Wireless Sensor Network Based Smart Grid Communications: Cyber Attacks, Intrusion Detection System and Topology Control

: The existing power grid is going through a massive transformation. Smart grid technology is a radical approach for improvisation in prevailing power grid. Integration of electrical and communication infrastructure is inevitable for the deployment of Smart grid network. Smart grid technology is characterized by full duplex communication, automatic metering infrastructure, renewable energy integration, distribution automation and complete monitoring and control of entire power grid. Wireless sensor networks (WSNs) are small micro electrical mechanical systems that are deployed to collect and communicate the data from surroundings. WSNs can be used for monitoring and control of smart grid assets. Security of wireless sensor based communication network is a major concern for researchers and developers. The limited processing capabilities of wireless sensor networks make them more vulnerable to cyber ‐ attacks. The countermeasures against cyber ‐ attacks must be less complex with an ability to offer confidentiality, data readiness and integrity. The address oriented design and development approach for usual communication network requires a paradigm shift to design data oriented WSN architecture. WSN security is an inevitable part of smart grid cyber security. This paper is expected to serve as a comprehensive assessment and analysis of communication standards, cyber security issues and solutions for WSN based smart grid infrastructure.


Introduction
The electrical grid is being revolutionarily transformed as Smart grid. Smart Grid is an automated and broadly distributed energy generation, transmission and distribution network. It is characterized by full duplex network with bidirectional flow of electricity and information. It is a close loop system for monitoring and response [1][2][3][4]. Smart Grid is being conceptualized and developed by various organizations around the world such as National Institute of Standards and Technology (NIST), Institute of Electrical and Electronics Engineers (IEEE), European Technology Platform (ETP), International Electrotechnical Commission (IEC), Electric Power Research Institute (EPRI), etc. Diverse set of standards and harmonization between various standards are also being rigorously researched by these organizations [4]. It can be defined in various ways as per its functional, technological or beneficial aspects. As per the definition given by U.S. department of energy, "A smart grid uses digital technology to improve reliability, security, and efficiency (both economic and energy) of the electric system from large generation, through the delivery systems to from implementation of smart microgrid using personal area network based on IEEE 802. 15.4 standard in University of Pisa in [39]. The relation between intricacy and security has been pointed out by Schumacher et al.: "Complexity is the worst enemy of security" [40] (p. 531). The technologies and policies for alternative fuels and vehicles are discussed in [41] and [42].
Smart Grid is an interconnected, hierarchical and heterogeneous network with enormous complexities and dynamics. Thus, gigantic network architecture of Smart Grid creates many security challenges. However, the above-mentioned statement is divergent for WSN as it contains tiny sensor nodes with limited computational and storage capabilities. For WSN, simplicity of design and implementation becomes the worst enemy of its security and makes it more vulnerable to attacks.
WSN is an essential component of Smart Grid communication infrastructure. Overall cyber security concerns and solutions for Smart Grid network are widely described in literature. Attacks, vulnerabilities and Security requirements for WSN are different from the overall network security essentials due to design and limitations of sensor nodes. The address specific security attacks and solutions may not be applicable for data oriented approach of WSN as communication between sensor nodes is concerned with the data rather than the address of a specific sensor node. For example, readings of humidity and temperature of a specific area can be sent by redundant nodes. Thus, the security of redundant sensor nodes should also be considered. Security of WSN is one of the most vulnerable, complicated and challenging aspects of Smart Grid security. It can be considered as a base of entire Smart Grid network security. This paper is focused on WSN attacks, detection methodologies and security aspects. It is expected to serve as a reference for design of secured sensor network for Smart Grid applications. Authors of this paper have described a layered communication architecture of Smart Grid, importance and complexities of WSN security as well as topology control methods for state of the art readers.

Application of Wireless Sensor Networks in Smart Grid
WSN is a cost effective solution for monitoring, control, measurement and fault diagnosis in various domains of smart grid network. A sensor node mainly contains sensors, memory, processor, power supply, transceiver and actuator. Sensors are used to sense various quantities like humidity, temperature, current, etc. Generally, WSN nodes are battery powered. Figure 1 shows the basic structure of WSN node. WSN facilitates both sensing and communication requirements [21][22][23]. Small sensor nodes collectively form a sensor network which is used for remote wireless communication in HAN, NAN and WAN [21]. Large scale deployment of sensor nodes can be used to communicate the conditions of various generation, transmission and distribution units. Wireless sensor nodes can provide economical solution for smart microgrid monitoring which facilitates high penetration of renewable energy sources. WSN is a significant part of advanced metering infrastructure (AMI). Sensing and communication are crucial for Plug in Hybrid Electric Vehicle (PHEV) system which is one of the most ingenious component of smart grid technology. PHEV contains gasoline or diesel engine with an electric motor as well as a large rechargeable battery. PHEV can be recharged from an electrical power outlet. It has the potential to reduce Green House Gases (GHG) emissions and carbon footprint [38][39][40]. PHEV facilitates flexibility as well as economy in fuel usage [41][42][43]. WSN can be used to communicate PHEV statistics to upstream network layers for operation and control of Smart Grid components. This information will be online available to various stakeholders through a web of sensor nodes [44][45][46][47][48]. An effective remote monitoring, diagnosis and control can prevent cascaded disastrous events and breakdowns [49][50][51]. Wireless sensor networks can be used for accurate monitoring of generation, transmission, distribution and consumption of electricity [52][53][54][55].
WSN is the most suitable solution for HAN, NAN, WAN and smart microgrid applications for integration and operation of renewable energy sources [54][55][56][57][58][59]. Figure 2 shows the application of WSN at different levels of smart grid. Wireless sensor networks are applicable for the following smart grid utilities:  Smart power generation Wireless sensor networks can be used at the generation side for monitoring and management of produced energy. It is a prominent solution for smart microgrid applications using renewable energy resources. WSN can be used in solar farm, wind farm, biogas plant, etc. to monitor and control intermittent energy. One of the objectives of smart grid is to expedite the use of renewable energy sources. Renewable energy resources are situated in harsh environments and hostile locations. Moreover, their unpredictable behavior creates more challenges during their operation and management. WSN nodes are economical solutions for monitoring the behavior of renewable energy resources [55][56][57][58]. Various parameters of generating equipment can be effectively measured, communicated and controlled using WSN.


Smart power transmission and distribution Transmission and distribution of power contains various components like overhead transmission lines, underground cable network, substations and distribution transformers. WSN is an essential element of SCADA system. Real time remote monitoring of these components is inevitable to prevent power failures due to equipment breakdown or malicious attacks. Wireless sensor networks can be used for power monitoring, fault detection and isolation, location discovery and outage detection [56][57][58][59].


Customer applications Wireless sensor network is an effective and prominent solution for home automation systems. It can be used for complete energy management of customer premises. Consumer plays an active role in smart grid technology. Consumers have the power to decide the time of use and rates of energy usage in smart homes [56][57][58][59]. For these applications, wireless sensor networks are inevitable for communication and processing of information. WSN is the backbone of smart home applications and HAN [57].

Challenges of Wireless Sensor Networks in Smart Grid Applications
WSNs are a vital part of self-healing smart grid network as sensor nodes communicate parameters pertaining to conditions of various equipment and energy sources. However, there are many challenges in deployment and operation of WSN due to limitations of sensor nodes and complexity of heterogeneous smart grid network [53][54][55][56][57][58]. The challenges are summarized in Table 1.

Severe ecological conditions
Wireless sensor nodes can be subjected to harsh environmental condition which may cause fault in wireless sensor node.

Various network topologies
Heterogeneous network topologies in energy distribution network due to various features and failure of sensor nodes may cause technical challenges in design of sensor nodes.
Limited capability Restricted processing and memory capabilities cause various challenges in design and deployment of wireless sensor networks.

Bit errors
In communication systems, high bit error rates are observed due to high noise level. This calls for various error detection and correction schemes. Detection and correction of errors require greater memory and processing facilities which make the design of sensor network challenging.

Security of sensor networks
Security of wireless sensor network is an indispensable and decisive requirement. The sensor nodes must be secured from physical tampering to hacking for smooth functioning of various smart grid applications. Physical tampering is also called node capture.

Quality of service necessities for smart grid environment
Zigbee is based on IEEE 802.15.4 standard. It is an energy proficient short range wireless communication technology. It functions in the ISM (Industrial, scientific and medical) band which is allocated for industrial, scientific and medical applications. Zigbee operates in the band of 2.4 GHz, 868 and 928 MHz with full duplex wireless data transmission. IEEE 802.15.4 standard describes physical layer and media access layer and Zigbee Alliance has expanded the configuration of an application layer and network layer. The maximum throughput achievable by Zigbee is 250 Kbps [51]. In the area of power automation, it is applicable for smart meters, power system monitoring and measurement of various electric parameters. Smart Grid integrates information and communication technology with an existing power system to improve the power grid network with the capabilities of self-healing, disaster recovery, interoperability, compatibility, energy efficiency and security [52][53][54][55]. Zigbee can play an imperative role in operation and maintenance of power grid, data accumulation, parameter measurement, security, monitoring and consumer interface [59].

Bluetooth
Bluetooth is a short distance wireless communication technology based on IEEE 802.15.1 standard. It uses short wavelength wireless transmission in the unlicensed ISM band from 2400 to 2480 MHz. It uses frequency hopping spread spectrum (FHSS) technology and around 1600 hops per second. Its key features are extensive availability, low power consumption and rapid data exchange. Bluetooth was initially developed in 1994 by Ericsson and then a group of firms formed a special interest group to retain and improve this technology. There are two network topologies used in Bluetooth which are termed as Piconet and Scatternet. A Piconet is created by a Personal Area Network in which one wireless client acts as a master and other wireless clients serve as slaves. Maximum eight devices can communicate with each other in one Piconet. A Scatternet is an arrangement of group of Piconets. Bluetooth is used for communications between smart consumer appliances, energy management system and smart meters. It has peak data throughput of 1 Mbps, 79 radio frequency channels and channel bandwidth of 1 MHz. It has nominal range of around 10 m. Bluetooth comprises of three power classes each having a different range [59].

Z-Wave
Z-Wave protocol is specifically designed for smart home applications. It can be adopted in Home area networks of smart grid. Z-Wave is a low data rate, short range radio frequency mesh networking standard operating on 908 MHz band. The maximum coverage area is 30 m indoor and 100 m outdoor. It does not require central coordinator but employs master and slave nodes. It can support 232 devices. The data rate is from 9.6 Kbps to 40 Kbps [59].

WirelessHART
WirelessHART protocol is designed for industrial automation. It is a real-time, centralized and multi-hop mesh network developed for industrial monitoring and control applications [60][61][62]. It is based on IEEE 802.15.4 compatible radio and operates on 2.4 GHz ISM band. It uses direct sequence spread spectrum (DSSS) technology. Besides DSSS, it uses Time Division Multiple Access (TDMA) technology in which 10 ms time slots are allocated to nodes. The range of this technology is up to 200 m. Security of communications is maintained using 128 bit AES encryption [62]. Individual session keys as well as common network encryption keys are shared among all nodes for broadcast services. WirelessHART overcomes all the shortfalls of Zigbee in terms of robustness, reliability, security and message delivery [60][61][62][63][64].

Wavenis
Wavenis is an emergent wireless communication technology for low power machine to machine communication applications. It can be used for distances up to 200 m for various indoor applications. This technology can be used in various metering applications in smart grid. It can be used in automatic meter reading, advanced metering infrastructure and remote communication applications. Wavenis operates in the bands of 868 MHz, 915 MHz, and 433 MHz. The data throughput of Wavenis ranges from 4.8 Kbps to 100 Kbps [59]. Table 2 shows the comparative analysis of various communication standards for WSN.
Wireless sensor network (WSN) uses various short distance communication technologies as they are suitable for power efficiency and wide spread availability. These communication standards use unlicensed ISM band for their operation which makes WSN effective for communication but vulnerable to attacks.

Security Issues and Cyber Attacks in Wireless Sensor Networks
The wireless sensor networks used for smart grid applications have different characteristics than communication networks used for other generic applications. These characteristics are in terms of deployment topology, data processing, environmental conditions and network throughput. The security issues are related to confidentiality, authentication, availability, integrity, authorization and newness. Confidentiality deals with secrecy of data communication. Authentication is necessary for prevention of fake messages from malicious sensor nodes. It ensures data authenticity. Availability means consistency in services in presence of attacks. Integrity means the data or messages are received in an unaffected form at the destination. Authorization means only authorized sensor nodes can communicate and unauthorized access of data must be prevented. Newness of data is inevitable to ensure that attackers do not replay the old data again to hinder the security of WSN [65][66][67]. It is very challenging to ensure the stated measures of data authenticity and security in WSN due to following constraints mentioned in Figure 3. They are deployed either in fixed manner or in random manner. They are remotely managed and controlled [66][67][68][69].  Communication Constraints: WSNs communicate using radio transmissions and most of them use unlicensed ISM band which is used for many other applications. Co-existence of various wireless standards is a major challenge for secured communication.
Due to above reasons, sensor nodes are prone to eavesdropping or jamming attacks. Moreover, limited processing and storage capabilities of sensor nodes prevent the use of advanced cryptographic methods. Public key cryptography requires expensive computational techniques. Limited batteries of WSN make them prone to denial of service attacks which can further drain the energy from sensor nodes. Tampering of sensor nodes and reprogramming of chip are the possible attacks. The security of WSN can be endangered by reverse engineering [65][66][67][68]. Smart grid also employs various devices and communication protocols at different network levels. Thus, WSN interoperability and security with other devices is also a matter of concern. The protection and privacy of WSN must be treated carefully for smart grid applications which is a broad network comprising of enormous networks and protocols. The public and private security measures for WSN in various smart grid applications require costly solutions [67][68][69]. The cost-security tradeoff must be carefully implemented. Figure 4 shows the complete overview of applications, security issues and objectives of WSN for smart grid network. In addition to above wireless network security issues, WSNs are vulnerable to cyber-attacks on various network layers [70]. Figure 5 depicts the classification of various cyber-attacks in WSN.

External Attack
Most cyber-attacks are external in nature, where an intruder is outside the range of WSN. External attacks are performed by jamming the network, exhausting the resources or denial of service (DOS) [71][72][73][74][75].

Internal Attack
In this type of attack, an intruder is in the range of WSN. This type of attack is performed by physical tampering of node, revelation of confidential information, causing denial of service to authorized node etc. [71][72][73][74][75].

Active
This type of attack is performed by an adversary through modification and theft of data. In active attacks, an attacker performs some operation in order to hamper the confidentiality, availability and integrity of data. An adversary performs the attack by packet modification, false data injection, resource exhaustion and node capturing. Spoofing, jamming, wormhole, hello flood, DOS, sink hole, etc. are the examples of active attacks [72][73][74][75][76].

Passive
Passive attacks are performed by an intruder mainly through an observation of network activities [75]. This type of attack impedes confidentiality of the network. This type of attack includes analysis of traffic, decryption of vulnerable data, capture of information, etc.
 Physical devastation/Node capture of wireless sensor nodes Wireless sensor nodes are deployed and distributed at various places for various applications and there are possibilities of physical damage or destruction of nodes by attackers. Node capturing may result into alteration of hardware and software of WSN node [75][76][77].


Utilization of network resources by intruders and making the scarce resources unavailable Attackers or intruders can exploit the limited network resources and make them unavailable for actual users.


Alteration of configurations of wireless sensor network by attackers The encryption and other WSN configuration aspects must be confidential and any alterations made by intruders make them unapproachable for genuine users. Jamming of network, camouflaging the wireless sensor network ambiance and physical attack on sensor nodes are common threats.

Misdirection Attack
In the misdirection attack, the information is routed towards fake path. It alters the routing information of network and affects the communication adversely. Misdirection is a network layer attack. Authentication techniques between transmitter and receiver, multi hop routing, etc. can be used to detect misdirection attack.

Selective Forwarding
Selective forwarding is a network layer attack. In this type of attack, a counterfeit node acts like an actual node and divert the packets to a wrong path but selectively drops some of the packets so that it becomes difficult to identify the intrusion. Acknowledgement based routing, multi data flow and detection based on neighboring information can be used to detect this type of intrusion [70][71][72][73][74][75].

Sink Hole Attack
Sink hole attack is a data link layer attack. In this attack, an intruder comes with an agreement with a sensor node or introduces a fake node in the sensor network. When a forged node attracts the network traffic, an attack is generated. Once the attack is successful, the forged node can perform various malfunctions like dropping all packets, dropping selective packets and alteration of data [70][71][72][73][74][75][76].

Sybil Attack
In Sybil attack, a malicious sensor node takes multiple identities to perform an attack. In wireless sensor network, all the sensor nodes work complaisantly but this type of attack targets this cooperation and disturbs the routing and communication process [76][77][78][79][80][81][82][83].

Wormhole Attack
Wormhole attack is a data link layer intrusion. In this type of attack, a malicious or fake node registers all the information and diverts it to wrong path. This attack can be formed without the knowledge of cryptography of actual wireless sensor node [84][85][86].

Hello Flood Attack
In wireless sensor networks, routing protocols use Hello packets for detection of neighbors. In this type of attack, fake packets are used to camouflage hello packets and to attract the sensor nodes [70]. Attackers with ample radio resources and processing capabilities can generate this type of attack. The victim node will identify false hello packet as normal node.
Various cyber-attacks concerning different network layers and their respective countermeasures are described in Table 3.

Intrusion Detection System
Encryption and authentication are unavoidable considerations for the security of any communication system. These parameters are vital for reliability, confidentiality and integrity of sensor network. Detection of attack or intrusion is also imperative to make the system robust against attacks. Intrusion detection system (IDS) can be termed as a subsequent security measure and defense mechanism against cyber-attacks [82]. IDS observes the network and identifies anomalous activities. IDS can be defined as a unification of hardware and software tools which are meant to detect internal or external cyber-attacks. IDS is essential for fault tolerance, security and reliability of WSN. IDS also investigates the physical tampering of sensor node. This ensures secured routing of information over WSN. The principle tasks of IDS are prevention and detection of attacks, situational awareness, evidence collection, and administration of connection topologies [77][78][79][80][81][82][83]. Security of WSN is more complex compared to Mobile and Adhoc Networks (MANETs) due to resource constraints in WSN [83]. This calls for an efficient IDS for secured WSN communication. The main components of IDS are as follows.


Sensor: It collects statistics from the system being monitored.  Detector: It analyzes collected data to identify intrusions.  Information Base: It supports the detector by providing attack signatures.  Response Manager: It manages the responses to the cyber-attacks. The general block diagram of IDS is shown in Figure 6. The strategies for attacks and detection system for WSNs are different from other wireless or wired networks due to their structure and limited battery life. Intrusion detection approaches for WSN are classified as follows [82][83][84][85][86].

Anomaly Detection
Different types of anomalies of WSN are node anomaly, network anomaly and data anomaly as described below.


Node Anomaly: This types of anomalies can be detected during failure of WSN node or power problems. Failure of solar panel, or fluctuations in power of different components can cause this type of anomaly. Node anomalies can be due to hardware or software issues in the WSN nodes [82].  Network Anomaly: Unexpected fluctuations in the signal strength and connection problems can be used to detect network anomaly. Complete loss of connectivity or episodic connectivity can be used to detect intrusions in the network.  Data Anomaly: An intrusion attempt can be detected from chaotic or disordered data communication.
Detection of a specific type of anomaly is very useful to decide the type and explication of cyber-attack. To detect the above-mentioned anomalies, different types of approaches are applied [86]. These anomaly detection approaches can be classified as game theoretic, statistical, machine learning, artificial immune system and data mining based approach [85][86][87][88].

Misuse Detection
Misuse detection is a signature based intrusion detection system to discover recognized attacks. The limitation of this type of detection system is that it cannot detect unknown attacks which are not predefined. Moreover, keeping signatures of attacks to generate data base is a complicated task for WSN due to its limited memory and processing capabilities. However, in very few studies, this method has been explored using watchdog approach [83][84][85][86]. Watchdog approach uses the abnormal behavior of a node to detect an intrusion. All nodes watch the performance of their neighbors and communicate the information about their behavior.

Hybrid Detection
An intrusion detection approach that does not qualify to be classified as either anomaly or misuse detection is called hybrid detection method. This approach is application specific and it is manually defined by an administrator. Hybrid approach can be a combination of anomaly and misuse detection approaches for accurate results [84][85][86][87][88].
IDS is inevitable for the security of WSN as it awakens the operators to take countermeasures against cyber-attacks. In addition to IDS, Intrusion Detection and Prevention System (IDPS) can also be developed for rapid and effective mitigation and avoidance of attacks [89]. In some instances, an intruder can be intelligent with detection avoidance competences. Moreover, range of communication, density of nodes, sensing algorithms, etc. play a crucial role in intrusion and its detection [90]. More research work is required against intelligent intrusion algorithms in order to maintain security, confidentiality and reliability of WSN.

Topology Control
Wireless sensor networks are the group of distributed sensor nodes which communicate various information for monitoring and control purpose. Sometimes placing WSN nodes in unsecured and hostile environments is required. Due to limited battery life and restricted processing and storage capabilities of WSN nodes, security against above mentioned attacks is the biggest challenge. WSN nodes are mandatory part of smart grid communication infrastructure. Apart from cryptographic approach, WSN can be designed in such a way that their topologies have specific connectivity properties [89]. Topology control can be a practical solution for WSN nodes with limited computational and communication capabilities [90][91][92][93]. An overview of various topology control schemes is described below.

Random Key Predistribution Scheme
This scheme is extensively recognized as an appropriate solution for secured WSN communication. There are two types of random key predistribution scheme.

Eschenauer-Gligor Random Key Predistribution Scheme (EG Scheme)
In EG scheme, there are n number of sensors in a keying network. This scheme uses an offline pool of keys containing Pn keys. Before deployment, each sensor is assigned Kn number of discrete keys selected from the pool of keys. Pn and Kn both are the functions of n for generality motives. Kn keys in each sensor establish sensor's key ring. After deployment, two sensors can establish secured communication link only if they have at least one key(s) in common i.e., 1 ≤ s ≤ Kn. Confidentiality and authenticity are achieved with symmetric key encryption mode [94,95]. Figure 7 shows the EG scheme.

s-Composite Random Key Predistribution Scheme
s-Composite Random Key Predistribution Scheme is better than EG scheme in a manner that it requires s overlapped key in order to establish communication between two sensor nodes. It requires s ≥ 2 for secured communication between sensor nodes. It is beneficial for small scale sensor attacks but becomes vulnerable for large scale attacks. s is selected according to desired resilience of the sensor network. Figure 8 shows the s-composite scheme.

Link Constraint Models
Various link constraint models are used to explore WSN using either EG or s-composite schemes. They are classified as follows.

Full Visibility Model
In this model, it is assumed that there is a communication link between any two sensor nodes in a network. According to this model, the two sensors can establish a secured communication link only if they have one key for EG scheme and s keys for s-composite scheme common among them [94][95][96][97]. This model requires mutual keys between sensor nodes which satisfy both the schemes.

On-Off Channel Model
This model contains independent channels, each of which is either on with probability Pn or off with probability (1−Pn), where Pn is a function of n for generality. This model requires the channel between two sensors to be on for communication [96][97][98][99].

Disk Model
This model requires sensor nodes to be within a specific radius rn to establish a communication link between them. For the node distribution, it is considered that all n nodes are uniformly and individually deployed in a bounded space of a Euclidean plane. Such network area A is either a torus T or a square S, each of unit area, depending on whether the boundary effect exists or not [98]. The boundary effect arises whenever part of the transmission area of a node may fall outside the network area A. T does not have the boundary effect, whereas S has the boundary effect [99][100][101].

Conclusions
The present power grid is going through a huge transformation with the deployment of smart grid technology. Smart grid is a complex hierarchical and heterogeneous network. Wireless sensor network is a prominent solution for various applications of smart grid. Wireless sensor networks are distributed collection of sensor nodes situated at various places for measurement and communication of various parameters such as temperature, voltage, current and humidity. These parameters are required for remote monitoring and control of different components of smart grid. WSNs are effective solutions for energy management system in home, industry and business applications. These small sensor nodes are extensively vulnerable to attacks as they are placed in hostile surroundings. Node capture results into complete control of attacker on the WSN node and tampering of hardware as well as software of the node. The energy exhausted sensor nodes can be easily victimized. Therefore, cryptographic security is essential to protect the communication between sensor nodes as well as to detect sensor capture and to invalidate the compromised security keys. The security of dispersed WSN nodes is a crucial technical challenge due to limited memory and computational capabilities of WSN nodes.
The data oriented design approach is required for deployment of WSN as the purpose of sensor nodes is to sense and communicate the parameters. Redundant nodes must be deployed to deal with node failures and degraded signal strengths. Traditional communication process involves physical address and IP address of a specific transmitter and receiver to establish a successful communication link. This type of communication is address oriented which is different from WSN approach. As an example, consider the measurement of average temperature of some area. In this case, temperatures from each and every node is not required but an average temperature can be calculated from received readings from sensor nodes positioned at various places. Identity of a specific node is secondary as soon as the readings from all areas are received. The challenges of WSN communication is different from the challenges faced by the usual communication networks. Rigorous and diversified research endeavors in the field of WSN security are inevitable as WSN forms the backbone of smart grid IoT applications. Moreover, various solutions for local and wide area networks are already available due to development of Internet technology. The security of gigantic network of tiny WSN networks is the most challenging aspect of smart grid reliability.
In this paper, the analysis of various wireless communication standards, cyber security issues and solutions for WSN are discussed. Nature of various attacks must be known for detection of attacks and development of different solutions at diverse network layers. Apart from well researched solutions such as IDS and cryptographic security, this paper explores topology control for cyber security of wireless sensor networks. Various solutions for intrusion detection and prevention system must be developed for inhibition of adverse effects of cyber-attacks. Secured interoperability between various communication standards is inevitable for robust hierarchical smart grid infrastructure. WSN security is a multi-faceted research topic due to limitations imposed by communication standards and sensor nodes. The cost-security tradeoff must be critically analyzed and implemented for future applications of wireless sensor networks in smart grid applications.

Conflicts of Interest:
The authors declare no conflict of interest.

Abbreviations
The following abbreviations are used in this manuscript: