A nature of WSN makes them vulnerable to various kinds of attacks. Vella [130] men- tioned that data obtained by sensing nodes need to keep as private and confidential. Without security mechanisms or security primitives, the malicious users could intercept the private information or send false messages to the neighbour nodes in WSNs. Security attacks can be categorised into two; passive and active.
2.2.1 Passive Attack
The listening and monitoring over the communication channels by the unauthorised and the malicious users are named as passive attack. Sensor nodes can sense and collect data from the environments in WSNs; as a result the network becomes vulnerable to potential abuse of these data resources. Since collected data are stored in the sensor nodes without tamper-proof or tamper-evident equipment, the privacy and confidentiality of data become important issues to protect from the passive attacks. The malicious users can gather in- formation by using passive monitoring and listening in anonymous manner. Some of the common passive attacks [120] in WSNs are explained as follows:
• Eavesdropping and Passive Monitoring
One of the most common forms of attack on privacy of data is eavesdropping and passive monitoring. If data messages are not encrypted, the adversary can easily understand their content and obtain information.
• Traffic Analysis
Traffic analysis can be performed to intersect data messages in order to analyse in- formation from the pattern in a communication channel, even if these messages are encrypted with cryptographic keys. The greater the number of data messages has been observed and intercepted by a malicious user, the more he can infer from the traffic. Eavesdropping is more effective when it is combined with traffic analysis.
2.2.2 Active Attack
The monitoring, listening and modifying the data stream or message in the communication channel is known as active attack. In active attack, an adversary can maliciously disturb the communication channels between the sensors nodes. In harmful active attacks, the adver- sary can alter and spoof the data packets and messages. In addition, it can even interfere
2.2 Security Vulnerabilities in WSNs 11 with the wireless signals to jam the network. Ng et al [100] argued that even if the infor- mation is protected from eavesdropping by means of encryption, the attacker may blindly modify that encrypted information and turn the information into meaningless information. The following active attacks [135] are most common in WSNs.
• Camouflage
A malicious user may try to compromise a sensor node and then use that masquer- aded node as a normal sensor node in WSNs to advertise false routing information. The camouflage node tries to attract other sensor nodes regarding packets forwarding. After data packets start receiving at the camouflaged node, it will forward to other powerful sensor nodes to analyse systematically.
• Sybil Attack
A particular harmful attack against sensor and ad hoc network is known as Sybil attack [99], where a node illegitimately claims to multiple identities. In Sybil attack, a malicious node presents multiple identities and sends incorrect information such as routing information, resource allocation, etc. to neighbour nodes in WSN. This means that a malicious user uses an identity of multiple nodes and routes multiple paths through a single malicious node. The sensor node authentication and encryption scheme can help to prevent that kind of attack in WSN [102].
• Wormhole Attack
Since the sensor nodes use a radio transmission medium to send information, the ma- licious users can eavesdrop the packets, tunnel them to another location and retransmit them in WSNs [73]. This kind of attack can make false information that the original sender of the packets is in the neighbourhood of the remote location. There are two proposed countermeasures to detect wormhole attack in WSNs called MDS-VOW [135] and Wormhole Attack Defence Mechanism (WODEM) [147].
• Replay Attack
A malicious user might use an old data message and attempt to send it at a later time for data access. When the sensor nodes receive that data message, they believe that it is an original message from the authorised user. Therefore, the malicious user can access to that data resource by using the old message. Freshness can be used in the query and request message to prevent replay attacks in WSN.
12 Literature Review of Current WSN Access Control Models • Hello Flood Attack
Hello Flood attack is introduced in WSNs by Karlof and Wagner [63]. In Hello flood attack, an attacker (laptop class user) uses high-power transmitter and advertises Hello message, which contains high quality route to the destination node or base station, to their neighbour nodes. It may cause a large number of sensor nodes to use that faulty routing information because the sensor nodes which receive Hello message, might attempt to transmit via that attacker node. In reality, the attacker nodes are out of the radio range of sensor network and far away from these neighbour nodes but the attackers use high-power transmitter to pretend that they are nearby.
• Sinkhole Attack
A malicious node attempts to attract all the traffic from a particular area to go through it. A malicious node might use incorrect routing tables to attract all the traffic from neighbouring nodes or nodes from specific area. The result is that these nodes are chosen a malicious node as next hop node to route their data packets through. This type of attack makes a selective forwarding message and all the traffic in the sensor network should flow through a single malicious node [120].
• Daniel of Service Attack
In WSNs, the attacks on availability of communication channels and data resources are referred to as Daniel of Service (DoS). WSNs are vulnerable for DoS attack due to resource limited and energy constrained. In DoS attack, an adversary usually attempts to disrupt, corrupt or destroy a network. Wood and Stankovic [141] discussed that DoS attack acts like an event that attempts to reduce a network’s capacity, to perform its expected function. In the published literatures, most of the defence mechanisms for DoS attacks need high computational overhead. This means that the defence mech- anisms from other wireless technologies are not suitable to apply in resource limited WSNs. There are several common techniques in the published literatures to cope with some of the common DoS attacks but the development of defence mechanism against DoS attack is still an open research issue in WSNs.
• Node Replication Attack
In node replication attack, an attacker tries to add a new sensor node in the existing network in order to do that it replicates and copies node identity of current sensor node [120]. Node replication attack can cause several disruptions in the communica- tion channels by forwarding and corrupting the data packets over incorrect routes. If a
2.3 Security Requirements in WSNs 13