Node Clone Detection Protocols for Protect the WSNs: A Survey

ABSTRACT


INTRODUCTION
The main security targets of WSNs include confidentiality, authenticity, data integrity, and availability [1].The network is unreliable and unsuitable for further connections when the enemy launches node-clone attacks, as all of these security objectives are affected.This is due to two reasons, the first of which is that proper detection protocols are not available and are not effective for identifying and nullifying attacks.Secondly, the probability offered by some detection systems is negligible [2].Much damage to the network due to a node cloning attack since it is considered a real node by its neighbor and can participate in network operations using encryption keys.The major purpose of the adversary's creation of these clones is to perform a variety of insider attacks, including network traffic monitoring, sensor spoofing, mock data injection, sabotage of data collection, signal jamming, denialof-service attacks, and disrupting network tasks [3].
The cloned node initially launches a variety of malicious attacks (also called malicious nodes) into the network.The cloned node contains its legitimate information (identifier and encryption keys) and is involved in network operations being a non-compromised node [4].Figure 1 shows the details [5].
To evaluate the performance of various node-clone detection protocols, many items are involved in the evaluation process.The main items include connection costs, storage expenses, security of data, probability and time-detection, cost and maintenance, power level and utilization, delivery-rate, end-to-end delay, service quality, packet loss, and so on.They are the following [4]: (1) Communication cost: It is the average number of messages sent by nodes when verifying site claims.
(2) Storage cost: It is calculated by the sensor node by the number of stored location claims.
(3) Data security: This refers to preventing unauthorized users from illegally using data.
(4) Likelihood of detection: How much a measure of the accuracy of a detection protocol in identifying and detecting clones.
(5) Discovery time: It is the rate of time that passes between the publication of a clone and its discovery in the network.
(6) Data Delivery Cost: It is the cost factor incurred to deliver the packet from the source node to the destination.(7) Energy efficiency: It is the minimum amount of energy used by a node to direct the beam to the desired location.
(8) Rate of Delivery: It is calculated by dividing the ratio of the number of packets received by total number of packets sent.
(9) Amount of lost packets: Due to congestion or network failure, some of the packets are lost.The number of packets that failed to arrive at the destination.(10) Avoidance: the canceling of something by an authority [4].
These selection criteria are based on their relevance to nodeclone detection protocol performance and their potential to highlight the advantages and disadvantages of various strategies.Considerations such as communication and storage costs are critical when examining a protocol's scalability, and data security and detection probability are essential for determining a protocol's efficacy in thwarting node-clone attacks.In a similar vein, assessing a protocol's performance with regard to power consumption and data transfer requires consideration of energy efficiency and delivery rate.All things considered, the assessment items offer a thorough framework for evaluating the efficacy of node-clone detection procedures and can aid in directing the choice and application of suitable strategies for thwarting node-clone attacks in WSNs.
The Wireless Sensor Networks (WSNs) security objectives are as follows [4]: (1) Availability: It ensures that network services are accessible even while under assault.The enemy attempts to undermine Provides network services by disrupting its operations due to a node-cloning attack.
(2) Authenticity: This term typically specifies the identities of the nodes taking part in network communication.It is challenging to recognize between a clone and an original/legitimate node as a result of a node clone attack because the clone has the same basic information as the parent node.
(3) Confidentiality: guarantees safe data transfer between trusted nodes.Due to node clone attacks, where the clonednode mimics the behavior of a normal node, people attempt to abuse the data carried through networks, turning private information into public information.
(4) Data integrity: It gives a guarantee of data reliability and immutability it is used to communicate between nodes.Because of the node-clone attack, the enemy can cram wrong data, reprogram node code, forge data, and so on, which makes the data unreliable for transmission [4].
Node cloning compromises the security objectives of WSNs in several cases: a.It undermines the availability of network services by disrupting its operations due to a node-cloning attack.
b.It challenges the authenticity of the identities of the nodes taking part in network communication, making it difficult to differentiate between a clone and an original/legitimate node.
c.It compromises the confidentiality of data transfer between trusted nodes, as node clone can mimic the behavior of a normal node and abuse the data carried through networks, turning private information into public information.
d.It impacts the data integrity of the network by cramming wrong data, reprogramming node code, forging data, and so on, which makes the data unreliable for transmission.

CATEGORY OF SELECTION CRITERIA
The four types of detection systems are used to gather and validate clone evidence.These are the type of device, the mechanism of detection, the deployment techniques, and the detection range.Each was grouped as depicted in Figure 2 [6].The type of device is further divided into (static, mobile, and hybrid) in the selection criteria, while the detection mechanism is divided into (centralized, distributed, and hybrid).Finally, deployment tactics are once more divided into two categories: random uniform and grid.The detecting range is finally divided into entire and local categories [6]. Figure 2 shows the details.

Type of devices (Static, Mobile, Hybrid)
The sensor network is classified according to the type of devices as static, mobile, and hybrid in nature.Sensor nodes are randomly deployed in the static state, and their location does not change after deployment and is called (static).As for the mobile, the sensor nodes move on their own even after they are deployed, and by controlling their movement, they interact with the physical environment [6].From my point of view and other reverences such as [7], a third type can be created, which is a hybrid WSN, which is a mixture of static and mobile devices in the same network.It can be used in a single network if the environment is a forest where the network monitors fire occurrences, or in a battle environment that requires, for example, a temperature sensor.Additionally, we might need a mobile drone equipped with a camera to film the scene and send the footage to the server, resulting in a hybrid Wireless Sensor Network (WSN).Static devices have the advantage of being more accurate and dependable than mobile ones.A mobile gadget, on the other hand, is more adaptable and location-neutral.The benefits of both stationary and mobile devices are combined in a hybrid device.Using a stationary device has the drawback of being less flexible than using a mobile one.Because of its mobility, a mobile device might be less accurate, and because of its complexity, a hybrid device might be more expensive.

Methodology of detection (centralized, distributed and hybrid)
In the centralized network, a node is responsible for the security of the network, but in the distributed network, each node is responsible for its security.The above two types can be combined into one network called a hybrid [8].Using a centralized system has the benefit of being simpler to administer and control.On the other hand, a distributed approach can manage large-scale systems and is more faulttolerant.The benefits of both distributed and centralized methods are combined in a hybrid mechanism.A centralized method may have a single point of failure, which is a drawback.There could be more complexity and communication overhead in a distributed system.In previous research, comparisons were made between centralized and distributed technologies in terms of: cost, power consumption, Accuracy, Dependency on additional hardware and Deployability [9].A hybrid system combines the benefits of distributed and centralized systems.
In some situations, it may be more efficient than a distributed or centralized system, including: (1) Constraints on resources: A hybrid system can effectively spread the load between its dispersed and centralized components in Wireless Sensor Networks (WSNs) with limited resources.
(2) Scalability: Due to its capacity to split the workload among several nodes and lessen the stress on individual nodes, a hybrid system can scale more effectively than a centralized or distributed system.
(3) Fault tolerance: Because a hybrid system can keep running even if some of its nodes or components fail, it can offer superior fault tolerance than a centralized or distributed system.
(4) Energy efficiency: Because a hybrid system may optimize each node's energy usage while lowering overall energy consumption, it may be more energy-efficient than a centralized or dispersed system.

Deployment strategies (Random, Grid)
The grid will perform better upon deployment than the random.The grid deployment ensures non-determinism and is useful in shielding the enemy from intelligent attack.The study is made simpler by the grid-based torus structure; It is a network diagram that dominates all directions north, south, east and west.A grid-based implementation offers great connectivity and resilience.The random deployment situation in some protocols results in significant collision probability and consequently somewhat high storage costs [6].Using a random uniform deployment strategy has the benefit of being straightforward and simple to put into practice.On the other side, a grid deployment strategy is more organized and has a wider coverage area.A random uniform deployment strategy has the drawback of potentially missing some places.A grid rollout strategy could be more expensive and complicated.A random uniform deployment approach could be preferable to a grid deployment approach in specific circumstances.For instance, a random uniform deployment approach might be more successful in covering all areas if the deployment area involves obstacles or is shaped unevenly.Furthermore, compared to a grid deployment method, a random uniform deployment technique might be more affordable and simpler to execute.Nonetheless, a grid deployment method might outperform a random uniform deployment strategy when it comes to preventing node clones.This is so that node clones may be detected and prevented more successfully.A grid deployment method offers improved coverage and connection.Furthermore, because a grid deployment method is less susceptible to sophisticated attacks and more structured, it can offer greater resilience against them.

Detection range (whole, local)
Since location claims are relayed to many zones and a strong attacker could corrupt an entire zone, the WSN network necessitates a higher communication cost.The localization strategy necessitates focused attention on the local area with no need to consider the network as a whole.Thus, the cost of communication and computation may be decreased [6].It can cover a bigger area and detect more clones when the entire detection range is used.In contrast, a local detection range can yield more precise results because it is more targeted.Using the whole detection range has the potential drawback of increased complexity and expense.Certain clones outside of a local detection range may go unnoticed by the range.There is a risk of missing localized or specific events, as the detection range may not be focused on specific target locations [10].In WSNs, there are trade-offs between whole-network and local detection algorithms in terms of resource consumption, accuracy, and scalability.Although whole-network detection techniques might offer more thorough coverage, they might not scale well for big WSNs and might need more resources.Although local detection techniques might not cover the whole network, they might yield more accurate results and aid in resource conservation.The particulars of the application and the resources at hand determine which of these approaches is best.

CLONE DETECTION PROTOCOLS
In this study, we will discuss the detection methodology of protection (sometimes called techniques, methodologies, or schemes) from node cloning attacks in WSNs from literature, which are of two types, central and distributed, as follows:

Clone-detection protocols in static WSNs
There are many protocols designed to detect replication of nodes in static WSNs that can be classified into work-centric and distributed-work protocols [5].

Centralized clone detection protocols in static WSNs
These technologies rely mainly on a strong base station "BS" regardless of being complex and having low overhead costs, for decision-making and information convergence, nodes send their location claims to the base station with the help of their neighbors.If a single identifier is found in more than one location and when the base station verifies the node identifiers, a cloning attack alert message is generated.These protocols are able to discover cloned nodes.But the sensor information remains unsecured, as the enemy can perform sabotage operations and spy on the information transmitted between the sensor node and the sink.In addition, the life of the network ends quickly because the nodes close to the basin node lose their energy quickly, depending on the capacity of their battery.Central detection techniques for static WSNs can be classified into one of the categories listed below, and their comparison is shown in Table 1 [5].

Distributed clone detection protocols in static WSNs
Replication detection differs from centralization in that each node in the network is responsible for its own security, which means that there is no central node of the authority designated to do the work.Even nodes in remote locations in the network participate in this task.Focusing on static wireless networks, there are several different types of detection techniques or schemes that we will mention in detail below and their comparison is shown in Table 1 [5], and Figure 3 shows the details of protocols.Table 1 explains the details of Figure 3, which includes some of the protocols proposed by the researchers within the mentioned sources, each according to his calculations for the purpose of preventing or repelling exposure to the attack of cloning nodes in WSNs, which we highlighted some of the important ones, as well as we mentioned their costs, advantages and disadvantages, so that the researcher does not make mistakes and also develops Its work is based on it, and it includes protocols for static node networks.

Clone detection protocols in mobile WSNs
Recently, mobile nodes have been used significantly within WSN.Because it plays a major role in implementing some needs in the network, in which it is needful for the nodes to be mobile to solve problems and provide many advantages over static wireless networks.Since static network protocols are not feasible and ineffective for detecting clones in mobile nodes, it has become necessary to develop and study some techniques for mobile WSNs to detect cloned nodes.These technologies are categorized into two main categories, centralized and distributed, and are detailed in Table 2.

Other clone detection protocols in WSNs
The way these protocols work is hybrid (centralized and distributed at the same time) and does not depend on any of the rules that we have discussed previously.Table 3 below explains more.[11] Fast [14] by: Ho et al.Scheme [11]

O(𝑛√𝑛 ) O(n)
A mobile node should never move at speeds greater than the maximum speed configured by the system.

This protocol does not carry the costs of the current generation of
Wireless Sensor Networks because it uses much more expensive equipment called GPS [14].

E(X))
Low memory cost because sensor node location information is not needed, only persistent communication is required [42].

THE POTENTIAL IMPACT OF THESE PROTOCOLS ON THE LIFESPAN OF A MOBILE WSN AND THEIR COMPATIBILITY WITH EXISTING NETWORK INFRASTRUCTURES
Node-clone detection algorithms have the potential to significantly affect how long a mobile WSN lasts.In the context of mobile WSNs, a number of protocols and procedures have been proposed to increase the network lifetime.To extend the lifespan of a network, these techniques include the usage of mobile relays, mobile sinks, and mobile sensor relocation [64].One way to extend the lifetime of a network is to deploy mobile nodes to take over the sensing and relaying duties of bottleneck nodes.For example, when a mobile node travels to co-locate with a bottleneck node and handles the bottleneck node's transmission and reception activities, the bottleneck node can sleep to conserve energy, extending its lifetime and enhancing the longevity of the network as a whole.To extend the lifespan of mobile WSNs, dynamic optimization of sensor node communication activity has also been investigated.Moreover, strategies to balance energy consumption and increase network lifetime have been studied, including the use of multi-path routing techniques and the exploitation of node mobility in mobile WSNs [65].The problem is that most research is not tested protocols within real WSN environments as a testbed or publishing operations, but only simulation, we hope in the future the testing should be real instead of using simulation.

CONCLUSION
This paper presents a review of previous research to compare and categorize some of the detection protocols for cloned nodes, highlighting their advantages and disadvantages, as well as costs in terms of memory, connectivity, and detection methodology.We have added new protocols alongside the previous ones, extracted their advantages and disadvantages through the three tables, and classified them according to the common classification in previous research.The classification was modified with simple additions from previous research as well, and this is what distinguishes our review from previous reviews.We have added new protocols alongside the protocols.We extracted its advantages and disadvantages through the three tables, and classified them according to the unified classification, and this is what distinguishes our review from previous reviews.We found that almost all detection schemes suffer from high communication and storage costs, but they still provide a high detection rate because such large costs affect the life of the network, especially when nodes do not have many capabilities to bear it.The types of sensors were also highlighted, which are central, distributed, and hybrid (a mixture of the two types).In static nodes, a node can be located once during initialization when sensor nodes are statically deployed.But, when the sensor nodes are mobile, they must periodically obtain their locations as they travel and this is one of the challenges faced by the mobile nodes.Therefore, they need to increase the time, energy, and speed of speed of execution.Presenting the significance of these findings for further study, the creation of fresh detection algorithms, or the area of mobile WSNs in general could improve the survey.For instance, the survey can go over how to overcome the difficulties found by creating new detection techniques or enhancing the ones that already exist.The impact of the results on the design of mobile WSNs and their applications might also be covered in the survey.By outlining these ramifications, the survey may offer a more thorough comprehension of the importance of the issues raised and their possible influence on the area of mobile WSNs.In the future, it is anticipated that the issues of high prices and network life impact in WSNs will be resolved by future technological developments in areas like energy efficiency, sustainability, and integration with edge computing and artificial intelligence.These advancements will support WSNs' long-term viability, sustainability, and efficiency, opening the door for further development and broad use.Including such research will affect the quality of future research, and benefiting from previous mistakes and not falling into them will enhance the improvement of energy use and speed of communication between nodes, and most importantly, it will increase the security of the network from cloning, due to the great need people have for Internet of Things networks, the most important of which are the medical, industrial, agricultural, and military fields in Nowadays, it is necessary to pay attention to its security in order to further enhance its quality in industries.

Table 2 .
Clone node detection protocols in mobile WSNs