- Email: [email protected]
Wireless multimedia sensor and actor networks for the next generation power grid
Ad Hoc Networks 9 (2011) 542–551
Contents lists available at ScienceDirect
Ad Hoc Networks journal homepage: www.elsevier.com/locate/adhoc
Wireless multimedia sensor and actor networks for the next generation power grid Melike Erol-Kantarci, Hussein T. Mouftah ⇑ School of Information Technology and Engineering, University of Ottawa, 800 King Edward Avenue, Ottawa, ON, Canada
a r t i c l e
i n f o
Article history: Available online 10 August 2010 Keywords: Electrical power grid Smart grid Wireless multimedia sensor and actor network
a b s t r a c t Electrical power grid is undergoing a major renovation, to meet the power quality and power availability demands of the 21st century. The new power grid, which is also called as the smart grid, aims to integrate the recent technological advancements in the Information and Communication Technology (ICT) field to the power engineering field. The present smart grid implementations focus on smart meter based utility-to-meter and utility-tocustomer communications. Although these features provide significant improvements on the customer management side, in the following decades, grid management will be one of the major ICT-dominant fields. Recently, adoption of Wireless Sensor Networks (WSN) for the power grid is gaining wide attention from the industry and the academia. Scalar sensor measurements bring valuable insights, however they can provide limited set of information. In the next generation power grid, limited-sensing, Supervisory Control and Data Acquisition (SCADA) based, centrally controlled operational architecture will be replaced with wireless connected, low-cost, multimedia sensors combined with distributed decision-making and acting systems, working in coordination with a supervisory software tool. In this paper, we discuss the potential applications and the challenges of employing wireless multimedia sensor and actor network (WMSAN) for the smart grid. Ó 2010 Elsevier B.V. All rights reserved.
1. Introduction Electrical power grid is among the high priority critical infrastructures of a nation. In last decades, power grid operators have been experiencing problems due to the imbalance between growing demand and diminishing fossil fuels, coupled with aging equipments and workforce. The resilience of the power grid has become questionable especially after the major blackouts in California and Northeast of the United States, in 2001 and 2003, respectively. The lack of pervasive and effective communications, automation, monitoring and diagnostic tools, has resulted in the failure of the power grid. Consequently, the opportunities that become available with the advances in Informa⇑ Corresponding author. Tel.: +1 613 562 5800; fax: +1 613 562 5664. E-mail addresses: [email protected] (M. Erol-Kantarci), [email protected] (H.T. Mouftah). 1570-8705/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.adhoc.2010.08.005
tion and Communications Technology (ICT) have necessitated the modernization of the existing power grid. Resilience and self-healing combined with distributed renewable energy generation, distributed storage, reduced Green House Gas (GHG) emissions, sophisticated energy management at the demand-side, easy adaptability to future systems, two-way flow of information and electricity are among the desired properties of the smart grid [1]. Recently, smart meters have been installed in the majority of consumer premises in North America to provide communication between utilities and consumers. While this is expected to increase the efficiency of the power grid, the opportunities that rise from the advances in Micro-Electro-Mechanical Systems (MEMS) are not fully exploited yet. Adoption of WSN in the smart grid is an attractive research topic. WSNs can be deployed rapidly and self-organize to form an intelligent monitoring platform [2]. For instance, low-cost sensor nodes that use wireless
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551
communications, can be spread over the field of operation of a utility and enhance utility asset monitoring capabilities. Although WSNs are decent tools for monitoring, in the following years, the next generation power grid technologies will employ WMSANs to enable distributed and autonomous operation, augmented with fine-grained monitoring capabilities. WMSAN is a group of distributed sensor and actor platforms that are connected via wireless communications where multimedia sensors record video, capture audio and still images, collect ambient scalar data, and actors are capable of operating the devices. WMSAN have been considered to be used in surveillance, traffic control, health care, elderly assistance, person locator and industrial process control applications previously [3]. In the smart grid, distributed power plants, overhead and underground power lines, transmission towers, substations, industrial consumer premises, commercial/residential buildings and homes, will require low-cost, long-lasting, distributed systems that can sense, communicate and act, i.e. WMSANs. In the traditional power grid, monitoring equipments such as, power donuts, weather stations, sagometers which generally communicate through wired communications, have been employed at several critical locations. However, wireline communications have high cabling costs especially due to the geographical spread of the utility assets. Moreover, the next generation power grid will need more than a monitoring system since the large-scale and real-time nature of the grid demand quick repose to a large number of incidents occurring at the same time and related with each other. As the decision process and the control algorithms mitigate from human operators and the centralized architecture toward distributed intelligent devices, the capabilities of the power grid will increase. An autonomous grid that can dynamically react to changing ambient conditions, failures and consumer demands can maximize the capacity of energy generation, power transmission and distribution. Furthermore, collecting multimedia content could highly improve the safety and security of the grid. As low-memory image processing algorithms [4] and low-cost hardware that ubiquitously collect multimedia data, such as CMOS cameras and microphones [3,5] become available, the widespread adoption of WMSANs in the smart grid will be possible. Despite the opportunities that become available with WMSANs, WMSANs have several challenges to be addressed. They suffer from low link quality, high latency, limited bandwidth and limited battery, as well as, lack of established techniques for in-network processing, efficient video streaming, QoS provisioning, distributed decision and control algorithms. Moreover, there are challenges specific to the power grid. Power grid is a large-scale system, in the sense of geographical region and the number of equipments. A single malfunctioning device may cause partial (isolated) outage, or it may cause overall system failure and affect a larger region and higher number of consumers. Similarly, under utilization of resources at a single transmission line or a generation unit may impact the overall system performance. In addition, the operation and maintenance of the grid require strict safety precau-
543
tions for personnel and public health, therefore WMSANs should not cause hazardous situations. In this paper, we discuss the potential applications of WMSANs in the next generation power grid. We explore new applications for energy generation facilities, transmission and distribution (T&D) equipments and consumer premises, while discussing the related challenges. In Section 2 we present the state-of-the-art in power grid monitoring tools and introduce possible WMSAN applications. We discuss the challenges of employing WMSANs in the power grid and point out the open issues in Section 3. Finally, we conclude the paper in Section 4. 2. Multimedia sensors for the power grid In today’s power grid, sensing equipments are employed in a limited number of critical assets such as the power transmission lines and substations. Generally, these sensing equipments are not interconnected, i.e., they transmit the collected data to a central location via wireline communications, where a supervisory control system and human operators at the utility headquarters manage grid operations. The power grid can be conceptually divided into three segments to represent the flow of electricity from a generation facility to the premises where it is consumed. These segments are, energy generation segment (supplier), T&D segment (utility) and consumption segment (demandside). State-of-the-art sensing equipments are mostly employed in the T&D segment and less frequently in energy generation facilities. In the next generation power grid, it is necessary to have sensors in the generation plants and consumer premises, as well. In this section, we present possible WMSAN applications in the next generation smart grid while briefly summarizing the sensing equipments in use today. 2.1. WMSANs for energy generation facilities In the traditional grid, energy is generated at power plants which use resources such as nuclear fission, hydro, or fossil fuels (e.g. coal, gas, diesel, natural gas). In several countries, wind tribunes and Photovoltaic (PV) panels also contribute to energy generation. Wind and solar power are called as renewable or green generation methods and they are preferred more than the fossil fuel based energy generation techniques due to their lower cost and lower GHG emissions. However, their availability is limited to the geographical conditions of the grid and the time of the year. For example, solar power is not available at night and wind power varies seasonally. The intermittent nature of these generation techniques makes it hard to employ them as primary power supplies for a grid unless distributed energy storage is available. In the traditional grid energy storage methods such as pumped hydro, compressed air and flywheel are available however they are not convenient for storing the energy generated at the solar and the wind farms. Energy storage in the smart grid is one the active research topics. The amount of energy to be stored depends on the amount of
544
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551
energy generation and the demand. Therefore, in a power plant or in a wind/solar farm, real-time generator monitoring, provides insights for storage decisions. Monitoring and controlling the distributed energy generation and storage requires low-cost, distributed sensor and actor systems. WMSANs offer promising solutions in this field. They have large coverage, and they provide fault tolerance due to dense deployment. They can be rapidly deployed and they are able to self-organize when new nodes are added, or some nodes become unavailable. In the following subsections, we explain the opportunities where WMSANs can be used in energy generation facilities. 2.1.1. Wind farms Wind farms consist of wind tribunes that transform the force of wind to electrical power. For a wind tribune, temperature, pressure, humidity and wind orientation are critical parameters affecting the power output [6]. These parameters can be derived from scalar sensor measurements. On the other hand, bird collisions with wind tribunes is a significant problem affecting the healthy operation of the tribune. A recent study proposes to use acoustic emission sensors, i.e. microphones, which are mounted on the wind tribune tower to detect bird collisions [7]. Previously, accelerometers and fiber-optic sensors have also been considered, however they have been found less efficient since they need to be installed on each of the blade of the wind tribune, and rotor blade vibrations affect the ability of accelerometers to detect collisions. On the other hand, the drawback of acoustic sensors is that, they are affected by the turbine operating noise and the ambient noise. In this case, WMSANs can be employed to collect both audio and vision data. For instance, upon detection of a collision by the acoustic sensors, image sensors may wake-up to capture images of the blades and the tower to identify the faulty component, e.g. blade, pitch, gear box. Remote detection of the faulty component provides knowledge of required parts to be replaced, in advance, and speeds up the restoration of the wind tribune. Meanwhile, during the restoration, actors can shut down the faulty tribune, communicate with the utility headquarters to learn the amount of demand. For example, if the demand is high and more power is required, actors can bring backup tribunes online. 2.1.2. Solar farms For solar panels, temperature, radiation, DC voltage and current measuring scalar sensors can be used for predicting the energy output of the panel. In addition, the efficiency of solar panels decrease when solar beams are blocked by clouds. To increase the reliability of the prediction of energy output, WMSANs can monitor the sky and weather conditions, such as cloudiness. A simple illustration of solar and wind farms with WMSANs is given in Fig. 1. 2.1.3. Traditional power plants In the traditional nuclear, hydro or fossil fuel based power plants, temperature and pressure sensors are frequently used in addition to humidity, flow, vibration, volt-
Fig. 1. WMSANs for solar and wind farms with distributed storage.
age, current and motion sensors. These sensors generally communicate over wires, i.e., they require cabling. Recently, employing wireless sensors in power plants have been considered in several works. Eaton Inc. has developed a low-cost wireless motor monitoring system that continuously measures line voltage and current to evaluate motor energy used in a power plant [8]. In addition, US Department of Energy funded Industrial Technologies Program (ITP) also supports wireless sensor and automation technologies in power generation [9]. The ITP has proposed several wireless sensor application areas which are emission monitoring, flame sensing, turbine control, boiler control, motor control and monitoring the health of rotating assets [10]. These monitoring tasks can be more accurately established with WMSANs than WSNs. In addition to equipment monitoring, WMSANs can improve personnel safety. For example, under emergency conditions entrance of the personnel to hazardous areas can be remotely restricted, or the personnel could be routed to safety showers rapidly in a nuclear plant. 2.2. WMSANs for T&D facilities Transmission and distribution segment covers substations, overhead power lines and underground power lines. These are critical assets where an equipment failure or breakdown may cause blackouts or may be dangerous for public health. Moreover securing these assets is more challenging than power plants since they can be reached easily from outside. A recent study proposes to use WSNs for securing the power grid [11]. In this section, we investigate the further advantages of using WMSANs. T&D power lines of a utility covers a large geographical region with sometimes remote or hard-to-access components. Consequently, in the traditional grid, sensing equipments are deployed on a limited number conductors and towers, where the measured properties include sag, conductor strength, temperature, heating, galloping, icing, wind speed, and contact with vegetation and animals. The next generation power grid will cover the whole T&D segment by employing a large number of sensor and actor nodes to retrieve multimedia data and establish an autonomous grid.
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551
2.2.1. Power lines Overhead conductors heat and expand due to the ambient temperatures and the electricity current flowing through the conductor. Thermal expansion together with wind and adverse weather conditions, results in sagging. Sag is one of the significant factors limiting the ampacity, i.e., current capacity of the conductor. Legal and operational constraints force certain ground clearance values for overhead power lines to ensure public safety and healthy grid operation. The load in the power line can be increased as long as the conductor does not exceed the maximum allowable sag level. Therefore ampacity of a span, i.e. conductor section between two towers, is limited by the amount of sag. In the power grid, the span with the lowest ampacity determines the capacity of the T&D system. Therefore, utilities monitor the overhead power lines in real-time to avoid penalties or hazards due to sagging. Utilities are usually precautious, and they keep the ampacity lower than the actual possible value. This leads to inefficient operation of the T&D system. Integration of accurate monitoring tools will enable accurate ampacity estimation which will lead to efficient utilization of the power lines. Presently power line sag monitor, sagometer, Power Line Sensor (PLS) and Power Donut2 techniques are used for sag measurements. The power line sag monitor [12] uses an accelerometer to measure the inclination angle, i.e. the angle between the conductor and the horizontal plane. Acceloremeter transmits its readings to a remote processor where the sag is computed. The sagometer uses a video camera which monitors the movement of a passive target mounted on a conductor [13]. The sagometer communicates with the SCADA system using either spreadspectrum radio, fiber optics, satellite or cellular communications. PLS is able to take photos of clearance to ground in addition to line temperature, ambient temperature, wind measurements [14]. Power Donut2 [15] measures sagging as well as current, line to ground voltage, conductor temperature, current and voltage waveform. Power Donut2 transmits its data to a specified server via GSM/GPRS communications. Power line sag monitor, sagometer, PLS and Power Donut2 are relatively expensive tools therefore they are used in a limited number of transmission lines. However, by the use of low-cost multimedia sensors, sag measurement for all overhead power lines can become feasible. Moreover, multiple video sensors can provide larger field of view and more field of regard than a single Pan-TiltZoom (PTZ) camera that is available in the present systems. In today’s power grid, technology to avoid sagging is also available but these are costly to apply to all of the T&D power lines. Sagging Line Mitigator (SLiM) and Aluminium Conductor Composite Core (ACCC) are recently developed techniques to reduce sag. SLiM is a small device that can extend and contract passively and it is attached to the conductor to reduce sag as the heat and load increase [16]. SLiM is designed to retract upon heating and extract upon cooling therefore it can adjust the effective conductor length and prevent sagging. ACCC uses a composite material with less sagging property [17]. Other methods to reduce sag include raising towers or recabling the span, which are both costly. In the next generation smart grid, WMSANs can be employed to detect and correct sagging more efficiently.
545
In the T&D segment there are also various sensors to measure the conductor temperature profile, wind and other weather conditions or leakage current [18]. These are generally data sensors. Video sensors may be employed in detecting the wearing of the conductors or towers. Moreover, contact with vegetation or animals may cause safety issues, video sensors can also monitor the environment, and in case of danger they can activate circuit breakers and reroute the current over another span. In addition, wireless underground sensors can be used for monitoring the underground cables. 2.2.2. Substations Substations have been considered as the last mile of the traditional power grid. The terminal units communicate the measured critical parameters either via power line communication, satellite communication, optical fiber communication or wireless communications [19,20]. Using WSNs at the substations have been explored in [19] however WMSANs can enable a larger set of applications in the next generation power grid. For instance, islanding and safety of personnel can be better achieved via WMSANs. In the conventional grid, when the distribution system experiences a fault, a circuit breaker isolates the specific region easily. However in the smart grid, distributed generation units may still be feeding the section [21]. Although this is obviously advantageous to prevent blackouts, from another point of view, lack of communications may cause hazards when intended islanding is required. In this case, sensors and actuators can coordinate distributed generation units of the smart grid. When a utility personnel is detected by the WMSAN who is close to an energized component an alarm can be raised. This could reduce the risk of safety hazards to utility workers when power lines could be still energized in cases where they are assumed to be disconnected. WMSANs are also powerful tools for surveillance. Power lines and substations may be vulnerable to physical attacks and terrorism. In this case, video sensors can be triggered with the detection of an anomaly by the data sensors [22]. 2.3. Sensors on consumer premises For the traditional power grid, substations have been considered as the last mile and the extension of the grid into the consumer premises has not been implemented. This is due to the large amount of consumers which is unmanageable in the traditional grid. However, with the advances in the smart grid technologies, it will become possible to communicate, monitor and possibly control the power consumption of the consumers without disturbing their business or comfort. Consumers can be roughly classified into three groups based on their power needs and usage patterns. These are industrial, commercial and residential consumers. Although, in the traditional power grid, there are various sensing equipments for all of the three categories of consumers, usually these equipments are not interconnected and they have simple control principles. 2.3.1. Industrial consumers Presently, sensors are used to ensure safe and healthy operation of the equipments. They collect measurements
546
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551
such as temperature and pressure, and they are capable of rising alarms when the measured entities are above predefined thresholds. Advanced sensors are used in the control of robotic arms, e.g. in automotive industry. In addition, vision sensors are used to verify product quality in a wide range of industrial applications. Generally, the sensing equipment is programmed or trained for a correct data set and the primary focus of the quality control applications is to determine if a new product matches the defined standards. These sensors work individually, they are sometimes supervised by human operators, and they do not have wireless communication capabilities. Moreover, they are not interconnected with the power grid, and they do not have power management abilities. When WMSANs are used, they can play a significant role in the industrial process control, as well as the automated demand respond for industrial consumers. For the traditional grid, demand response programs are available for large-scale consumers. Industrial consumers such as plants or commercial consumers such as corporate buildings, may opt in the demand response program which is usually handled by an aggregator company. The primary focus of the demand response is to reduce peak hour consumption. In the grid, managing peak hour consumption is important for healthy grid operation and it is also one of the factors affecting the cost of energy. In deregulated electricity markets, independent system operators use load forecasting tools to predict the demand for the next-hour or the next-day [23]. Electricity suppliers, generate or purchase energy in matching amounts with the demand and they supply bids for the amount of energy they are able to dispatch. Usually the suppliers with lower bids are selected [24] which settles the price of electricity for a certain time interval. Regulation fees, taxes, transmission, distribution, maintenance costs are added to determine the final fixed price of electricity for the consumers. This means, although the price of electricity purchased by the utility varies in time, the price of electricity sold to the consumers is fixed. The reason for the varying cost of energy on the supplier side mostly depends on the energy generation method. Suppliers operate base power plants and peak power plants (peaker plants). Since energy cannot be stored in large amounts, generation and load have to be balanced. The supplier operates base plants which are able to accommodate an average load level. Generally, depending on the demand, they can be partially shut down and started again in a daily cycle. For example, during overnight hours, the demand is low, and less energy is generated, while in the morning, demand rises and more energy is generated by starting up the extra facilities. However, during the day, there may be several hours where the demand increases above the available capacity of the base plants. For example, the power demand of the HVACs will increase in hot summer days, especially in the afternoon. These hours are called as the peak hours. Extending the capacity of the base plants according to peak demand, e.g. building a new nuclear power plant which will be used only for several hours a day is costly. In most of the cases, peaker plants are brought online to accommodate the peak load. Peaker plants can respond to power demand in a short
time but they generally use fossil fuels which are more expensive. For this reason, the price of electricity rises during peak hours. If the base and peaker generation capacity become insufficient consumers may experience outage. In the traditional grid, aggregators have been managing the demand response by basically calling consumers to reduce their consumption at peak hours. For instance, in commercial buildings, HVACs are cycled off to avoid penalties or to receive credits. In a factory, equipment cooling can be cycled off during peak hours as long as the sensor readings are in safe operation margins. Recently, automated demand response has been implemented in California where demand response is managed by utility signals [25]. This could be further improved for a demand response program exploiting the dynamic pricing of the smart grid. The smart grid, aims to bill the consumers based on the time of consumption. This is called as Time of Use (ToU) pricing [26]. Furthermore, using price signals can also provide demand response for all of the consumers, including the residential consumers. In this case, WMSANs can be used in monitoring and managing energy consumption. During peak hours, scalar sensors can measure the electric current with several techniques such as inductor, Hall effect, GMR sensor and magnetic force on MEMS and video sensors can be used to monitor the unused equipments and actors can shut them off safely. 2.3.2. Commercial consumers Corporate buildings and shopping malls are the examples of commercial consumers. For these consumers, occupancy and temperature sensors are widely used. In addition, air quality measurement becomes important as HVACs can be cycled off for energy management [27]. Air quality is generally monitored by electrochemical sensors that measure CO2 level. Typical applications of WSN for the commercial buildings in smart grids are discussed in [28]. WMSANs have additional application fields, as well. WMSANs can be used in demand response for commercial buildings similar to industrial consumers. Moreover, the acoustic and video sensors can improve occupancy sensing for commercial consumers. WMSANs can be effectively used in parking lots of the commercial buildings and shopping malls for Plug-in Hybrid Electric Vehicle (PHEV) applications, as well. PHEVs are hybrid vehicles that use a mixture of gas and electricity. Electrical power is stored in a Lithium-Ion battery which can be charged from an electrical outlet. PHEVs are expected to be on the roads shortly, however their charging properties bare significant challenges for the power grid. Inevitably, PHEVs will increase the load on the grid and they may cause resilience problems if they are charged in an uncontrolled way. The impact of PHEV charging profile on the power grid has been explored in [29]. There is consensus on the need for controlled smart charging, yet standard technique is missing. Several smart charging proposals, including price prediction [30] and optimization [31] exists in literature. Besides regulating the time of charging, there are also emerging vehicle-to-grid (V2G) applications where PHEVs are considered for distributed power storage. The opportunities and challenges of PHEV-to-grid connections is out of the scope of this paper, however we outline the applica-
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551
tions where WMSANs can be used for PHEV charging and discharging decisions. In [32,33], PHEVs are proposed to be used for ancillary services. Ancillary services include spinning reserves and regulation. Spinning reserves are the idle sources that can dispatch power to the grid within 10 min of a request. These are the services which are generally paid to be immediately available and they do not generate power unless requested. Regulation services are generally under real-time control of the system operator and they are used for keeping the system frequency as close to 60 Hz as possible. A fleet of PHEVs can be contracted to provide the ancillary services. In this case, pricing of consumption and revenue from supporting the grid, should be considered in detail. Although permanent contracts with fleet operators is feasible, it does not provide flexibility for the individual PHEV owners. For this purpose, public charging stations located at parking lots can utilize WMSANs. Vision sensors can be used for identifying license plates and they can arrange billing accordingly. Similarly, electricity theft at the public stations could be prevented by the help of WMSANs. Moreover, multimedia sensors that are mounted on PHEVs can communicate with the other vehicles in a lot to share maps and traffic data. This information can be used to determine the duration of charging or the amount of energy available for discharging, by predicting the amount of battery power required to reach the next destination of the vehicle. For example, if the daily driving pattern of a driver includes a visit to a shopping mall, then to the school to pick up children and then driving home, PHEV can communicate with the other vehicles in the shopping mall, gather traffic information and construction information along the route of the driver and predict the required power to complete the trip. The charging duration or the discharging amount could be transmitted to the WMSAN of the parking lot where actors coordinate charging of the other vehicles.
2.3.3. Residential consumers Using WSN for energy management in homes, have recently been explored in [34,35]. In-home energy management has received attention from the academia, as well as the industry. In this course, smart home concept [36] has regained importance with a focus on energy efficiency [37]. For example, Intel has developed the Home Energy Dashboard concept to provide a simple interface to consumers to monitor their monthly bills and the electricity consumption of the individual appliances [38]. Besides providing savings for the consumers, energy management schemes can reduce the carbon footprint of the households [39]. In the smart grid, consumers are motivated to supply their energy from solar panels and small wind tribunes. Monitoring these equipments is important for the efficiency of the whole grid and the savings of the individual consumers where WMSANs can play a significant role. Real-time monitoring of solar panels and wind tribunes may increase equipment lifetime, increase the predictability of their output and report malfunctioning devices. An
547
Fig. 2. WMSAN application for residential consumers and PHEVs.
illustration of the WMSAN for the smart home with PHEV is given in Fig. 2. 3. Challenges of WMSAN applications in the next generation power grid Major challenges related with using WMSANs for the power grid, rise from the real-time operational restrictions, coupled with the physical channel handicaps of WMASNs; such as, low link quality, low data rate, high or variable latency. Additional challenges are related with processing, storing and transmitting the huge amount of data generated by the large-scale power grid. Multimedia sensors and actors are battery-limited, therefore energy efficiency is another challenge. Sensor nodes also have limited memory and processing capabilities which make video coding both challenging and necessary. A WMSAN may involve a combination of sensors that collect data, sound, motion and video to monitor a certain region, where multiple sensors possibly get correlated or redundant data. The correlation and redundancy can be advantageous when handled properly. In addition, transport protocols need to be able to provide QoS provisioning and they should incorporate priority mechanisms for applications that require strict real-time operation. 3.1. Link quality Wireless link quality in power plants, T&D system and the premises of industrial consumers is expected to be low. In literature, there are few works addressing this issue [19,40] where the authors focus on IEEE 802.15.4-based WSN operation in industrial conditions. Physical wireless channel models for communications around power lines and communications inside the substations, the impact of link quality on the connectivity and the topology of a WSN, the impact of connectivity on the real-time operation of the power grid are still open issues. The impact of harsh environmental conditions on sensors and communications is also unexplored. For outdoor sensor nodes, corrosive conditions, e.g. solar radiation, wind, rain, humidity or
548
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551
other weather factors, could affect the performance, whereas for indoor environments such as substations or power rooms; dust and heating could be potential problems. Additionally, low link quality may degrade the performance of WMSANs more than it does in WSN. In WMSAN, a single bit error may cause an entire sequence of frames to be discarded. For example, in video streaming, if an independently coded (I) frame is discarded due to channel errors, it may lead to the loss of an entire sequence [41]. The challenges raised by the low link quality can be addressed by transport protocols or cross-layer approaches which are still open issues. 3.2. QoS provisioning Time-varying and location-dependent link capacity and packet errors may cause some critical packets to be lost. In smart grid applications, it is essential to develop novel schemes to give priority to the packets of real-time applications. A cross-layer approach has been proposed in [42]. This study considers a substation scenario and employs Ultra Wide Band (UWB) communications which have high data rates within tens of meters. QoS provisioning for WMSANs that are employed in other grid assets, is still an open issue. 3.3. Latency Power systems generally have tight latency requirements. Most of the decisions are in real-time where delayed information may lead to less efficient operation or wrong decisions [19]. For this reason, variable latency has to be addressed by the transport protocols, as well as the decision and control algorithms. 3.4. Video streaming In the next generation power grid, cyber and physical security will have major importance. Surveillance of utility assets can be provided via video sensors of WMSANs. However, video streaming in WMSAN still needs to be improved in terms of throughput, fairness and encoding/ decoding techniques. Traditional video streaming principles are contradictory with sensor network architecture, as complex algorithms for encoding are present on power-constrained and resource-limited sensor nodes while decoding is done by more powerful nodes. There are several works in literature addressing this issue. In [41], the authors propose a compressed sensing technique to reduce the computational complexity at the encoder. Another technique combining compressed sensing and rate control algorithm has been proposed in [43], as well. Moreover, in a WMSAN, video sensors may simultaneously transmit streams where maximizing the throughput and fairness of each transmission become another challenging issue. In [44], the authors propose a rate control algorithm that uses RTT and SNR measurements and calculates the optimal ratio of video encoder rate to channel encoder rate. This algorithm can be applied to WMSANs for smart grids, to increase the performance of the real-time streaming,
especially when high quality video from different parts of the network are required. 3.5. In-network processing In-network processing will be an important part of the smart grid applications. It improves the performance of WMSANs by decreasing the amount of data to be transmitted to the sink and the centralized supervisory system, i.e. SCADA. Moreover, when it is combined with computer vision techniques, it can provide fast reacting opportunities for the actors. 3.6. Battery Wireless sensors inherently have energy limitations. In the power grid, some of the sensors will be deployed in locations that are hard to access and replace. For example, replacing sensors for underground cables is costly. For this reason, energy scavenging or energy harvesting becomes crucial [45]. In the power grid, there are numerous opportunities for energy harvesting, such as vibration, magnetic field, motion, solar, ambient light, air flow, temperature gradients, and pressure gradients. Although there are examples of sensors applying energy harvesting from the energized electric conductors, possible opportunities have not been fully exploited yet. 3.7. Decision and control algorithms The next generation power grid will be autonomous where actors will manage load balance, power switching, islanding and a large amount of the other grid management operations. Efficient grid management will be possible by distributed decision and control algorithms [46]. In a local operation field, actors will be applying the decisions of distributed control algorithms and at the same time communicating with the other actors on the power grid to synchronize a specific task. Today’s power grid is controlled centrally by human operators and the SCADA system. However, in the next generation power grid, supplier and consumer roles are redefined where consumers are able to sell power to grid, and the grid operators are able to manage the consumption. In this complex real-time system, decisions have to be taken rapidly. Developing robust decision and control algorithms for the power grid and novel tools to improve SCADA are essential. A recent tool, Dynamic Monitoring and Decision Systems (DYMONDS) aims to transform SCADA and develop a software tool with distributed decision and control abilities [47]. 3.8. Data management In the next generation power grid, the amount of data collected by audio, vision and scalar sensors will be huge. It is crucial to aggregate data since correlated or redundant data may be collected by sensors close to each other and sensors that have the same field of view. Moreover, employing an hierarchical architecture and allowing some portion of the data to be processed by lower tiers before it
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551
is unnecessarily transmitted to the utility headquarters may help in data management. 3.9. Large-scale WMSAN for the power grid, is inherently have the potential to become a very large-scale network. First, the geographical span of the power grid will be large. Second, the number of sensors and actors employed in each component will be dense. For a typical utility with 25,000 km of high voltage power lines, thousands of transformers, capacitors and breakers, it is estimated that monitoring only the T&D assets could require more than 100,000 sensors [18]. Management of such a large-scale network is challenging. 3.10. Safety Using sensors and actors close to the power grid components should be safe. Especially in power plants and substations, sensor nodes should not be capable of igniting materials, even under faulty conditions. For example, when gas leakage from an underground pipe occurs around an underground power line, which is being restored by actors, the motion of the actors may ignite the gas. Field tests under adverse conditions have to be established before WMSANs can be safely integrated to the power grid. 4. Conclusion Growing demand for energy, diminishing fossil fuels, efforts to reduce GHG emissions and resilience problems that showed up in the electrical power grid have led to common consensus on the need for a renovation of the power grid. The key to this renovation is the integration of ICT to enable addressing the imbalance between demand and supply, widespread adoption of renewable energy farms, monitoring utility assets and self-healing under failures. Presently, researchers in academia and industry are focusing on applying WSN technologies to the smart grid. In the next generation power grid, applications that go beyond WSN capabilities will be demanded. The scalar wireless sensors will be replaced by wirelessly connected multimedia sensors that are able to collect audio/video, and actors that are able to control the grid operations. WMSANs will enable fine-grained monitoring by multimedia sensors and distributed control of the grid by actors. In this paper, we have outlined the possible WMSAN applications for the power grid while discussing their challenges and pointing out the open issues to be addressed before they are widely adopted. WMSANs can be used in energy generation facilities such as nuclear, hydro, fossil fuel based power plants, wind farms and solar panels. Coordination of distributed power generation and storage is possible with fine-grained monitoring, as well as distributed control algorithms. Moreover, T&D equipments can be monitored and maintained by the WMSANs. The efficiency of the power grid and its ability to accommodate certain levels of load are restricted by safety measures of the overhead power lines. For this
549
reason, the energized power lines need to be monitored. In a WMSAN, vision sensors can monitor the sag and provide real-time estimates of load. WMSANs can be also used in the consumer premises (demand-side) to coordinate supply and consumption balance, and to improve demand response programs. Meanwhile, the use of WMSANs for the power grid introduces several challenges that need to be addressed. Link quality may be affected by weather, ambient conditions and the electromagnetic field of the electricity current. Low link quality may result in very low throughput especially for video streaming applications. Moreover, the power grid has strict real-time requirements where variable latency of the WMSANs may result in misconfigurations. QoS provisioning is another important issue to be addressed. Multimedia sensors have the potential of generating larger amount of data. Moreover, considering the large geographical region covered by a utility and the number of equipments present in the grid implementation, transmission of the collected raw data to the utility headquarters is not feasible. The reason is partly because the limited bandwidth and latency bounds and partly due to the limited sensor batteries. In-network processing techniques needs to be explored to handle this issue. To relax battery constraints energy scavenging techniques may be employed however energy scavenging in the power grid is not fully exploited. In the next generation smart grid, actors can play an active role in management and operation of the grid which requires robust distributed control algorithms. Finally, the sensor and actor nodes need to be safetyproved before they can be employed in the power grid. They should not be capable of ignition, overheating or yielding any other reaction that could cause hazardous situations. Acknowledgment This work was partially supported by an ORF-RE grant from the ministry of Research and Innovation of the Government of Ontario. References [1] S.M. Amin, B.F. Wollenberg, Toward a smart grid: power delivery for the 21st century, IEEE Power and Energy Magazine 3 (5) (2005) 34– 41. [2] I.F. Akyildiz, W. Su, Y. Sankarasubramaniam, E. Cayirci, Wireless sensor networks: a survey, Computer Networks 38 (4) (2002) 393– 422. [3] I.F. Akyildiz, T. Melodia, K.R. Chowdhury, A survey on wireless multimedia sensor networks, Computer Networks 51 (4) (2007) 921–960. [4] S. Rein, M. Reisslein, Low-memory wavelet transforms for wireless sensor networks: a tutorial, submitted for publication.. [5] I.F. Akyildiz, T. Melodia, K.R. Chowdhury, Wireless multimedia sensor networks: applications and testbeds, Proceedings of the IEEE 96 (10) (2008). [6] L. Shen, H. Wang, X. Duan, X. Li, Application of Wireless Sensor Networks in the Prediction of Wind Power Generation, Wireless Communications, Networking and Mobile Computing, Dallian, China, October 2008. [7] A. Pandley, J. Hermence, R. Harness, Development of a Cost-Effective System to Monitor Wind Turbines for Bird and Bat Collisions – Phase I: Sensor System Feasibility Study, Tech. Rep. CEC-500-2007-004,
550
[8]
[9] [10] [11]
[12] [13] [14] [15] [16]
[17] [18]
[19] [20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32] [33]
[34]
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551 January 2007.. B. Lu, T.G. Habetler, R.G. Harley, J.A. Gutierrez, Applying wireless sensor networks in industrial plant energy management systems. Part I. A closed-loop scheme, in: Proceedings of the IEEE Sensors, Irvine, CA, USA, October 30–November 3, 2005. US Department of Energy Funded Industrial Technologies Program. . R.A. Leon, V. Vittal, G. Manimaran, Application of sensor network for secure electric energy infrastructure, IEEE Transactions on Power Delivery 22 (2) (2007). R.M. Hayes, A. Nourai, Power Line Sag Monitor, US Patent 6 205 867, March 27, 2001. EDM International, Inc. . Power Line Sensor. . 3M. . Y. Yang, F. Lambert, D. Divan, A Survey on Technologies for Implementing Sensor Networks for Power Delivery Systems, Power Engineering Society General Meeting, 2007. V.C. Gungor, F.C. Lambert, A survey on communication networks for electric system automation, Computer Networks 50 (2006) 877–897. M. Qureshi, A. Raza, D. Kumar, S. Kim, U. Song, M. Park, H. Jang, H. Yang, B. Park, A survey of communication network paradigms for substation automation, IN: IEEE International Symposium on Power Line Communications and Its Applications, Jeju Island, South Korea, 2008, pp. 310–315. V.K. Sood, D. Fischer, J.M. Eklund, T. Brown, Developing a Communication Infrastructure for the Smart Grid, IN: IEEE Electrical Power & Energy Conference (EPEC), Montreal, Canada, October 2009. M. Satyajayant, M. Reisslein, G. Xue, A survey of multimedia streaming in wireless sensor networks, IEEE Communications Surveys and Tutorials 10 (4) (2008) 18–39. S.K. Aggarwal, L.M. Saini, A. Kumar, Electricity price forecasting in deregulated markets: a review and evaluation, International Journal of Electrical Power & Energy Systems 31 (1) (2009) 13–22. P. Bajpai, S.N. Singh, Bidding and gaming in electricity market: an overview and key issues, in: Proc of National Power System Conference (NPSC), Chennai, 2004, pp. 338–346. M.A. Piette, G. Ghatikar, S. Kiliccote, P. Palensky, C. McParland, E. Koch, D. Hennage, Open Automated Demand Response Communications Specification (Version 1.0), Lawrence Berkeley National Laboratory Technical April 2009.
[35]
[36]
[37]
[38] [39]
[40]
[41]
[42]
[43]
[44]
[45] [46] [47]
Computers and Communications (ISCC’10), Riccione, Italy, June 2010. M. Erol-Kantarci, H.T. Mouftah, TOU-aware energy management and wireless sensor networks for reducing peak load in smart grids, Green Wireless Communications and Networks Workshop (GreeNet) in IEEE VTC2010-Fall, September 6–9, 2010. M. Chana, D. Estevea, C. Escribaa, E. Campo, A review of smart homes – present state and future challenges, Elsevier Computer Methods and Programs in Biomedicine 91 (2008) 55–81. K. Kok, S. Karnouskos, D. Nestle, A. Dimeas, A. Weidlich, Cor Warmer, P. Strauss, B. Buchholz, S. Drenkards, N. Hatziargyriou, V. Lioliou, Smart houses for smart grid, in: 10th International Conference on Electricity Distribution CIRED, Prague, June 2009. Intel Home Energy Management Dashboard..
Melike Erol-Kantarci received her B.S. (2001), M.Sc. (2004) and Ph.D. (2009) degrees from Computer Engineering Department, Istanbul Technical University, Turkey. Currently, she is a postdoctoral fellow at School of Information Technology and Engineering, University of Ottawa. From September 2006 to August 2007, she has been at the Computer Science Department, UCLA, as a Fulbright visiting researcher. Her main research interests are residential energy management, smart grids, underwater sensor networks and wireless sensor networks.
Hussein T. Mouftah joined the School of Information Technology and Engineering (SITE) of the University of Ottawa in 2002 as a Tier 1 Canada Research Chair Professor, where he became a University Distinguished Professor in 2006. He has been with the ECE Dept. at Queen’s University (1979–2002), where he was prior to his departure a Full Professor and the Department Associate Head. He has six years of industrial experience mainly at Bell Northern Research of Ottawa (became Nortel Networks). He served IEEE ComSoc as Editorin-Chief of the IEEE Communications Magazine (1995–1997), Director of Magazines (1998–1999), Chair of the Awards Committee (2002–2003), Director of Education (2006–2007), and Member of the Board of Gover-
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551 nors (1997–1999 and 2006–2007). He has been a Distinguished Speaker of the IEEE Communications Society (2000–2007). Currently he serves IEEE Canada (Region 7) as Chair of the Awards and Recognition Committee. He is the author or coauthor of 6 books, 32 book chapters and more than 950 technical papers, 10 patents and 140 industrial reports. He is the joint holder of 12 Best Paper and/or Outstanding Paper Awards. He has received numerous prestigious awards, such as the 2008 ORION Leadership Award of Merit, the 2007 Royal Society of Canada Thomas W. Eadie Medal, the 2007–2008 University of Ottawa Award for Excellence in Research, the 2006 IEEE Canada McNaughton Gold Medal, the 2006 EIC
551
Julian Smith Medal, the 2004 IEEE ComSoc Edwin Howard Armstrong Achievement Award, the 2004 George S. Glinski Award for Excellence in Research of the U of O Faculty of Engineering, the 1989 Engineering Medal for Research and Development of the Association of Professional Engineers of Ontario (PEO), and the Ontario Distinguished Researcher Award of the Ontario Innovation Trust. Dr. Mouftah is a Fellow of the IEEE (1990), the Canadian Academy of Engineering (2003), the Engineering Institute of Canada (2005) and the Royal Society of Canada RSC: The Academy of Science (2008).
Contents lists available at ScienceDirect
Ad Hoc Networks journal homepage: www.elsevier.com/locate/adhoc
Wireless multimedia sensor and actor networks for the next generation power grid Melike Erol-Kantarci, Hussein T. Mouftah ⇑ School of Information Technology and Engineering, University of Ottawa, 800 King Edward Avenue, Ottawa, ON, Canada
a r t i c l e
i n f o
Article history: Available online 10 August 2010 Keywords: Electrical power grid Smart grid Wireless multimedia sensor and actor network
a b s t r a c t Electrical power grid is undergoing a major renovation, to meet the power quality and power availability demands of the 21st century. The new power grid, which is also called as the smart grid, aims to integrate the recent technological advancements in the Information and Communication Technology (ICT) field to the power engineering field. The present smart grid implementations focus on smart meter based utility-to-meter and utility-tocustomer communications. Although these features provide significant improvements on the customer management side, in the following decades, grid management will be one of the major ICT-dominant fields. Recently, adoption of Wireless Sensor Networks (WSN) for the power grid is gaining wide attention from the industry and the academia. Scalar sensor measurements bring valuable insights, however they can provide limited set of information. In the next generation power grid, limited-sensing, Supervisory Control and Data Acquisition (SCADA) based, centrally controlled operational architecture will be replaced with wireless connected, low-cost, multimedia sensors combined with distributed decision-making and acting systems, working in coordination with a supervisory software tool. In this paper, we discuss the potential applications and the challenges of employing wireless multimedia sensor and actor network (WMSAN) for the smart grid. Ó 2010 Elsevier B.V. All rights reserved.
1. Introduction Electrical power grid is among the high priority critical infrastructures of a nation. In last decades, power grid operators have been experiencing problems due to the imbalance between growing demand and diminishing fossil fuels, coupled with aging equipments and workforce. The resilience of the power grid has become questionable especially after the major blackouts in California and Northeast of the United States, in 2001 and 2003, respectively. The lack of pervasive and effective communications, automation, monitoring and diagnostic tools, has resulted in the failure of the power grid. Consequently, the opportunities that become available with the advances in Informa⇑ Corresponding author. Tel.: +1 613 562 5800; fax: +1 613 562 5664. E-mail addresses: [email protected] (M. Erol-Kantarci), [email protected] (H.T. Mouftah). 1570-8705/$ - see front matter Ó 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.adhoc.2010.08.005
tion and Communications Technology (ICT) have necessitated the modernization of the existing power grid. Resilience and self-healing combined with distributed renewable energy generation, distributed storage, reduced Green House Gas (GHG) emissions, sophisticated energy management at the demand-side, easy adaptability to future systems, two-way flow of information and electricity are among the desired properties of the smart grid [1]. Recently, smart meters have been installed in the majority of consumer premises in North America to provide communication between utilities and consumers. While this is expected to increase the efficiency of the power grid, the opportunities that rise from the advances in Micro-Electro-Mechanical Systems (MEMS) are not fully exploited yet. Adoption of WSN in the smart grid is an attractive research topic. WSNs can be deployed rapidly and self-organize to form an intelligent monitoring platform [2]. For instance, low-cost sensor nodes that use wireless
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551
communications, can be spread over the field of operation of a utility and enhance utility asset monitoring capabilities. Although WSNs are decent tools for monitoring, in the following years, the next generation power grid technologies will employ WMSANs to enable distributed and autonomous operation, augmented with fine-grained monitoring capabilities. WMSAN is a group of distributed sensor and actor platforms that are connected via wireless communications where multimedia sensors record video, capture audio and still images, collect ambient scalar data, and actors are capable of operating the devices. WMSAN have been considered to be used in surveillance, traffic control, health care, elderly assistance, person locator and industrial process control applications previously [3]. In the smart grid, distributed power plants, overhead and underground power lines, transmission towers, substations, industrial consumer premises, commercial/residential buildings and homes, will require low-cost, long-lasting, distributed systems that can sense, communicate and act, i.e. WMSANs. In the traditional power grid, monitoring equipments such as, power donuts, weather stations, sagometers which generally communicate through wired communications, have been employed at several critical locations. However, wireline communications have high cabling costs especially due to the geographical spread of the utility assets. Moreover, the next generation power grid will need more than a monitoring system since the large-scale and real-time nature of the grid demand quick repose to a large number of incidents occurring at the same time and related with each other. As the decision process and the control algorithms mitigate from human operators and the centralized architecture toward distributed intelligent devices, the capabilities of the power grid will increase. An autonomous grid that can dynamically react to changing ambient conditions, failures and consumer demands can maximize the capacity of energy generation, power transmission and distribution. Furthermore, collecting multimedia content could highly improve the safety and security of the grid. As low-memory image processing algorithms [4] and low-cost hardware that ubiquitously collect multimedia data, such as CMOS cameras and microphones [3,5] become available, the widespread adoption of WMSANs in the smart grid will be possible. Despite the opportunities that become available with WMSANs, WMSANs have several challenges to be addressed. They suffer from low link quality, high latency, limited bandwidth and limited battery, as well as, lack of established techniques for in-network processing, efficient video streaming, QoS provisioning, distributed decision and control algorithms. Moreover, there are challenges specific to the power grid. Power grid is a large-scale system, in the sense of geographical region and the number of equipments. A single malfunctioning device may cause partial (isolated) outage, or it may cause overall system failure and affect a larger region and higher number of consumers. Similarly, under utilization of resources at a single transmission line or a generation unit may impact the overall system performance. In addition, the operation and maintenance of the grid require strict safety precau-
543
tions for personnel and public health, therefore WMSANs should not cause hazardous situations. In this paper, we discuss the potential applications of WMSANs in the next generation power grid. We explore new applications for energy generation facilities, transmission and distribution (T&D) equipments and consumer premises, while discussing the related challenges. In Section 2 we present the state-of-the-art in power grid monitoring tools and introduce possible WMSAN applications. We discuss the challenges of employing WMSANs in the power grid and point out the open issues in Section 3. Finally, we conclude the paper in Section 4. 2. Multimedia sensors for the power grid In today’s power grid, sensing equipments are employed in a limited number of critical assets such as the power transmission lines and substations. Generally, these sensing equipments are not interconnected, i.e., they transmit the collected data to a central location via wireline communications, where a supervisory control system and human operators at the utility headquarters manage grid operations. The power grid can be conceptually divided into three segments to represent the flow of electricity from a generation facility to the premises where it is consumed. These segments are, energy generation segment (supplier), T&D segment (utility) and consumption segment (demandside). State-of-the-art sensing equipments are mostly employed in the T&D segment and less frequently in energy generation facilities. In the next generation power grid, it is necessary to have sensors in the generation plants and consumer premises, as well. In this section, we present possible WMSAN applications in the next generation smart grid while briefly summarizing the sensing equipments in use today. 2.1. WMSANs for energy generation facilities In the traditional grid, energy is generated at power plants which use resources such as nuclear fission, hydro, or fossil fuels (e.g. coal, gas, diesel, natural gas). In several countries, wind tribunes and Photovoltaic (PV) panels also contribute to energy generation. Wind and solar power are called as renewable or green generation methods and they are preferred more than the fossil fuel based energy generation techniques due to their lower cost and lower GHG emissions. However, their availability is limited to the geographical conditions of the grid and the time of the year. For example, solar power is not available at night and wind power varies seasonally. The intermittent nature of these generation techniques makes it hard to employ them as primary power supplies for a grid unless distributed energy storage is available. In the traditional grid energy storage methods such as pumped hydro, compressed air and flywheel are available however they are not convenient for storing the energy generated at the solar and the wind farms. Energy storage in the smart grid is one the active research topics. The amount of energy to be stored depends on the amount of
544
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551
energy generation and the demand. Therefore, in a power plant or in a wind/solar farm, real-time generator monitoring, provides insights for storage decisions. Monitoring and controlling the distributed energy generation and storage requires low-cost, distributed sensor and actor systems. WMSANs offer promising solutions in this field. They have large coverage, and they provide fault tolerance due to dense deployment. They can be rapidly deployed and they are able to self-organize when new nodes are added, or some nodes become unavailable. In the following subsections, we explain the opportunities where WMSANs can be used in energy generation facilities. 2.1.1. Wind farms Wind farms consist of wind tribunes that transform the force of wind to electrical power. For a wind tribune, temperature, pressure, humidity and wind orientation are critical parameters affecting the power output [6]. These parameters can be derived from scalar sensor measurements. On the other hand, bird collisions with wind tribunes is a significant problem affecting the healthy operation of the tribune. A recent study proposes to use acoustic emission sensors, i.e. microphones, which are mounted on the wind tribune tower to detect bird collisions [7]. Previously, accelerometers and fiber-optic sensors have also been considered, however they have been found less efficient since they need to be installed on each of the blade of the wind tribune, and rotor blade vibrations affect the ability of accelerometers to detect collisions. On the other hand, the drawback of acoustic sensors is that, they are affected by the turbine operating noise and the ambient noise. In this case, WMSANs can be employed to collect both audio and vision data. For instance, upon detection of a collision by the acoustic sensors, image sensors may wake-up to capture images of the blades and the tower to identify the faulty component, e.g. blade, pitch, gear box. Remote detection of the faulty component provides knowledge of required parts to be replaced, in advance, and speeds up the restoration of the wind tribune. Meanwhile, during the restoration, actors can shut down the faulty tribune, communicate with the utility headquarters to learn the amount of demand. For example, if the demand is high and more power is required, actors can bring backup tribunes online. 2.1.2. Solar farms For solar panels, temperature, radiation, DC voltage and current measuring scalar sensors can be used for predicting the energy output of the panel. In addition, the efficiency of solar panels decrease when solar beams are blocked by clouds. To increase the reliability of the prediction of energy output, WMSANs can monitor the sky and weather conditions, such as cloudiness. A simple illustration of solar and wind farms with WMSANs is given in Fig. 1. 2.1.3. Traditional power plants In the traditional nuclear, hydro or fossil fuel based power plants, temperature and pressure sensors are frequently used in addition to humidity, flow, vibration, volt-
Fig. 1. WMSANs for solar and wind farms with distributed storage.
age, current and motion sensors. These sensors generally communicate over wires, i.e., they require cabling. Recently, employing wireless sensors in power plants have been considered in several works. Eaton Inc. has developed a low-cost wireless motor monitoring system that continuously measures line voltage and current to evaluate motor energy used in a power plant [8]. In addition, US Department of Energy funded Industrial Technologies Program (ITP) also supports wireless sensor and automation technologies in power generation [9]. The ITP has proposed several wireless sensor application areas which are emission monitoring, flame sensing, turbine control, boiler control, motor control and monitoring the health of rotating assets [10]. These monitoring tasks can be more accurately established with WMSANs than WSNs. In addition to equipment monitoring, WMSANs can improve personnel safety. For example, under emergency conditions entrance of the personnel to hazardous areas can be remotely restricted, or the personnel could be routed to safety showers rapidly in a nuclear plant. 2.2. WMSANs for T&D facilities Transmission and distribution segment covers substations, overhead power lines and underground power lines. These are critical assets where an equipment failure or breakdown may cause blackouts or may be dangerous for public health. Moreover securing these assets is more challenging than power plants since they can be reached easily from outside. A recent study proposes to use WSNs for securing the power grid [11]. In this section, we investigate the further advantages of using WMSANs. T&D power lines of a utility covers a large geographical region with sometimes remote or hard-to-access components. Consequently, in the traditional grid, sensing equipments are deployed on a limited number conductors and towers, where the measured properties include sag, conductor strength, temperature, heating, galloping, icing, wind speed, and contact with vegetation and animals. The next generation power grid will cover the whole T&D segment by employing a large number of sensor and actor nodes to retrieve multimedia data and establish an autonomous grid.
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551
2.2.1. Power lines Overhead conductors heat and expand due to the ambient temperatures and the electricity current flowing through the conductor. Thermal expansion together with wind and adverse weather conditions, results in sagging. Sag is one of the significant factors limiting the ampacity, i.e., current capacity of the conductor. Legal and operational constraints force certain ground clearance values for overhead power lines to ensure public safety and healthy grid operation. The load in the power line can be increased as long as the conductor does not exceed the maximum allowable sag level. Therefore ampacity of a span, i.e. conductor section between two towers, is limited by the amount of sag. In the power grid, the span with the lowest ampacity determines the capacity of the T&D system. Therefore, utilities monitor the overhead power lines in real-time to avoid penalties or hazards due to sagging. Utilities are usually precautious, and they keep the ampacity lower than the actual possible value. This leads to inefficient operation of the T&D system. Integration of accurate monitoring tools will enable accurate ampacity estimation which will lead to efficient utilization of the power lines. Presently power line sag monitor, sagometer, Power Line Sensor (PLS) and Power Donut2 techniques are used for sag measurements. The power line sag monitor [12] uses an accelerometer to measure the inclination angle, i.e. the angle between the conductor and the horizontal plane. Acceloremeter transmits its readings to a remote processor where the sag is computed. The sagometer uses a video camera which monitors the movement of a passive target mounted on a conductor [13]. The sagometer communicates with the SCADA system using either spreadspectrum radio, fiber optics, satellite or cellular communications. PLS is able to take photos of clearance to ground in addition to line temperature, ambient temperature, wind measurements [14]. Power Donut2 [15] measures sagging as well as current, line to ground voltage, conductor temperature, current and voltage waveform. Power Donut2 transmits its data to a specified server via GSM/GPRS communications. Power line sag monitor, sagometer, PLS and Power Donut2 are relatively expensive tools therefore they are used in a limited number of transmission lines. However, by the use of low-cost multimedia sensors, sag measurement for all overhead power lines can become feasible. Moreover, multiple video sensors can provide larger field of view and more field of regard than a single Pan-TiltZoom (PTZ) camera that is available in the present systems. In today’s power grid, technology to avoid sagging is also available but these are costly to apply to all of the T&D power lines. Sagging Line Mitigator (SLiM) and Aluminium Conductor Composite Core (ACCC) are recently developed techniques to reduce sag. SLiM is a small device that can extend and contract passively and it is attached to the conductor to reduce sag as the heat and load increase [16]. SLiM is designed to retract upon heating and extract upon cooling therefore it can adjust the effective conductor length and prevent sagging. ACCC uses a composite material with less sagging property [17]. Other methods to reduce sag include raising towers or recabling the span, which are both costly. In the next generation smart grid, WMSANs can be employed to detect and correct sagging more efficiently.
545
In the T&D segment there are also various sensors to measure the conductor temperature profile, wind and other weather conditions or leakage current [18]. These are generally data sensors. Video sensors may be employed in detecting the wearing of the conductors or towers. Moreover, contact with vegetation or animals may cause safety issues, video sensors can also monitor the environment, and in case of danger they can activate circuit breakers and reroute the current over another span. In addition, wireless underground sensors can be used for monitoring the underground cables. 2.2.2. Substations Substations have been considered as the last mile of the traditional power grid. The terminal units communicate the measured critical parameters either via power line communication, satellite communication, optical fiber communication or wireless communications [19,20]. Using WSNs at the substations have been explored in [19] however WMSANs can enable a larger set of applications in the next generation power grid. For instance, islanding and safety of personnel can be better achieved via WMSANs. In the conventional grid, when the distribution system experiences a fault, a circuit breaker isolates the specific region easily. However in the smart grid, distributed generation units may still be feeding the section [21]. Although this is obviously advantageous to prevent blackouts, from another point of view, lack of communications may cause hazards when intended islanding is required. In this case, sensors and actuators can coordinate distributed generation units of the smart grid. When a utility personnel is detected by the WMSAN who is close to an energized component an alarm can be raised. This could reduce the risk of safety hazards to utility workers when power lines could be still energized in cases where they are assumed to be disconnected. WMSANs are also powerful tools for surveillance. Power lines and substations may be vulnerable to physical attacks and terrorism. In this case, video sensors can be triggered with the detection of an anomaly by the data sensors [22]. 2.3. Sensors on consumer premises For the traditional power grid, substations have been considered as the last mile and the extension of the grid into the consumer premises has not been implemented. This is due to the large amount of consumers which is unmanageable in the traditional grid. However, with the advances in the smart grid technologies, it will become possible to communicate, monitor and possibly control the power consumption of the consumers without disturbing their business or comfort. Consumers can be roughly classified into three groups based on their power needs and usage patterns. These are industrial, commercial and residential consumers. Although, in the traditional power grid, there are various sensing equipments for all of the three categories of consumers, usually these equipments are not interconnected and they have simple control principles. 2.3.1. Industrial consumers Presently, sensors are used to ensure safe and healthy operation of the equipments. They collect measurements
546
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551
such as temperature and pressure, and they are capable of rising alarms when the measured entities are above predefined thresholds. Advanced sensors are used in the control of robotic arms, e.g. in automotive industry. In addition, vision sensors are used to verify product quality in a wide range of industrial applications. Generally, the sensing equipment is programmed or trained for a correct data set and the primary focus of the quality control applications is to determine if a new product matches the defined standards. These sensors work individually, they are sometimes supervised by human operators, and they do not have wireless communication capabilities. Moreover, they are not interconnected with the power grid, and they do not have power management abilities. When WMSANs are used, they can play a significant role in the industrial process control, as well as the automated demand respond for industrial consumers. For the traditional grid, demand response programs are available for large-scale consumers. Industrial consumers such as plants or commercial consumers such as corporate buildings, may opt in the demand response program which is usually handled by an aggregator company. The primary focus of the demand response is to reduce peak hour consumption. In the grid, managing peak hour consumption is important for healthy grid operation and it is also one of the factors affecting the cost of energy. In deregulated electricity markets, independent system operators use load forecasting tools to predict the demand for the next-hour or the next-day [23]. Electricity suppliers, generate or purchase energy in matching amounts with the demand and they supply bids for the amount of energy they are able to dispatch. Usually the suppliers with lower bids are selected [24] which settles the price of electricity for a certain time interval. Regulation fees, taxes, transmission, distribution, maintenance costs are added to determine the final fixed price of electricity for the consumers. This means, although the price of electricity purchased by the utility varies in time, the price of electricity sold to the consumers is fixed. The reason for the varying cost of energy on the supplier side mostly depends on the energy generation method. Suppliers operate base power plants and peak power plants (peaker plants). Since energy cannot be stored in large amounts, generation and load have to be balanced. The supplier operates base plants which are able to accommodate an average load level. Generally, depending on the demand, they can be partially shut down and started again in a daily cycle. For example, during overnight hours, the demand is low, and less energy is generated, while in the morning, demand rises and more energy is generated by starting up the extra facilities. However, during the day, there may be several hours where the demand increases above the available capacity of the base plants. For example, the power demand of the HVACs will increase in hot summer days, especially in the afternoon. These hours are called as the peak hours. Extending the capacity of the base plants according to peak demand, e.g. building a new nuclear power plant which will be used only for several hours a day is costly. In most of the cases, peaker plants are brought online to accommodate the peak load. Peaker plants can respond to power demand in a short
time but they generally use fossil fuels which are more expensive. For this reason, the price of electricity rises during peak hours. If the base and peaker generation capacity become insufficient consumers may experience outage. In the traditional grid, aggregators have been managing the demand response by basically calling consumers to reduce their consumption at peak hours. For instance, in commercial buildings, HVACs are cycled off to avoid penalties or to receive credits. In a factory, equipment cooling can be cycled off during peak hours as long as the sensor readings are in safe operation margins. Recently, automated demand response has been implemented in California where demand response is managed by utility signals [25]. This could be further improved for a demand response program exploiting the dynamic pricing of the smart grid. The smart grid, aims to bill the consumers based on the time of consumption. This is called as Time of Use (ToU) pricing [26]. Furthermore, using price signals can also provide demand response for all of the consumers, including the residential consumers. In this case, WMSANs can be used in monitoring and managing energy consumption. During peak hours, scalar sensors can measure the electric current with several techniques such as inductor, Hall effect, GMR sensor and magnetic force on MEMS and video sensors can be used to monitor the unused equipments and actors can shut them off safely. 2.3.2. Commercial consumers Corporate buildings and shopping malls are the examples of commercial consumers. For these consumers, occupancy and temperature sensors are widely used. In addition, air quality measurement becomes important as HVACs can be cycled off for energy management [27]. Air quality is generally monitored by electrochemical sensors that measure CO2 level. Typical applications of WSN for the commercial buildings in smart grids are discussed in [28]. WMSANs have additional application fields, as well. WMSANs can be used in demand response for commercial buildings similar to industrial consumers. Moreover, the acoustic and video sensors can improve occupancy sensing for commercial consumers. WMSANs can be effectively used in parking lots of the commercial buildings and shopping malls for Plug-in Hybrid Electric Vehicle (PHEV) applications, as well. PHEVs are hybrid vehicles that use a mixture of gas and electricity. Electrical power is stored in a Lithium-Ion battery which can be charged from an electrical outlet. PHEVs are expected to be on the roads shortly, however their charging properties bare significant challenges for the power grid. Inevitably, PHEVs will increase the load on the grid and they may cause resilience problems if they are charged in an uncontrolled way. The impact of PHEV charging profile on the power grid has been explored in [29]. There is consensus on the need for controlled smart charging, yet standard technique is missing. Several smart charging proposals, including price prediction [30] and optimization [31] exists in literature. Besides regulating the time of charging, there are also emerging vehicle-to-grid (V2G) applications where PHEVs are considered for distributed power storage. The opportunities and challenges of PHEV-to-grid connections is out of the scope of this paper, however we outline the applica-
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551
tions where WMSANs can be used for PHEV charging and discharging decisions. In [32,33], PHEVs are proposed to be used for ancillary services. Ancillary services include spinning reserves and regulation. Spinning reserves are the idle sources that can dispatch power to the grid within 10 min of a request. These are the services which are generally paid to be immediately available and they do not generate power unless requested. Regulation services are generally under real-time control of the system operator and they are used for keeping the system frequency as close to 60 Hz as possible. A fleet of PHEVs can be contracted to provide the ancillary services. In this case, pricing of consumption and revenue from supporting the grid, should be considered in detail. Although permanent contracts with fleet operators is feasible, it does not provide flexibility for the individual PHEV owners. For this purpose, public charging stations located at parking lots can utilize WMSANs. Vision sensors can be used for identifying license plates and they can arrange billing accordingly. Similarly, electricity theft at the public stations could be prevented by the help of WMSANs. Moreover, multimedia sensors that are mounted on PHEVs can communicate with the other vehicles in a lot to share maps and traffic data. This information can be used to determine the duration of charging or the amount of energy available for discharging, by predicting the amount of battery power required to reach the next destination of the vehicle. For example, if the daily driving pattern of a driver includes a visit to a shopping mall, then to the school to pick up children and then driving home, PHEV can communicate with the other vehicles in the shopping mall, gather traffic information and construction information along the route of the driver and predict the required power to complete the trip. The charging duration or the discharging amount could be transmitted to the WMSAN of the parking lot where actors coordinate charging of the other vehicles.
2.3.3. Residential consumers Using WSN for energy management in homes, have recently been explored in [34,35]. In-home energy management has received attention from the academia, as well as the industry. In this course, smart home concept [36] has regained importance with a focus on energy efficiency [37]. For example, Intel has developed the Home Energy Dashboard concept to provide a simple interface to consumers to monitor their monthly bills and the electricity consumption of the individual appliances [38]. Besides providing savings for the consumers, energy management schemes can reduce the carbon footprint of the households [39]. In the smart grid, consumers are motivated to supply their energy from solar panels and small wind tribunes. Monitoring these equipments is important for the efficiency of the whole grid and the savings of the individual consumers where WMSANs can play a significant role. Real-time monitoring of solar panels and wind tribunes may increase equipment lifetime, increase the predictability of their output and report malfunctioning devices. An
547
Fig. 2. WMSAN application for residential consumers and PHEVs.
illustration of the WMSAN for the smart home with PHEV is given in Fig. 2. 3. Challenges of WMSAN applications in the next generation power grid Major challenges related with using WMSANs for the power grid, rise from the real-time operational restrictions, coupled with the physical channel handicaps of WMASNs; such as, low link quality, low data rate, high or variable latency. Additional challenges are related with processing, storing and transmitting the huge amount of data generated by the large-scale power grid. Multimedia sensors and actors are battery-limited, therefore energy efficiency is another challenge. Sensor nodes also have limited memory and processing capabilities which make video coding both challenging and necessary. A WMSAN may involve a combination of sensors that collect data, sound, motion and video to monitor a certain region, where multiple sensors possibly get correlated or redundant data. The correlation and redundancy can be advantageous when handled properly. In addition, transport protocols need to be able to provide QoS provisioning and they should incorporate priority mechanisms for applications that require strict real-time operation. 3.1. Link quality Wireless link quality in power plants, T&D system and the premises of industrial consumers is expected to be low. In literature, there are few works addressing this issue [19,40] where the authors focus on IEEE 802.15.4-based WSN operation in industrial conditions. Physical wireless channel models for communications around power lines and communications inside the substations, the impact of link quality on the connectivity and the topology of a WSN, the impact of connectivity on the real-time operation of the power grid are still open issues. The impact of harsh environmental conditions on sensors and communications is also unexplored. For outdoor sensor nodes, corrosive conditions, e.g. solar radiation, wind, rain, humidity or
548
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551
other weather factors, could affect the performance, whereas for indoor environments such as substations or power rooms; dust and heating could be potential problems. Additionally, low link quality may degrade the performance of WMSANs more than it does in WSN. In WMSAN, a single bit error may cause an entire sequence of frames to be discarded. For example, in video streaming, if an independently coded (I) frame is discarded due to channel errors, it may lead to the loss of an entire sequence [41]. The challenges raised by the low link quality can be addressed by transport protocols or cross-layer approaches which are still open issues. 3.2. QoS provisioning Time-varying and location-dependent link capacity and packet errors may cause some critical packets to be lost. In smart grid applications, it is essential to develop novel schemes to give priority to the packets of real-time applications. A cross-layer approach has been proposed in [42]. This study considers a substation scenario and employs Ultra Wide Band (UWB) communications which have high data rates within tens of meters. QoS provisioning for WMSANs that are employed in other grid assets, is still an open issue. 3.3. Latency Power systems generally have tight latency requirements. Most of the decisions are in real-time where delayed information may lead to less efficient operation or wrong decisions [19]. For this reason, variable latency has to be addressed by the transport protocols, as well as the decision and control algorithms. 3.4. Video streaming In the next generation power grid, cyber and physical security will have major importance. Surveillance of utility assets can be provided via video sensors of WMSANs. However, video streaming in WMSAN still needs to be improved in terms of throughput, fairness and encoding/ decoding techniques. Traditional video streaming principles are contradictory with sensor network architecture, as complex algorithms for encoding are present on power-constrained and resource-limited sensor nodes while decoding is done by more powerful nodes. There are several works in literature addressing this issue. In [41], the authors propose a compressed sensing technique to reduce the computational complexity at the encoder. Another technique combining compressed sensing and rate control algorithm has been proposed in [43], as well. Moreover, in a WMSAN, video sensors may simultaneously transmit streams where maximizing the throughput and fairness of each transmission become another challenging issue. In [44], the authors propose a rate control algorithm that uses RTT and SNR measurements and calculates the optimal ratio of video encoder rate to channel encoder rate. This algorithm can be applied to WMSANs for smart grids, to increase the performance of the real-time streaming,
especially when high quality video from different parts of the network are required. 3.5. In-network processing In-network processing will be an important part of the smart grid applications. It improves the performance of WMSANs by decreasing the amount of data to be transmitted to the sink and the centralized supervisory system, i.e. SCADA. Moreover, when it is combined with computer vision techniques, it can provide fast reacting opportunities for the actors. 3.6. Battery Wireless sensors inherently have energy limitations. In the power grid, some of the sensors will be deployed in locations that are hard to access and replace. For example, replacing sensors for underground cables is costly. For this reason, energy scavenging or energy harvesting becomes crucial [45]. In the power grid, there are numerous opportunities for energy harvesting, such as vibration, magnetic field, motion, solar, ambient light, air flow, temperature gradients, and pressure gradients. Although there are examples of sensors applying energy harvesting from the energized electric conductors, possible opportunities have not been fully exploited yet. 3.7. Decision and control algorithms The next generation power grid will be autonomous where actors will manage load balance, power switching, islanding and a large amount of the other grid management operations. Efficient grid management will be possible by distributed decision and control algorithms [46]. In a local operation field, actors will be applying the decisions of distributed control algorithms and at the same time communicating with the other actors on the power grid to synchronize a specific task. Today’s power grid is controlled centrally by human operators and the SCADA system. However, in the next generation power grid, supplier and consumer roles are redefined where consumers are able to sell power to grid, and the grid operators are able to manage the consumption. In this complex real-time system, decisions have to be taken rapidly. Developing robust decision and control algorithms for the power grid and novel tools to improve SCADA are essential. A recent tool, Dynamic Monitoring and Decision Systems (DYMONDS) aims to transform SCADA and develop a software tool with distributed decision and control abilities [47]. 3.8. Data management In the next generation power grid, the amount of data collected by audio, vision and scalar sensors will be huge. It is crucial to aggregate data since correlated or redundant data may be collected by sensors close to each other and sensors that have the same field of view. Moreover, employing an hierarchical architecture and allowing some portion of the data to be processed by lower tiers before it
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551
is unnecessarily transmitted to the utility headquarters may help in data management. 3.9. Large-scale WMSAN for the power grid, is inherently have the potential to become a very large-scale network. First, the geographical span of the power grid will be large. Second, the number of sensors and actors employed in each component will be dense. For a typical utility with 25,000 km of high voltage power lines, thousands of transformers, capacitors and breakers, it is estimated that monitoring only the T&D assets could require more than 100,000 sensors [18]. Management of such a large-scale network is challenging. 3.10. Safety Using sensors and actors close to the power grid components should be safe. Especially in power plants and substations, sensor nodes should not be capable of igniting materials, even under faulty conditions. For example, when gas leakage from an underground pipe occurs around an underground power line, which is being restored by actors, the motion of the actors may ignite the gas. Field tests under adverse conditions have to be established before WMSANs can be safely integrated to the power grid. 4. Conclusion Growing demand for energy, diminishing fossil fuels, efforts to reduce GHG emissions and resilience problems that showed up in the electrical power grid have led to common consensus on the need for a renovation of the power grid. The key to this renovation is the integration of ICT to enable addressing the imbalance between demand and supply, widespread adoption of renewable energy farms, monitoring utility assets and self-healing under failures. Presently, researchers in academia and industry are focusing on applying WSN technologies to the smart grid. In the next generation power grid, applications that go beyond WSN capabilities will be demanded. The scalar wireless sensors will be replaced by wirelessly connected multimedia sensors that are able to collect audio/video, and actors that are able to control the grid operations. WMSANs will enable fine-grained monitoring by multimedia sensors and distributed control of the grid by actors. In this paper, we have outlined the possible WMSAN applications for the power grid while discussing their challenges and pointing out the open issues to be addressed before they are widely adopted. WMSANs can be used in energy generation facilities such as nuclear, hydro, fossil fuel based power plants, wind farms and solar panels. Coordination of distributed power generation and storage is possible with fine-grained monitoring, as well as distributed control algorithms. Moreover, T&D equipments can be monitored and maintained by the WMSANs. The efficiency of the power grid and its ability to accommodate certain levels of load are restricted by safety measures of the overhead power lines. For this
549
reason, the energized power lines need to be monitored. In a WMSAN, vision sensors can monitor the sag and provide real-time estimates of load. WMSANs can be also used in the consumer premises (demand-side) to coordinate supply and consumption balance, and to improve demand response programs. Meanwhile, the use of WMSANs for the power grid introduces several challenges that need to be addressed. Link quality may be affected by weather, ambient conditions and the electromagnetic field of the electricity current. Low link quality may result in very low throughput especially for video streaming applications. Moreover, the power grid has strict real-time requirements where variable latency of the WMSANs may result in misconfigurations. QoS provisioning is another important issue to be addressed. Multimedia sensors have the potential of generating larger amount of data. Moreover, considering the large geographical region covered by a utility and the number of equipments present in the grid implementation, transmission of the collected raw data to the utility headquarters is not feasible. The reason is partly because the limited bandwidth and latency bounds and partly due to the limited sensor batteries. In-network processing techniques needs to be explored to handle this issue. To relax battery constraints energy scavenging techniques may be employed however energy scavenging in the power grid is not fully exploited. In the next generation smart grid, actors can play an active role in management and operation of the grid which requires robust distributed control algorithms. Finally, the sensor and actor nodes need to be safetyproved before they can be employed in the power grid. They should not be capable of ignition, overheating or yielding any other reaction that could cause hazardous situations. Acknowledgment This work was partially supported by an ORF-RE grant from the ministry of Research and Innovation of the Government of Ontario. References [1] S.M. Amin, B.F. Wollenberg, Toward a smart grid: power delivery for the 21st century, IEEE Power and Energy Magazine 3 (5) (2005) 34– 41. [2] I.F. Akyildiz, W. Su, Y. Sankarasubramaniam, E. Cayirci, Wireless sensor networks: a survey, Computer Networks 38 (4) (2002) 393– 422. [3] I.F. Akyildiz, T. Melodia, K.R. Chowdhury, A survey on wireless multimedia sensor networks, Computer Networks 51 (4) (2007) 921–960. [4] S. Rein, M. Reisslein, Low-memory wavelet transforms for wireless sensor networks: a tutorial, submitted for publication.
550
[8]
[9] [10] [11]
[12] [13] [14] [15] [16]
[17] [18]
[19] [20]
[21]
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32] [33]
[34]
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551 January 2007.
[35]
[36]
[37]
[38] [39]
[40]
[41]
[42]
[43]
[44]
[45] [46] [47]
Computers and Communications (ISCC’10), Riccione, Italy, June 2010. M. Erol-Kantarci, H.T. Mouftah, TOU-aware energy management and wireless sensor networks for reducing peak load in smart grids, Green Wireless Communications and Networks Workshop (GreeNet) in IEEE VTC2010-Fall, September 6–9, 2010. M. Chana, D. Estevea, C. Escribaa, E. Campo, A review of smart homes – present state and future challenges, Elsevier Computer Methods and Programs in Biomedicine 91 (2008) 55–81. K. Kok, S. Karnouskos, D. Nestle, A. Dimeas, A. Weidlich, Cor Warmer, P. Strauss, B. Buchholz, S. Drenkards, N. Hatziargyriou, V. Lioliou, Smart houses for smart grid, in: 10th International Conference on Electricity Distribution CIRED, Prague, June 2009. Intel Home Energy Management Dashboard.
Melike Erol-Kantarci received her B.S. (2001), M.Sc. (2004) and Ph.D. (2009) degrees from Computer Engineering Department, Istanbul Technical University, Turkey. Currently, she is a postdoctoral fellow at School of Information Technology and Engineering, University of Ottawa. From September 2006 to August 2007, she has been at the Computer Science Department, UCLA, as a Fulbright visiting researcher. Her main research interests are residential energy management, smart grids, underwater sensor networks and wireless sensor networks.
Hussein T. Mouftah joined the School of Information Technology and Engineering (SITE) of the University of Ottawa in 2002 as a Tier 1 Canada Research Chair Professor, where he became a University Distinguished Professor in 2006. He has been with the ECE Dept. at Queen’s University (1979–2002), where he was prior to his departure a Full Professor and the Department Associate Head. He has six years of industrial experience mainly at Bell Northern Research of Ottawa (became Nortel Networks). He served IEEE ComSoc as Editorin-Chief of the IEEE Communications Magazine (1995–1997), Director of Magazines (1998–1999), Chair of the Awards Committee (2002–2003), Director of Education (2006–2007), and Member of the Board of Gover-
M. Erol-Kantarci, H.T. Mouftah / Ad Hoc Networks 9 (2011) 542–551 nors (1997–1999 and 2006–2007). He has been a Distinguished Speaker of the IEEE Communications Society (2000–2007). Currently he serves IEEE Canada (Region 7) as Chair of the Awards and Recognition Committee. He is the author or coauthor of 6 books, 32 book chapters and more than 950 technical papers, 10 patents and 140 industrial reports. He is the joint holder of 12 Best Paper and/or Outstanding Paper Awards. He has received numerous prestigious awards, such as the 2008 ORION Leadership Award of Merit, the 2007 Royal Society of Canada Thomas W. Eadie Medal, the 2007–2008 University of Ottawa Award for Excellence in Research, the 2006 IEEE Canada McNaughton Gold Medal, the 2006 EIC
551
Julian Smith Medal, the 2004 IEEE ComSoc Edwin Howard Armstrong Achievement Award, the 2004 George S. Glinski Award for Excellence in Research of the U of O Faculty of Engineering, the 1989 Engineering Medal for Research and Development of the Association of Professional Engineers of Ontario (PEO), and the Ontario Distinguished Researcher Award of the Ontario Innovation Trust. Dr. Mouftah is a Fellow of the IEEE (1990), the Canadian Academy of Engineering (2003), the Engineering Institute of Canada (2005) and the Royal Society of Canada RSC: The Academy of Science (2008).