The smart gird is the integration of the 20th century traditional electrical power grid with the most recent 21st telecommunication and information technologies. Such integration enables efficient resource utilization to optimize energy consumption, install and manage distributed energy sources, as well as to exchange the generated power. In other words, the power flow and communications will be in two-ways. Many utility companies around the globe started to install renewable energy sources such as solar and wind energy nearby the consumption sites. Also, residential home owners started to install smart home appliances and renewable energy resources in their premises to generate and consume electrical power efficiently.
As the smart grid concepts emerged as a fast growing research and development topic in the last few years, the National Institute of Standards and Technology (NIST) developed a conceptual model for the smart grid to set the stage for a better understanding to the smart grid technology. The NIST conceptual model consists of seven domains, namely bulk generations, transmissions, distributions, consumers, markets, operations and service providers. Smart grid users communicate in two-way directions by utilizing several wireless and wired communication protocols such as Zigbee, WiFi, Homeplug, power line carrier, GPRS, WiMax, LET, Lease line, and Fibers. Several software packages were updated and many are being developed to accommodate the new grid operation, maintenance and management such as, distribution management system (DMS), geographic information systems (GIS), outage management systems (OMS), customer information systems (CIS), and supervisory control and data acquisition system (SCADA).
Figure 1: Smart grid communications protocols
As a result of the smart grid evolution, some recent enabling technologies have emerged to reduce the number of communication protocols and handle big amounts of data. The Internet of Things (IoT) is one the most recent enabler for the smart grid.
Internet of Things (IoT) is a recent new concept, on which Internet evolves from connecting machines and peoples towards connecting (smart) objects/things. Thus, we can say that IoT communications is the evolution of M2M communications. According to Cisco, by 2020 there will be over 50 billion connected objects against a population of 7 billion. An object can be anything/device/entity equipped/embedded with computation, storage and communication capabilities with different capacities (sensor, actuator, mobile phone, desktop, laptop, printer, car, fridge, oven, etc). While smart objects are already connected through proprietary non-IP solutions in different applications (Zigbee, HART/ Wireless HART, Z-Wave, etc.) and at a small scale, IoT aims at connecting the objects at a large scale using IP-based solutions (IP, TCP/UDP, etc.), directly or through gateways if IP support is not possible, while allowing them to interact with any other communicating party on/over the Internet.
The Smart Grid (SG), the intelligent power grid, could be seen as the largest instantiation of the IoT network in the next future. The whole power grid chain, from the energy power plant generation to the final electricity consumers (houses, building, factories, public lightning, electric vehicles, smart appliances, etc.), including transmission and distribution power networks, will be filled with intelligence and two-way communication capabilities to monitor and control the power grid anywhere, at a fine granularity and a high accuracy. For instance, smart houses, will be equipped with smart meters and smart appliances, whereas power generators and electric transmission and distribution networks will be equipped with various sensors and actuators. The aim of the SG is to keep a real-time balance between energy generation and consumption, by allowing a fine-grained monitoring and control over the power chain, thanks to the huge number of the two-way communicating smart objects (smart meters, smart appliances, sensors, actuators, etc.)
While the use of IoT is very prominent in the context of the SG, it could also lead to disasters. Indeed, as a critical infrastructure, the SG will now be more attractive to cyber-attacks, since its monitoring and control could be done over standard internet-based protocols and solutions, and may rely on public communication infrastructure. As a consequence, an attacker could cause financial loses to the utility and make damage to the electric assets by breaking the real-time balance between energy consumption/production, through manipulating data generated by the smart objects or sent from the utility.
Figure 2: 6LowPAN frame structure for smart grid applications.
Figure 3: 6LowPAN frame structure for smart grid applications
Smart Grid Communication Protocols
Smart grid communications are based on wireless and wired networks technologies. Regardless of the technology, these networks can be classified based on their functionality within the smart grid. These classifications, as reported in the literature, are: home area network, neighborhood area network, access network, backhaul network, core and external networks. These networks connect many smart grid objects such as home appliances, smart meters, switches, reclosers, capacitors bank, integrated electronic devices (IEDs), transformer, relays, actuators, access points, concentrators, routers, computers, printers, scanners, cameras, field testing devices, and other devices. All these appliances and devices are geographically distributed throughout the grid, starting from residential units to substations and up to utility data and command centers.
As mentioned in the introduction, each device can access and exchange data via different communication protocols. Figure 1 shows the smart grid communications protocols layers. The bandwidth and latency requirements for the smart grid appliances and devices vary from few msecs to several minutes and from few kbps to few hundred kbps as shown in Table 1.
IoT Smart Grid Conceptual Model
As mentioned in the previous sections, smart homes have several appliances and some form of renewable energy resources. These appliances and resources can be considered as IoT technologies. Each can upload and download data and commands from utilities and home owners. In addition, the grid at large has many devices that can be considered as IoT objects such as reclosers, switches, capacitor banks, transformers, IEDs, smart sensors, and actuators in the substations. In general, smart grids for large cities or countries may have millions of home appliances and thousands of grid devices.
This research proposes that each one of the appliances and devices can have a unique IP address. For example, a dishwasher has a unique IP address a transformer’s IP address. This requires the smart grid to have a large number of IP addresses. This is not an issue as the IPV4 is extending from 32-bits to 128-bits address size IP addresses. The IPV4 can address up to 232 devices (4-billionunique addresses). Moreover, IPV6 can address up to 2128 (Trillions of unique addresses).
One outcome of such addressing schema is the 6Low PANcommunication protocol. It embarks on top of IPV6 and is designed to be used over the IEEE 802.15.4 standard. The 6LowPAN frame sized is limited to 127 bytes including a payload of 21 bytes for TCP and 33 bytes for UDP. With some techniques, the payload may increase to 65 – 75 bytes. This is adequate for the smart grid appliances and devices monitoring and controlling applications. This protocol is the backbone of the IoT communication media.
To model the smart grid within the IoT context, smart home appliances, renewable energy resources, substation devices and workforce tools will be assigned IPV6 address as follows:
- Smart home appliances:
Recent smart homes are equipped with smart appliances and each appliance is considered as a thing (object). These things can be an air-conditioner, water-heater, dishwasher, refrigerator, smart energy/gas/water meters, in- home-display, automated lights, solar energy cell, wind mill, electrical rechargeable vehicle, and storage battery [9]-[11]. In the proposed model a unique IP address is assigned to each appliance and device. Each appliance or device can be accessed through the internet by an authorized personnel such as a utility’s operator or homeowner. The appliance status can be transmitted (uploaded) or control command to be received (downloaded). The exchange data and control commands utilize the payload portion of the 6LowPAN frame as shown in Figure 2.
- Substations devices:
The power substation has many devices (things) such as transformers, breakers, switches, reclosers, meters, relays, IEDs, capacitor banks, voltage regulators, cameras, and several other things. Similarly to smart homes, each device (thing) in the substation is considered as an object and is assigned a unique IP address. Each object (thing) can transmit its status and receive control commands from the utility authorized operator via the Internet. The payload is few bytes and can be accommodated using the 6LowPAN protocol as shown in Figure 3.
- Distributed renewable energy resources:
The distributed renewable energy resources are one of the major smart grid enablers that can be installed around the residential neighborhoods, distributed transformers and substations. It supplements power sources that can be installed quickly to be used during the peak hours, as well as on other times of the day when is needed. Each one of these source can supply power to operate, monitor and control. An IP address can be as- signed to each appliance and device. The payload size and other related 6LowPAN frames are shown in Figure 2.
- Mobile workforce tools and devices:
To operate the grid efficiently, a mobile workforce should be on the move 24 hours a day, 7 days a week to fix issues related to residential power outages, feeders, transformers, meters, power lines, and other related issues. The workforce operators are equipped with ragged laptop, smart meters, mobile phone, and cameras. Each of these devices is assigned an IP address and can be accessed as in the above mentioned devices and appliances in Sections 1 – 3.
- Utility data and control center infrastructure:
This center has many applications and database services such as, distribution management system (DMS), geographic information systems (GIS), outage management systems (OMS), customer information systems (CIS), and supervisory control and data acquisition system (SCADA). Each service has its own IP address.
- Echo systems:
The echo systems could be external power server providers, marketing and third parties power providers. Each of which should have point of access through an IP address.
Figure 3 depicts the above mentioned proposed conceptual model for the smart grid within the Internet of Things contexts. It mimics and integrates the about appliances and devices in model that is scalable. The proposed conceptual model introduces other challenges in security and handling big data that are beyond the scope of this paper. It is worth mentioning that cloud computing is a paradigm that enables a solution to the smart grid environment requirements related to computational power, storage, and high availability of resources.
Figure 4: The Smart Grid Conceptual Model
Figure 5: General View of the AMI
Internet of Things And Smart Grid
Internet of Things
The term IoT often makes reference to the integration of (resource-constrained) objects, such as sensors, actuators, RFID tags or any device involving a communicating interface and a computing capability, into the Internet. Objects of the physical world (fridge, window, heater, switch, washing-machine, etc.) could now be easily accessible, manageable and communicate through Internet using internet-based protocols (IPv6, UDP/TCP, HTTP, etc.). For the most resource-constrained devices, especially, those compliant with the IEEE 802.15.4 standard, the IETF (Internet Engineering Task Forces), proposed several protocols for their efficient integration and at different layer to the Internet:
- 6LowPAN: IPv6 over Low Power Wireless Personal Area Networks, an adaptation layer to support the IPv6 protocol on IEEE 802.15.4 networks
• RPL Routing Protocol for Low-Power and Lossy Networks
• CoAP: Constrained Application Protocol, is a specialized web transfer protocol for use with constrained nodes and constrained networks
Even for those objects that still couldn’t support IP natively, or updated to support it (due to extremely resources constraint or other considerations like preserving legacy systems), integration to the global Internet network is still possible through gateways, where proprietary non-IP stack protocols (Zigbee v1, HART, Z-Wave, etc.), are translated to/from IP stack protocols, but at a highly cost and without achieving and end-to-end communication.
Smart Grid
The SG can be seen, in its simple form, as the classical power grid augmented with the massive use of ICT technologies (software, hardware, networks), in addition to the integration of distributed renewable energy generation and storage capacities. As seen in Figure 4, in the SG there are two flows:
- Electric flow (dashed line) from the plant generation to the end customer, which is the main flow of the classical power grid. However, in the vision of the SG, the electric flow could be bidirectional, where the end-customer will buy and could also sell energy.
• Information flow (regular lines): A large-scale two-way communication flow between the different shareholder and components of the SG. Most of the communication flow is due to the massive use of sensors/actuators and other smart objects alongside the transmission and distribution areas, in addition to the use of smart meters and other smart objects (smart appliances, electric vehicles, etc.) at the end-customer side.
The SG involves, amongst others, two key elements, which are Smart Meters and Advanced Metering Infrastructure:
• Smart/Advanced meters (SMs), equip houses, factories, institutions, etc. (see Figure 5). They record energy consumption data and other information for billing or management purposes. They can report data periodically, upon request or in response to some events to the utility and also respond to requests from the utility (e.g., software update, real-time pricing, load shedding, energy cut-off, etc.), thanks to their two-way communication capability. They may optionally play the role of local energy management system, by controlling or managing the energy consumption of the smart devices on the house (fridge, oven, air-conditioner, electric cars, etc.)
• Advanced Metering Infrastructure (AMI), as shown in Figure 5, is responsible for collecting, analysing, storing and providing the metering data sent by the SMs to the appropriate authorized parties (e.g., energy provider, utility, SG’s operator, Meter Data Management Service, etc), so they can proceed them (billing, outage management, demand forecasting, etc). The AMI is also responsible for transmitting requests, commands, pricing-information and software updates from the authorized parties to the SMs.
IoT-based SG
Compared to classical power grid, the SG highly integrates ICT on the whole energy chain (from producers to end-consumers), through the large-scale deployment of different kind of sensing, actuating and other embedded devices, in addition to the use of smart meters, smart appliances and e-cars, all of them sharing the capacities of computing and communication.
What has made Internet universally popular is the use of standard communication protocols, mainly the TCP/IP stack. Any two computers situated anywhere in the world, could easily have an end-to-end communication, regardless their access technology. IoT extends the reachability of Internet to reach, through standardized communication protocols (or a gateway in the extreme case), everything that could communicate and be individually addressed. This copes with the huge number of devices/objects deployed on the SG and the crucial need of near-real time communication with them through unified standard-based communication protocols (based on the TCP/IP stack), rather than proprietary solutions (Zigbee v1, (W)HART, Z-Wave, etc.).
Assuming that the SG of a country involves 20 million smart meters, in addition to 40 million sensors and actuators deployed to monitor the whole power grid infrastructure. For the SG’s operator, it will be interesting to remotely manage and configure the smart meters and the sensors/actuators– regardless their manufacturer- in addition to get information on the last mile grid’s status. For the energy providers, it will be interesting to get remotely energy consumption from SMs in-order to accurately bill the customers, in addition to detect attempts of tampering with the SMs (ex, energy theft). For the end-user, it will be also interesting to get up-to date prices (assuming dynamic pricing), to well manage its consumption, in addition to get early alerts about planed disconnection. Obviously, all these bidirectional end-to-end interactions and communications, will highly benefit from IP based communication protocols (unless it is impossible or not appropriate), and even public communication infrastructures to make them easily scalable and to make induced costs lower.
IoT-based Smart Grid’s Security Issues and Challenges
The added ICT dimension to the classical power grid, introduced new security issues and challenges that were not (or rarely) present on the classical power grid. Those security issues and challenges could hamper the rapid deployment and adoption by end-users of the IoT-based SG. Hereafter, we briefly describe the most important security issues and challenges faced on the IoT-based SG.
Security Issues
As a cyber-physical system, the IoT-based SG will face several security issues:
- Impersonation/Identity Spoofing
This attack aims at communicating on behalf of a legitimate thing in an unauthorized way, by making use of its identity. An attacker could spoof the identity of some one’s smart meter, in order to make it paying for its energy consumption.
- Eavesdropping
Since objects/devices on the IoT-based SG communicate, often using public communication infrastructure, an attacker can easily have access to their exchanged data. An attacker can easily know the energy consumption of households
- Data tampering
An attacker can modify exchanged data, such as dynamic prices sent prior to peak periods, making them lowest prices. As a consequence, this could make households increasing their consumption (charging e-cars, etc.) instead of reducing them, thus, resulting in overloaded power network.
- Authorization and Control Access issues
Since several devices could be monitored and configured remotely, such as smart meters, or field deployed sensors and actuators in distribution substations, an attacker or even an angry employee, could try to gain an unauthorized access rights, to manipulate them, thus damaging physical assets (ex, transformers) or leading to power outages.
- Privacy issue
Smart meters and smart appliances in residential houses, could tell more than the energy consumption. Their generated fine-grained data could harm the privacy of the end-user, by divulging information about their habits (wake up, sleeping and dinner times, etc.), if they are in or away from house, if they are on vacation, etc.
- Compromising and Malicious code
Since objects of the SG are computation and communication enabled, they are target to compromising physically or remotely. Moreover, since they run different kind of software, they could be target of different kinds of software infection or malicious code infection in-order to control and manipulate them (ex, targeting smart meters, or smart appliances in households). Moreover, massively deployed objects with constrained devices (sensors, etc.) are usually non-tamper-resistant devices, making physical compromising an easy task
- Availability and DoS issues
In the classical power grid, it was difficult, if not impossible, to target the availability of assets (electricity meters, substations, etc.), especially, at a large scale. In the SG, ICT will be integrated even in the vital assets of the power grid, thus making it possible to target them, making them partially or totally unavailable resulting on DoS attack. Moreover, assuming that most devices/things are IP-enabled and do not run proprietary protocols stacks, the task of a familiar Internet attacker will be easier.
- Cyber-attack
The SG could be seen as the largest Cyber-Physical-System (CPS)6, involving Physical systems representing the physical assets of the SG (transformers, circuit breakers, smart meters, cables, etc.) and ICT systems, where ICT elements control/manage physical entities. Now, a Cyber-attack could harm the physical assets – as was the case with the Stuxnet attack -, which was difficult in the classical power grid.
Security Challenges
When dealing with security algorithms, protocols and policies for the IoT-based SG, several challenges need to be taken into consideration:
- Scalability
The SG could span over large areas (several cities or the entire country), and involves a large number of smart devices and objects. This will make it difficult to conceive scalable security solutions, such as key management and authentication5.
- Mobility
With mobile devices/objects, such as e-cars and on-the field technical agents, there will be a continuous need for authentication and secure communication with a changing surrounding (smart meters, electric charging stations, etc.).
- Deployment
Since the SG could span to the entire country, objects/devices are deployed at a large scale, work unattended, and could be placed on remote places with no physical perimeter protection, making them easily accessible. Security solutions should be able to detect any attempt to tamper with them.
- Legacy systems
Already deployed systems and devices, could have a little or no support for security, since they were based mostly on proprietary solutions (hardware and software), deployed on isolated islands with no communication, or through private communication infrastructure. Integrating those legacy systems to the IoT- based SG is a real challenge, since in most cases there is no way to replace them with new systems, or update them so they can support the desired security solutions.
- Constrained Resources
Several devices/objects of the SG, especially, those massively deployed are resource constrained. Special care need to be taken when developing security solutions, to be sure that their limited resources could accommodate the solutions. This make applying classical security solutions, especially, those based on public-key cryptography or on PKI, a challenge.
- Heterogeneity
Due to the discrepancy on the resources of the devices/objects on the SG (memory, computation, bandwidth, energy autonomy, time-sensitivity, etc.), and their implemented protocols and communication stacks (for non IP-based devices) achieving secure end-to-end communications is a challenging task, requiring the most often adaptation of existing solutions or even using gateways.
- Interoperability
It could be seen as one of the consequences of protocols and communication stacks heterogeneity, between devices/objects in the SG. Legacy system and devices/objects that couldn‘t support TCP/IP stack (ex, Zigbee v1, HART) couldn’t communicate with IP-based systems and devices or objects, unless through gateways, making end-to-end secure communication impossible. Interoperability could also be seen between two devices implementing the same protocols and communication stacks, but different feature capabilities: one with fully support, the other with partial support (ex, DTLS with/without certificate support)
- Bootstrapping
How to efficiently bootstrap the millions of devices or objects of SG with the necessary initial keying materials (cryptographic keys, cryptographic functions/algorithms and parameters, etc.)?
- Trust Management
Objects/devices on the SG could be managed by different entities (end-users for smart appliances, SG’s operator for smart meters and sensors, etc). Objects/devices couldn’t communicate if a minimal trust level isn’t established. While objects or devices owned or managed by the same entity could easily establish a trust relationship, building trust between objects or devices owned or managed by different entities is a challenge, especially in such large-scale network.
- Latency or Time Constraint
Some parts of the SG need to respond on a real-time basis to events and messages. For instance, electric SCADA (Supervisory Control and Data Acquisition) system, used on transmission and distribution sub-stations, must respond on a real-time basis to any variation on current, voltage or frequency values of the electricity in addition to other meteorological parameters influencing equipment’s functioning all provided by different kind of smart objects (sensors, actuators, etc. etc.), in order to keep the assets safe and prevent the propagation of anomalies (power overload or outage) to other parts of the power grid. This makes time-consuming operations (i.e. public-key operations) not suitable.
Security Services for IoT-Based Smart Grid
Hereafter, we briefly list the major security services that should be considered for the IoT-based SG:
- Authentication
The capability to check/ensure the identity of any communicating device/object/ in the SG. For instance, the energy provider needs to authenticate each smart meter, in order to bill the corresponding user.
- Data Integrity
Ensures that (received) data were not modified in an unauthorized way. For instance, smart meters need to ensure the integrity of a software update, in addition, to source origin.
- Confidentiality
Ensures that data (stored or transmitted) is accessible only to the intended recipients. For instance, end-users’ consumption need to be known by the SG’s operator and the energy provider only.
- User’s Privacy
Guarantees that any data related to the user (energy consumer end-user) – brut, inferred or computed data- could not be obtained without its explicit approval, and will be used only for the intended purposes. For instance, energy consumption data used for billing purpose couldn’t be used for other purposes
- Authorization and Control Access
Guarantees that an authenticated object or person, is authorized to accomplish some tasks, or has been granted the necessary rights to access some resources. For Instance, an on-the field agent needs authorization and access control rights, to perform manual configuration on a smart meter.
Internet of Things, is the next step towards a globally and pervasive connection to any communication and computation enabled objects or devices, regardless their access technology, available resources and location. The Smart Grid can highly benefits from the IoT vision, where smart objects/devices are deployed alongside the energy path, from the generation plant to the end-customer. However, security is the main concern for the IoT, and the large-scale adoption and deployment of the SG.
In this paper, we briefly reviewed the main security issues and challenges for the SG, and dressed the major required security services. In the future, we will study on-depth the security of a key-element of the SG, which is the AMI, where we focus on how we can securely integrate energy aware smart home, equipped with smart meters and smart appliances, in the SG, so that end customer could actively and securely participate in the energy consumption or production equilibrium.
If you want to share thoughts or feedback then please leave a comment below.
