recovery in Smart Grids”,
Proceedings of IEEE Globecom 2011.
[Esmalifalak_2011] M. Esmalifalak, H. Nguyen, R. Zheng, and Z. Han, “Stealth False Data Injection using Independent Component
Analysis in Smart Grid”, Proceedings of IEEE SmartGridComm 2011.
[Kolesnikov_2011] V. Kolesnikov, W. Lee, and J. Hong, “MAC Aggregation Resilient to DoS Attacks”, Proceedings of IEEE
SmartGridComm 2011.
[Zhu_2011] T. Zhu, S. Xiao, Y. Ping, D. Towsley, and W. Gong, “A Secure Energy Routing Mechanism for Sharing Renewable Energy
in Smart Microgrid”, Proceedings of IEEE SmartGridComm 2011.
[Jokar _2011] P. Jokar, H. Nicanfar, V. Leung, “Specification-based Intrusion Detection for Home Area Networks in Smart Grids”,
Proceedings of IEEE SmartGridComm 2011.
204 IEEE GLOBECOM'11
References (3) [Chim_2011] T. Chim, S. Yiu, L. Hui, and V. Li, “PASS: Privacy-preserving Authentication Scheme for Smart Grid Network”,
Proceedings of IEEE SmartGridComm 2011.
[Rajagopalan_2011] S. Rajagopalan, L. Sankar, S. Mohajer, and V. Poor, “Smart Meter Privacy: A Utility-Privacy Framework”,
Proceedings of IEEE SmartGridComm 2011.
[Kim_2011] Y. Kim, E. Ngai, and M. Srivastava, “Cooperative State Estimation for Preserving Privacy of User Behaviors in Smart Grid”,
Proceedings of IEEE SmartGridComm 2011.
[Casey_2011] P. Casey, N. Jaber, and K. Tepe, “Design and Implementation of a Cross-Platform Sensor Network for Smart Grid
Transmission Line Monitoring”, Proceedings IEEE SmartGridComm 2011.
[Berganza_2011] I. Berganza, A. Sendin, A. Arzuaga, M. Sharma, and Badri Varadarajan, “PRIME on-field deployment - First summary
of results and discussion”, Proceedings IEEE SmartGridComm 2011.
[Kikkert_2011] C. Kikkert, “Effect of Couplers and Line Branches on PLC Communication Channel Response”, Proceedings IEEE
SmartGridComm 2011.
[Lu_2011] X. Lu, W. Wang, A. Juneja, and A. Dean , “Talk to Transformers: An Empirical Study of Device Communications for the
FREEDM System”, Proceedings IEEE SmartGridComm 2011.
205 IEEE GLOBECOM'11