Analysis the Effect of Dynamic Clustering and Lightweight Symmetric Encryption Approaches on Network Lifetime in WSNs


 Energy consumption and security efficiency are still main challenges in Wireless Sensor Networks (WSNs) due to their hardware resource-constrained nature. The 6LoWPAN protocol was developed to improve WSNs communication, security, and node management optimization. Hence the protocol energy efficiency and security can be improved. In this paper, we address the WSN nodes’ power consumption by analyzing the Dynamic Cluster Head (DynCH) technique, which automates the process of selecting WSN Cluster Head (CH) nodes based on the WSN nodes’ energy and nodes’ distances among each other in mobile WSN nodes. Moreover, this analysis covers the complexity of DynCH in different environments to prove its efficiency compared to the steady CH mechanism. In addition, we present the performance analysis of different lightweight systematic block encryption algorithms along with DynCH scheme on WSNs lifetime. In particular, Speck128, FlexenTech, Tiny Encryption Algorithm (TEA), and Advanced Encryption Standard (AES) algorithms are used in order to determine the amount of energy consumed by the sensor nodes and their effect on the network lifetime. Furthermore, the lightweight key management mechanism is used to secure the data and keys exchange between WSN nodes in all different systematic encryption algorithms. The Cooja simulator with Contiki operating system are used to evaluate our analysis. Finally, the outcome of the analysis has showed that DynCH improves the wireless network lifetime by 45% compared to the steady clustering approach. Moreover, the analysis also shows that, Speck128 consumed 26%, FlexenTech consumed 52%, TEA consumed 65%, and AES consumed 78% of wireless network lifetime compared to unsecure wireless networks communication, respectively.

addition, key-management, which includes creating, updating, saving and allocating the encryption keys on the sensor nodes, is another crucial process for securing WSN. In symmetric encryption, the secrete key, which is used for both encrypting and decrypting data are used to build the secure communication link between the WSN nodes [24]. Therefore, it is important to protect the key and to exchange it securely within the WSN. In many circumstances, WSN requires security and availability, therefore, this paper analyses the performance of the DynCH algorithm [14] for MWSN while using several lightweight symmetric encryption algorithms to see their effect on the lifetime of WSN networks. A summary of the main paper contributions are as follows: 1. The performance of the DynCH clustering algorithm, specially designed for MWSNs, is analyzed and compared with the static clustering technique. The obtained simulation results show the e ciency of the DynCH algorithm.
2. The performance of using various lightweight symmetric encryption algorithms in WSN nodes is analyzed and compared with each other by in uencing the lifetime of WSNs and to determine the best implementation in 6LoWPAN.
The rest of the paper is organized as follows: Sect. 2 presents background information and summarizes the most related work. Section 3 discusses the DynCH dynamic clustering algorithm and the lightweight symmetric encryption algorithms. Section 4 presents the simulation environment, setup, results and the performance metric. Section 5 provides a discussion of the attained simulation results under different environment and using different encryption algorithms, Finally, Sect. 6 concludes the paper and draw future work.

Background And Related Works
The purpose of this section is to provide the technical background and literature review of topics of interest to this paper. Particularly, the 6LoWPAN Protocol, clustering and cryptographic algorithms in WSNs.

6LoWPAN Protocol
6LoWPAN is a protocol built on top of the IEEE 802.15.4 utilizing the 2.4 GHz spectrum and standardized by the Internet Engineering Task Force (IETF), which is designed to support low range, power, memory usage and cost. Thus making it very suitable for WSNs. The prime feature of this protocol comes with its capability to support the IPV6 stack, which makes it capable of connecting to other types of wireless network nodes that support IPv6 such as Bluetooth, WiFi and sub-1 GHz low power radio frequency [25] by using an edge router. Figure 1 depicts a WSN running the 6LoWPAN protocol.
In order to further reduce the WSN running 6LoWPAN protocol, two modi cations occurred on the IEEE 802. 15.4 protocol, the e and g, i.e. IEEE 802. 15.4e/g. Moreover, reliability between WSN has been improved by running a set of rules to manage and control data exchange between the nodes.
Furthermore, the 6LoWPAN physical (perception) layer is responsible for transforming the digital bits into analog wave to be transmitted over the air.

Clustering Technology in WSN Networks
In order to prolong the WSN lifetime and preserve the nodes' energy, clustering is often used, where nodes are grouped into clusters. Each cluster elects a Cluster Head (CH) node, which is responsible for gathering the sensed data from the sensor nodes, aggregating and compressing them, then send them to a remote site for data collection, more advanced processing and decision process [28].
In the literature, several works have been published in the area of WSN clustering and routing. The authors in [7] presented a joint clustering and routing protocol that aims at reducing long-distance communication with the sink node and optimize the power consumption. The AoI is divided into different layers until reaching the sink node. Furthermore, each layer is divided into different identical clusters, whereas the number of clusters increases as becoming closer to the sink node. Two Cluster Heads (CH) are assigned for each cluster in order to balance the energy consumption within the cluster nodes. The rst CH is called the leader CH that is responsible for sending the collected data to the sink node, while the other CH is responsible for gathering data from the cluster nodes.
The authors in [6] modi ed the previous work by proposing to have a forwarding-head node on each layer as a replacement of the leader-head CH for each cluster. However, both proposals suffered load-balancing challenges which were resolved in [29]. In [8], the authors proposed different criteria for selecting the CH such as nodes' remaining energy, number of neighbour nodes and the distance between the nodes' and the sink node (ER). Another work that alternates between the nodes' selection for being a CH is proposed in [30]. The alternation depends on the system's overall nodes' energy such that the CH selection process is uniform among the nodes, thus prolonging the network lifetime.
Arumugam et al. [31] proposed an energy aware clustering algorithm and routing protocol that aims at optimizing network energy consumption. The CHs for each cluster are chosen for each cluster to minimize the energy consumption and optimize the sensor nodes' resources. The routing protocol chooses the node with the highest residual energy which helps to provide a high packet delivery ratio. Another distinctive clustering strategy was presented in [5], the authors utilized different parameters that are related to the nodes' remaining energy, degrees and centrality factors to select the CH node.
In dynamic clustering technique, the authors in [10] presented a self-organization clustering scheme to maintain relay networks of mobile data collectors. The proposed algorithm distributed the WSN nodes to different logical groups based on the convex hull algorithmic problem to reduce signalling overhead. Moreover, each local group has a locale mobility management. In [11], the authors used a genetic algorithm to choose the CH amongst WSN nodes based on different factors such as WSN node coverage range, the distance between WSN nodes, the ER, and the expected consumed power of WSN node to select CH nodes for both single-hop and multi-hop models. Furthermore, the authors in [12] used K-means to perform the dynamic CH selection. The ER used it to learn the preferences and priorities of a group of WSN nodes to attain optimal solutions over time using the information at the network level, whereas, partial learning enabled each individual WSN node to learn about their preferences and priorities to achieve optimal solutions over time using neighbourhood information. Another work which combines static and dynamic clustering is presented in [13]. The authors condidered a mixture of static and dynamic combinations along with layered-based multi-hop communication. Initially, the ER connects the WSN nodes to different layers based on their distance from the ER which is used for inter-cluster multihop connections. It also calculates the adjacent characteristics of the WSN nodes. The CHs are then shaped in a distributed manner, whereby the chance of WSN node becoming a CH depends on its energy level, harvested energy, neighbourhood attributes, and proximity to the ER.
Our literature review showed that the majority of the clustering algorithms deal with static nodes, whereas the clustering algorithms proposed for mobile wireless sensor network have high complexity, thus making them not suitable for the limited computational and energy sensor nodes, needless to mention the excessive delay for making the CH selection process, which is necessary to be minimized, especially in the mobile and dynamic environment of the mobile wireless sensor network.

Cryptographic Algorithms in WSNs
Providing an e cient cryptographic algorithm that suits the limited WSN node's constraints is challenging, especially that most of the existing and well know algorithms are not well designed and optimized for resource constraints devices. In the literature, several papers [23], [32]- [34] have evaluated the existing algorithms that are based on symmetric, or asymmetric algorithms that use private keys only or private and public keys jointly, and studied their effect on the nodes' energy consumption. In the literature, several research papers have been suggested to provide cryptographic security for small devices based on the symmetric, public key, or hybrid encryption scheme in restricted environments such as WSNs. Some works [23], [32]- [34] have concentrated on analyzing current encryption algorithms that have a direct effect on WSN nodes' energy and performance. Key management and authentications are other crucial processes that should be optimized to ensure energy e cient yet secure performance of the resources' limited sensor nodes. As such, several researchers proposed e cient key management and authentication mechanisms [35], [36].
Furthermore, several works have been designed with the objective of having lightweight block ciphers with simple and e cient architecture that suits low resources devices. For example,, PRESENT [37], TEA [38], SIT [39], Speck [40] and SIMON [40]. Speck is a lightweight block ciphers algorithm proposed by the National Security Agency (NSA) along with SIMON. Speck is designed to be implemented in software, while SIMON is better implemented via hardware. Speck is an Add-Rotate-XOR (ARX) cipher that supports different blocks and key sizes which both determine the number of rounds. TEA on the other hand uses XOR, ADD, and SHIFT operations to achieve nonlinearity. A block cipher version of TEA known as Block TEA was proposed by Wheeler et al. [41] that can deal with any number of plaintext blocks.
Despite the fact that these lightweight ciphers are specially designed to preserve energy without compromising security. However, some of the proposed lightweight ciphers have suffered attacks on roundest [2] and some of them have shown low performance compared with typical ciphers [32]. Theodore, their adoption in WSN is not without risk. Indeed, a systematic review and evaluation for lightweight ciphers for WSN taking into consideration the ciphers' security and energy e ciency are provided in [42]. The security-performance trade-off for several lightweight ciphers for WSN are studied in [14], [42], [43]. The authors researched the various aspects of lightweight cryptography for limited resource devices in [19] and de ned the block, key sizes, number of rounds and the con guration of the cipher as the main parameters affecting ad performance algorithm security. In addition, the authors in [43] assessed the in uence of these parameters on the lightweight cipher's architecture and performance. The authors have shown in [42] that a limited block size of 32-64 bits should be selected in a lightweight cipher over the conventional block size of 64-128 bits. In addition, the authors demonstrated that the security of the cipher is largely dependent on the key length. FlexenTech [22] is another lightweight symmetric block cipher considered for resource-limited IoT and WSN devices, where primitive computations are used to encrypt a plaintext.
Shannon [44] proved that the key size should be at least as big as the block size to reach an optimal con dentiality rate. In addition, the authors examined the most powerful key lengths and the lightweight block ciphers in [45]. The authors found that the time needed for the whole key space k to be exhausted is proportional to the time required for 2 k encryption operations to be carried out. Zhang et al. [15] analysed and compared the effects on WSN energy consumption of various cryptographic algorithms. They found that various variables such as packet size, cipher mode of operation, the initialization vector (IV) of the algorithm used, and the e ciency of the communication channel would in uence the energy.
Another line of research investigated the effect of various cryptography algorithms on the nodes' and devices' resources usage such as the memory used, processing time and power consumption [46]- [48]. In [46], the energy consumption of AES, DES (Data Encryption Standard), RSA (Rivest-Shamir-Adleman) and RC4 (Rivest Cipher 4) algorithms have been evaluated for WSN. Several key sizes and different cipher block modes and their effects on energy consumption have been investigated. In [47], the authors study the power consumption of deploying several cipher algorithms on Android smartphones where several les with different sizes have been examined. Particularly, the AES nalist algorithm RC6, Two sh, Serpent, and Mars algorithms. The paper results showed that the least power was consumed by the Two sh and RC6 algorithms followed by Mars and Serpent. Another work that compared several ciphering algorithms such as the RSA, RC6, 3-DES and AES was presented in [48]. The comparison metrics were: execution time, memory needed to store the code and data. Their work showed that the RC6 algorithms consumed the least resources. However, the Electronic Code Book (ECB) mode of operation was used which is considered insecure. Another widely used cipher algorithm that showed its effectiveness is the AES developed by the National Institute of Standards and Technology (NIST) [40].
However, none of these lightweight schemes have been analyzed for their effect on WSN network lifetime or WSN node power consumption. In this paper, we will analyze and compare the effect of using Speck128, TEA, AES and FlexenTech along with DynCH algorithm on WSN network lifetime, which both are essential and high relevance, according to our conducted literature review.

The Dynch Scheme For Mobile Wireless Sensor Networks
The DynCH framework consists of two key stages: a complex local cluster selection technique is used in the rst stage to constantly adjust the CH to match the energy consumption of the nodes and maximize its lifetime. A lightweight symmetric encryption algorithm is used to protect data transfer in the second stage. These symmetric block schemes will be evaluated on the basis of network lifetime complexity and power usage. Moreover, the key exchange process between the WSN nodes, CHs and the Edge router is implemented using a key management protocol [14]. Figure 2 depicts the WSN architecture deploying the DynCH algorithm won its different phases.
As depicted in Fig. 2, the WSN nodes (n 1 , n 2, …, n N ), where N is the number of WSN nodes within the ER).
The WSN nodes sense the environment and collect relevant data. Then the sensed data are sent to the Edge Router via the CH. The WSN nodes are assumed to be mobile and can move in different directions. In the beginning, the CH is chosen depend on several conditions such as the WSN node energy, the number of neighbouring nodes and the distance between them. Once the CH is chosen and the clusters are formed. The key management exchange process will be used and data are sent securely using one of the lightweight symmetric encryption algorithms.
The rst phase of the local cluster selection methodology is illustrated in Fig. 3. The parameters d, d T , E T , w, and ω represent the distance between WSN nodes, the distance threshold, the WSN node energy threshold, the average distance of candidate CHs to neighbouring nodes, and the weight of CHs selection, respectively by the owchart. However, all these parameters were explained in details in [14]. The algorithm starts by continuous veri cation (period of time) of the distance between each WSN node (n) and its CH.
Since the WSN nodes travel at a low speed of approximately 5 m/s, the distance is tested once per second [49]. If the distance starts reaching the value of the distance threshold (d T ), the WSN node starts searching for the nearest CH to pair with it. To target the CH, a WSN node sends a Join_Request message, then sends the Release_Request message to its CH source. This process uses WSN mobile nodes that lead to the reassembling and regrouping. If n travels away from its CH source and when the distance between them becomes less than d T , n continues to scan for the closest CH, if there is, n binds to the CH target and then releases it from the CH source. The target CH is the nearest CH found by n. The CH source is a CH node bound by n. n initiates the process of forming a new CH between the neighboring WSN nodes if there is no nearby CH, and then invites them to join. If the CH moves away from its related WSN nodes, on the other hand, it has a few of them (c), where c is the lower WSN nodes that are attached to each CH. CH will release the linked WSN nodes and become a periodic WSN node in order to join to the nearest CH. Moreover, the WSN nodes that have been free can also check for or build new cluster classes.
The Low Power and Lossy Networks (RPL) protocol is nevertheless designed to catch devices that merge mesh and tree topologies in 6LoWPAN networks. The root uses the broadcast message DIO in regular RPL to draw an instance of topology [26]. After updating, each WSN node that receives a DIO message will add the sender as its parent and then forward it to its neighboring WSN node [50]. Ultimately, the WSN node chooses the best route to act as the default gateway, depending on the parent list.
This work, however, relies primarily on ER to identify the CHs that are deployed in the ER area from the WSN nodes. This decision will be dispersed to the numerous zoning areas near the ER, after which a selforganized CH rotation will be carried out. Next, by sending the CH-Announce message, the CH starts transmitting its identi er within its own province. The distance to the neighboring CHs is measured by neighboring WSN nodes and the closest one is calculated. However, the upper limit of the WSN nodes attached to each CH is not provided by the DynCH scheme.
Since the data in WSN networks is transmitted in unsecured wireless networks, all connections in the networks must be secured. The key management system and encryption schemes are employed should keep the overhead costs of key generation, agreement, distribution and data ciphering in low-power networks. In the second phase of the DynCH scheme, we will use some of the symmetric block ciphers to evaluate their performance on WSN network lifetime along it. Moreover, these cipher schemes are usually containing simple key schedules running on primary processes such as AND or XOR. Even more, they are supporting various block sizes (such as 32, 48, 64, 96, and 128 bit). Therefore, these schemes are often aimed at hardware or software applications. Hardware-associated lightweight ciphers symmetric blocks algorithms include SIMON, LED, Piccolo, and PRESENT, while LEA, Speck, and Chaskey are among the software-associated lightweight ciphers [51]. Most of these encryption schemes are found secure enough to be used in real world applications. The Speck128, TEA, and FlexenTech are used to establish a secure and reliable connection between WSN nodes and their linked CHs, and ultimately reach the ER. Moreover, the key management algorithm proposed in [14] has been used in this analysis. These symmetric encryption algorithms are selected due [51]to their reliability and lightness compared to others [51]. 6LoWPAN uses AES for data encryption. In AES, the block size of this encryption is 128 bits, with key sizes of 128, 192, and 256 bits. The number of rounds depends on key size, 10 rounds for a 128-bit key, 12 rounds for a 192-bit key, and 14 rounds when using a 256-bit key. Moreover, it is based on a swaptransformation network structure, whose primary procedures being SubBytes, ShiftRows, MixColumns, and AddRoundKey. The best attack on full AES is the biclique, nonetheless it is slightly better than brute force [23]. In TEA, cryptography emphasis Feistel iteration using 64-bit block size and a 128-bit key size where individual rounds employ K[0; 1] and even rounds employ K [2; 3]. With using 64 rounds (32 cycles), the TEA authors say 16 cycles may be appropriate but they recommend 32 cycles using the constant" C" (to stop simple attacks depend on the symmetry of the rounds) [52]. Moreover, Speck has 10 alternative forms where the block size is 32, 48, 64, 96 or 128 bits, the key size is 64, 72, 96, 128, 144, 192 or 256 bits and the round numbers rely on the given parameters. A block always involves two words and the word size can be 16, 24, 32, 48 or 64 bits. Moreover, the corresponding key consists of 2, 3 or 4 words. Each round involves two cycles, adding the right word to the left word, XORing the key to the left word, then XORing the left word to the right word.
In terms of recon guration exibility, FlexenTech and Speck cryptographic schemes are designed with smoothed exibility allowing for some different block and key sizes in use. Speck gets its nonlinearity from the standard summation process; key lengths of less than 80 bits have been shown to not offer a high level of security. Hence, the Speck128 is selected for this study. In managing sophistication and level of security, FlexenTech provides stability. The algorithm only conducts simple calculations using a mixture of, substitutes, random permutations and bit-level rotations to encrypt single info [22]. Table 1 shows the key schedule and block-size variations between their techniques.

Simulation And Environments
We address the simulation setting in this section and the parameters are used rst. The DynCH model is evaluated in order to gure out the lifetime of the WSN network depending on the various simulation conditions. First, the DynCH is constructed, then symmetrical block structures are implemented. The performance of the DynCH scheme is compared to constant clustering techniques [5]- [9] called "SteadyCH".

Simulation and Performance Metrics
Cooja simulator [53] is used to simulate the wireless network architecture of 6LoWPAN. Cooja is working on a Contiki OS [54] which is an open-source operating system designed for Intent of Thing (IoT) technology to manage and control the device's hardware and software. The simulation runs on a 1.8 GHz Intel Core i7 processing computer with 8 MB of cache and 8 GB of RAM. In the simulation, the different WSN nodes' numbers are initially distributed with dimensions of 500 x 500 terrain connected to different CHs numbers at the initial values as shown in Fig. 4 (a & b). After that, the WSN nodes start moving in a random direction at a rate of 5 meters per second.
In the simulation, the numbers of the different WSN nodes are initially distributed with dimensions of 500 x 500 terrain linked to the initial values of different CHs numbers, as seen in Fig. 4 (a & b). After that, at a rate of 5 m/s, the WSN nodes start traveling in a random direction.

Experimental Results
For key e ciency metrics, the following are used to test the DynCH scheme based on WSN node communications clustering and security and expanding the lifetime of the WSN network.
Effecting of the number of CHs to network lifetime on both DynCH and SteadyCH mechanisms.
Evaluate the effecting of the WSN nodes numbers on network lifetime for both DynCH and SteadyCH mechanisms.
Effecting of different simulation time to network lifetime on both DynCH and SteadyCH mechanisms.
Execution speed of the lightweight of symmetric encryption schemes on different data sizes.
Effecting of the lightweight of symmetric encryption schemes to network lifetime on both DynCH and SteadyCH mechanisms.

The Results
The lifetime of the network is calculated when the power of some WSN nodes reaches 0. Therefore, in our experimental results, we rely on different parameter values to plot under different conditions. With regard to the evaluation of the "DynCH" proposal scheme, we set the initial energy of the WSN node to 1 joule and the simulation time set to 200 seconds to allow some of the WSN nodes' energy to reach 0. Moreover, the WSN nodes number set to 200 as well.
In the second circumstance, we assess the impact of lightweight security algorithms in relation to the use or without the use of cryptographic algorithms in static clustering "steadyCH" and dynamic clustering "DynCH" techniques. Hence, we set the simulation time to 400 seconds to allow an of WSN nodes energy that does not have an encryption mechanism to reach 0.

Complexity Analysis for the DynCH on the Network Lifetime
In m/s, the nodes start to move away from thier CH. The DynCH scheme advises each node to rejoin to another closest CH or selecting a new CH among adjacent WSN nodes, whereas in the steadyCHs mechanism, each node remains connected to its CH and the distance increases will increase the consuming WSN node energy. The result in Fig. 5 shows the DynCH improves network lifetime by 72%, 66%, 52%, 44%, and 12% compared to SteadyCHs in different numbers of CHs, respectively.
The analysis effect of WSN nodes number to the network lifetime is depicted in Fig. 6. The number of nodes is increased by 100 nodes each time and the number of CHs is set to 8 CHs for both schemes. As illustrated from Fig. 6, an increase in WSN nodes affects both techniques by reducing the network lifetime. The reason behind this decrement is the effect of over-connecting the ordinary WSN nodes to the CHs. However, the DynCH algorithm shows the improvement of network lifetime compared to the steadyCH technique because of the CH dynamic selection. Nevertheless, Fig. 6 shows a decrement in network lifetime as the WSN nodes increase from 200 to 300. The reasons behind this are the density of distributed WSN nodes and the maximum number of WSN nodes parameter is not taken into account for each CH node. Each CH can accept any WSN node whose d is less than the threshold value.
In order to allow the WSN nodes' energy to reach 0 when the simulation time is less than 200 seconds for network lifetime detection, we need to reduce the WSN nodes' initial energy. Therefore, in Fig. 7, the WSN nodes' initial energy is set at 0.5 J, the CHs number is set at 8, and the WSN nodes' number is set at 200 as well. In addition, the simulation program runs a different number of simulation times each time. The speci c simulation times are 50, 100, 150, and 200 seconds.
As illustrated from the same Fig. 7, the effect of simulation time on the network lifetime in both schemes is slightly different. The network lifetime in DynCH begins to increase when the simulation time increases while in SteadyCHs starts to decrease. This differentiation is due to the positive relationship between the DynCH function and the duration time. The duration time will increase the distance (d) between WSN nodes and their CHs, thus, the DynCH function will be launched and then start to create new CHs or join other CHs. This process will save power in both nodes (CH and ordinary), while in steadyCH the increment of duration time will increase the distance between WSN nodes and their CHs, thus, the consuming power will increase and the network lifetime will decrease.
The result shows the DynCH and steadyCH started slightly the same when the simulation time at the 50s.
The DynCH then starts to improve the network lifetime by 22%, 31%, and 53% compared to steadyCH in a different number of WSN nodes, respectively.

Complexity Analysis for the lightweight encryption algorithms on the DynCH scheme for extending lifetime
We test the performance e ciency of the Speck128, FlexenTech, TEA, and regular AES encryption algorithms relative to each other during the encryption and decryption time. With rounds ranging from 4 6 Conclusion And Future Work In this paper, we have analyzed the DynCH scheme in WSN networks' lifetime and saving energy in the 6LoWPAN protocol. The DynCH depends on distances and nodes' power to reformulate CH nodes between WSN nodes, which keep the WSN nodes' connection and their energy close to the best value.
Moreover, the analyses of lightweight symmetric block algorithms were done to evaluate their power consumption in wireless sensor networks. The lightweight key management was used to unify the authentication method in different symmetric block algorithms and reduce power consumption in establishing and managing cryptographic key between WSN nodes. The Cooja simulator have used to obtain the network lifetime and symmetric encryption time in different environments. The results of DynCH analyses show that improvement of network lifetime by 58% on an average of using different CHs, 47% on an average of using different numbers of WSN nodes, 36% on an average of using different simulation times, and 45% on an average of using different initial power compared to SeadyCH approach. Moreover, the Speck128 algorithm shows the best results in execution time and network lifetime prolonging. The Speck128 improved the average network lifetime by 34%, 52%, and 70% compared to FlexenTech, TEA, and AES, respectively. In addition, the Speck128 used 26%, FlexenTech used 52%, TEA used 65%, and AES used 78% of the WSN network's average lifetime compared to Nonencryption connection, respectively.
In future research, we intend to update DynCH to cover the minimum and the maximum number of WSN nodes in each CH. Moreover, we intend to update it to cover different ERs with alteration of handover process to be suitable with wireless senor network. The Software De ned Network (SDN) controller could be used for this proposal.