INTERNATIONAL JOURNAL OF COMPUTERS COMMUNICATIONS & CONTROL Online ISSN 1841-9844, ISSN-L 1841-9836, Volume: 18, Issue: 2, Month: August, Year: 2023 Article Number: 5144, https://doi.org/10.15837/ijccc.2023.2.5144 CCC Publications Blockchain-assisted Secure Routing Protocol for Cluster-based Mobile-ad Hoc Networks N. Ilakkiya, A. Rajaram N. Ilakkiya Department of Masters in Computer Application E.G.S Pillay Engineering College Nagapattinam, Tamilnadu, 611002, India *Corresponding Author: Email: illalliyaphd@gmail.com A. Rajaram Department of Electronics and Communication Engineering E.G.S Pillay Engineering College Nagapattinam, Tamilnadu, 611002, India drrajaram@egspec.org Abstract MANETs aredecentralized network that involves mobile nodes. As the overall network is mobile and has no centralization, network management, routing, and security become very challenging. Though many works have been presented, still there is a lack in organizing the network due to unauthorized access, centralized security schemes, and the dynamic nature of the nodes. This paper proposed a novel Blockchain-assisted Secure Routing (Block-Sec) protocol for MANETs. All mobile nodes are authenticated by Distributed One-Time Passcode (DOT) based authorization scheme. All authorized nodes are segregated into multiple clusters based on Weight based Dynamic Clustering (WDC) algorithm in which multiple metrics are considered in clustering and re-clustering processes. After cluster formation, each cluster is elected with optimal Cluster Head (CH) by Strawberry Optimization (SBO) algorithm with a new objective function. After cluster formation, the optimal route is selected by Fast Neural Net-assisted Fuzzy (FNNF) algorithm by combining multiple variables. Data transmission is secured by Efficient Elliptic Curve (E2C2) algorithm. With the combined algorithms, the proposed approach obtainedimproved efficiency in packet delivery ratio (PDR), throughput, time analysis, and security level. Keywords: Blockchain, Distributed Security, Mobile Nodes, Dynamic Clustering, MANETs. 1 Introduction With the advancements in wireless standards such as WiFi, Radio networks,etc, there are many applications have been developed. However, the overall network fully depends upon centralized base stations (BSs) or routers [1]. MANETs resolve the problem of centralization by distributing the data transmission throughout the network [2]-[4]. MANETs ensure ease of network deployment and data transmission. MANET’s general architecture is given in figure.1. As shown in the figure, MANET has no central BS or routers. It is defined as the collection of wireless movable nodes that need no infrastructure for communication. The nodes presented in the network are generally mobile. https://doi.org/10.15837/ijccc.2023.2.5144 2 Figure 1: MANET architecture The mobile nodes are allowed to move around the network limit [5], [6]. Each mobile node is capable to act as a transmitter, receiver as well as router. Due to its wide usage, MANET is emerging in many growing Internet of Things (IoT) applications such as environmental monitoring, pollution monitoring, etc [7]. Due to the lack of infrastructure, routing becomes challenging in MANET [8]. Mainly, routing depends upon on-demand routing and distance vector-based protocols. DSR and AODV routing protocols are the two major on-demand protocols [9]. Both protocols initiate route discovery only if needed for data transmission. However, the existing MANET protocols fail to achieve the following efficacy, • Unable to choose the best route • No evaluation is considered in route selection • High possibility for link breakages • Unable to ensure prompt data delivery • Increases the number of retransmission • Route breakages due to mobility To resolve the above issues, routing is performed by considering network dynamics [10]. That is many algorithms use mobility as the major routing metric. However, it is proved that using combined metrics for route selection improves the efficiency of data transmission [11]. Generally, routing and network organization are interrelated. Routing will become simple and effective when the network is organized properly. For proper network manage- ment, MANET is clustered into multiple groups [12]. Cluster formation splits the network into multiple groups which makes data transmission easy. For cluster formation, k-nearest neighbor (KNN), and k-means algorithms have been utilized [13]. However, these conventional algorithms only consider the distance metric which is not effective. The combined cluster-based routing approaches assist in optimal data transmission [14]. On the other hand, security is the major concern in MANETs [15]. Because the network is probably deployed in the open and remote areas thus it increases the chances of tampering, data modification, and theft [16]. Network Layer Attacks/Issues Physical •Jamming •Tampering •Denial of Service (DoS) Link • Eavesdropping, Overhearing, • Man- In-The-Middle Network • Routing Attacks such as Blackhole, • sinkhole, • wormhole Transport • Unauthorized access, • IP spoofing Application • Malicious codes, • viruses, • application thefts In table.1, the major security issues existing in MANET are summarizedconcerning the layers [17]. As the network has no infrastructure, there is no control over the network nodes. Many research works have proposed security approaches such as authentication [18], encryption [19], and so on. However, security provisioning is https://doi.org/10.15837/ijccc.2023.2.5144 3 still challenging due to, • Network mobility • Mobile attackers • Strength of attackers • Central server-based authentication On the whole, data transmission and security provisioning are two major aspects of MANET networks. 1.1 Blockchain Network Blockchain is a technology that allows for the transparent, secure, and decentralized storage and transmission of information [20], [21]. It serves as a sizable archive that preserves a record of all user interactions ever since the blockchain was founded. The distributed design of blockchain, which is not hosted by a single server but rather by a small group of users, is a fantastic feature. Components of the blockchain contain security measures to safeguard the system and do not need an intermediary to verify the chain’s authenticity or the accuracy of the data. Blockchains are shared, decentralized, and fault-tolerant databases that are available to everyone on the network and are not under the jurisdiction of any one entity. The technology is made to function against enemies in high-stakes circumstances. Blockchains are open, distributed, and fault-tolerant databases that anybody on the network may use; they are not under the jurisdiction of any one entity. The technology is built to function in dangerous scenarios against foes. In figure.2, integration of Blockchain and MANET is depicted [22], [23]. Figure 2. Blockchain and MANET integration Each member of the network stores an identical copy of the blockchain and is in an essentially equal position. Blockchain is employed in various application scenarios and is regarded as one of the crucial strategies to speed up global growth due to its high level of security and dependability. Huge research works had been carried out for multimedia wireless networks in terms of routing, congestion avoidance, security features, packet loss and reducing delay during transmission, etc. Some of them are discussed here to relate the proficiency of our proposed work. 1.2 Research Challenges Research Challenges existing in MANET are, Dynamic Network Topology Because of the dynamic network structure, nodes may migrate at any moment and in any direction. This leads to link breakages during routing Energy Consumption MANET involves mainly resource-constrained devices. Frequent route discovery and selection lead to high en- ergy consumption. Security Breaches The involvement of the central security server increases the single node problem and is unable to provide a security level. https://doi.org/10.15837/ijccc.2023.2.5144 4 1.3 Research Contributions To resolve the research problems, this paper summarizes the following research contributions, • A novel Blockchain-assisted Secure Routing (Block-Sec) protocol is presented for MANETs • All mobile nodes are authenticated by Distributed One-Time Passcode (DOT) based authentication scheme which uses blockchain for node validation • The network is formatted with Weight based Dynamic Clustering (WDC) algorithm with Strawberry Optimization (SBO) algorithm-based CH selection process • For optimal route selection, Fast Neural Net-assisted Fuzzy (FNNF) algorithm is introduced with multiple metrics • Data transmission security is ensured by Efficient Elliptic Curve (E2C2) algorithm. 1.4 Paper Organization The remaining paper are arranged as follows: section II provides a literature survey on existing works held on MANET secure routing. In section III, the major problem is stated which is resolved by the proposed approach. In section IV, the proposed approach is discussed. Section V evaluates the overall performance through extensive simulation observations. In section, VI the contributions are concluded with future research directions. 2 Related works This research uses the well-known PSO (i.e., particle swarm optimization) approach to solve the problem of node mobility, and an adaptive k-nearest neighbor cluster algorithm is also developed [24]. To help with cluster formation, a multi-objective fitness function of PSO is considered. General testing in a simulated networked environment shows that the recommended method works. The complexity of MANET, where optimum energy is one of the key elements, was increased by mobile nodes migrating at random area of interest. It is far more difficult to maintain long battery life while simultaneously allowing for frequent changes in the architecture of mobile nodes. An effective parallel computing method that manages topological structures well is the honeycomb-based paradigm [25]. Additionally, the most important objective of MANET is to choose the best routes while taking energy efficiency into account. To do this, this study introduces the IEEHR paradigm for MANET. The model reduces the broadcasting range while performing pathfinding by combining Honeycomb-based area coverage with LAR. Since mobile nodes have limited energy, efficient energy use is also necessary for MANET in addition to good routing. A network’s performance and durability will both improve by how efficiently energy is spent in the network. Here, more energy is saved when the mobile nodes are asleep, which is further used up during the efficient routing process. The trust of nodesin direct and indirect pathways are principally evaluated in this research [26], and the secure multipath is chosen while also identifying and isolating the vulnerable nodes. To protect the data packets from assaults of data transmission, the DPs are then encrypted using the SH2E method. In identifying the finest way out of the multipath chosen, the LF-SSO algorithm is then used. By determining a path trust-based path, residual energy of the node, and the path’s distance, this approach increases the network lifespan. Then, using the found optimum path, the encrypted DPs are sent from the sender to taget, and lastly, relayed to the base station.[27-31] Because of the constant node migration and the limited resources available, security management is a severe challenge [32-37]. Rekeying is only done for the so-called clusters of subnetworks to avoid having to repeatedly renew the group key for the entire wide network. This research proposes an integrated strategy of the HDGK management and FTBC to deal with this problem. The FTBC uses fuzzy logic principles to separate misbe- having nodes from genuine data transfer and classify trusted and untrusted nodes. No one methodis used for all application types, of simple clustering and improved weighted distributed clusteringare proposed to meet various demands. Although multi-hop routing is essential in MANET, it might present difficult problems during communication including a lack of data privacy. ECC is combined with the Bee clustering strategy to offer a secure and resource-efficient data transport system [38]. Data dependability is still questionable because of attacks like data dumping attacks and black hole attacks even if it guarantees data. The neighbour routers in these situ- ations employ the overhearing approach, and the packet forwarding statistics are calculated using the ratio of received to forwarded packets. If the network’s packet forwarding ratio is low, an attack can be identified and a trustworthy another channel can be found for information transfer. The suggested work involves the SC-AODV integration, ECC, and a further overhearing mechanism into the Bee clustering strategy, which overall assures data secrecy, data dependability, and energy efficiency. https://doi.org/10.15837/ijccc.2023.2.5144 5 To improve PDR and network lifetime, a new routing protocol for MANETs was proposed in Hybrid Optimization- Based MPR-DC-Based MANET. This protocol uses a hybrid optimization technique that combines ACO and PSO [39]. A multipath-based routing algorithm for air pollution monitoring in MANETs was presented in [40]. This algorithm considers the nodes’ available energy and the links’ quality to choose the best routes and enhance network performance. [41] to control the broadcasting of many packets and guarantee the stability of all mobile nodes in a MANET, the proposed TCSSR algorithm makes use of timer count scheduling and spectator routing. To decrease packet loss and lengthen network lifespan, clustering, and stifle limitation methods are also utilized. [42] to extend the lifespan of inexpensive and compact sensor nodes in WSNs, the SICRA is presented. To save energy and balance the network load, SICRA calculates the ideal routing route by examining the transmission coverage, connection link, and other criteria.NMRouting is a neural controller based on perceptron and an MCDM controller that determines the optimum pathways for sending data packets in a neural-MCDM-based routing protocol that has been suggested for MANETs. To increase network dependability, three different queue types are also used [43]. With the use of fuzzy theory, MCDM, and an RGB model, the FSB-System is a de- tection system that is presented for estimating the likelihood of fire, asphyxia, and burns. Fuzzy controllers are utilized to calculate probability while temperature, smoke, and light sensors are employed to make judgments under various scenarios. Clusters of sensor nodes are set up, and data is sent from non-cluster heads to cluster heads [44]. The drawbacks of existing techniques including the complexity of MANET, where optimum energy is one of the key elements, are increased mobile node’s random movement within a region of interest. It is more diffi- cult to maintain long battery life while allowing for frequent changes in the architecture of mobile nodes. The IEEHR paradigm reduces the broadcasting range while performing pathfinding by combining Honeycomb-based area coverage with LAR, which can limit the scope of the network. While the proposed DXOR-RC6 with FE technique ensures data security and integrity in the E2-SR system, the security level of the other approaches is not explicitly mentioned. The use of multiple algorithms and techniques, such as the LF-SSO algorithm, may increase the complexity of the network and require more resources for implementation and maintenance. To overcome these drawbacks, the proposed protocol uses a DOT based authorization scheme to authenticate all mobile nodes in the network, providing a secure and reliable way of identifying legitimate nodes, The WDC algorithm is used to segregate authorized nodes into multiple clusters based on multiple metrics, which allows for better organization and management of the network. SBO algorithm is used to elect the optimal CH with a new objective function, which helps in the efficient management of the cluster and network. FNNF algo- rithm is used to select optimal routes by combining multiple variables, which helps in efficient routing and better performance. The proposed protocol uses the E2C2 algorithm to secure data transmission, ensuring the confidentiality and integrity of data. The combined algorithms used in the proposed protocol achieve better efficiency in PDR, throughput, time analysis, and security level, providing a more robust and reliable network 3 Problem statement The issues related in secure routing for mobile networks is a critical issue because of the dynamic and decentralized network. One of the main challenges is to ensure that data is transmitted securely between nodes while minimizing energy consumption and maximizing the delivery rate. However, attackers can exploit vul- nerabilities in the network to intercept or manipulate data, compromising the security and reliability of the network. To address this problem, a Min-Max function is formulatedfor balancingbetween energy consumption, delivery rate, and security level trade-off. The objective is to minimize energy consumption and the number of retransmissions while maximizing the delivery rate and security level. This objective is achieved by formulating the problem as a Min-Max function, where the energy consumption is minimized and the no. of retransmissions is to be optimized while the maximum delivery rate and the security level are to be maintained. The presence of attackers further complicates the problem, as they can exploit vulnerabilities in the network to compromise the security and reliability of the network. Therefore, the problem is formulated to minimize and maximize the consecutive functions in the presence of attackers. This formulation ensures that the network is secure, reliable, and efficient, even in the presence of attackers. The problem of secure routing in the mobile network is formulated as the problem of the Min-Max function as follows, Where EC , RT represent energy consumption and the number of retransmissions and DR,SL represent delivery rate and security level respectively. The problem is formulated as minimizing and maximizing the consecutive functions in the presence of attackers. 4 Proposed Work The proposed Block-Sec work aims at achieving high-level security with a reasonable data transmission rate. The overall work is explained in detail. https://doi.org/10.15837/ijccc.2023.2.5144 6 4.1 Network Model The proposed mobile network has n number of mobile nodes M1,M2,..Mn with the mobility range of denote the lower and upper bound of the mobility range respectively. All mobile nodes are allowed to move around the network with varying mobility speedswithin the range. The overall network is divided into multiple clusters for ease of management. All mobile nodes within the network are authorized before cluster formation. The proposed Block-Sec architecture is illustrated in figure.3. With the help of the Blockchain network, the nodes are validated in a distributed manner. For data, transmission is carried through the optimal route selected in the network. Overall data transmission is secured by using lightweight cryptography techniques. 4.2 Mobile Node Authentication Authentication is the process of validating the authorization of the nodes presented in the network. For authentication, the DOT protocol is proposed which uses the distributed Blockchain network. In general, au- thentication credentials are stored in a single server which lacks with single node failure problem. To avoid this major issue, we presented a Blockchain network for authentication. F ¯ igure.3 Proposed Block-Sec Architecture WDC is an algorithm used for the formation of clusters in MANETs. The algorithm considers multiple metrics such as node mobility, distance, and energy levels to create and update clusters dynamically. WDC-based cluster formation helps to improve performance of the overall network by reducing the overhead and improving the routing efficiency. A blockchain is a decentralized, distributed ledger that is used to record transactions securely and transparently. In the context of mobile networks, a blockchain network can be used to secure data transactions and ensure data integrity. The use of blockchain in mobile networks helps to reduce the risks of unauthorized access and enhances the security of the overall network. The Cluster Head (CH) is an important node in the cluster-based routing protocol for MANETs. The optimal selection of CH is critical for the performance of the network. Strawberry Optimization (SBO) is an optimiza- tion algorithm that can be used for selecting the optimal CH in MANETs. SBO-based CH selection helps to improve the network performance by reducing energy consumption and increasing the reliability of the network. DOT-based Authentication: Distributed One-Time Passcode (DOT) is a scheme used for node authentication in MANETs. In this scheme, each node is assigned a unique passcode, which is used for authentication. DOT-based authentication helps to improve the security of the network by preventing unauthorized access and ensuring that only authorized nodes can participate in the network. FNNF-based Optimal Routing: Fast Neural Net-assisted Fuzzy (FNNF) is a routing algorithm used in MANETs. The algorithm uses multiple variables such as distance, mobility, energy levels, and security levels to select the optimal route for data transmission. FNNF-based optimal routing helps to improve network performance by reducing the number of retransmissions, improving the delivery rate, and increasing the security of the network. The proposed DOT involves two main phases such as, 1. Registration in DOT – This process is initiated when a node is joining into the network. The proposed DOT protocol receives three main authentication credentials such as mobile node ID (M_ID), password (M_PW) and certificate (MC ). The certificate is generated by the main server in the network. The server is connected with the Blockchain network (i.e.) single node failure can be avoided. If i^ththe mobile node needs to register, https://doi.org/10.15837/ijccc.2023.2.5144 7 then it initiates the RegIni message with MI D(i),MP W (i) and MC (i) as follows, Mi →ReglniMI D (i)×MP W (i) .MC (i) After successful registration, the server replies RegSuc messages to Mi with required private and public key pairs (Kpu (i),Kpr (i)) with time stamp (TS ) as follows, Block→RegSucKpu (i),Kpr (i),TS Also, the server generates unique passcode (UI D (i)) by using the XOR function as follows, UI D (i)=H(MI D)⊕H(MP W )|Kpu H()represent the corresponding hash value. With the unique ID creation, the registration phase is completed. All these credentials are stored in the Blockchain network in a distributed manner. Authentication Phase – Once the node is registered, then it will have three major credentials with public and private key pairs. Whenever the node needs to participate in the network the node needs to be authenticated with the Blockchain. For authentication, the node needs to initiate AuthIni message to the Blockchain network. The AuthIni message is composed of, AuthIni→SignMI D (i),MP W (i)+UI D (i) The credentials need to be signed digitally by the mobile nodes. On receiving AuthIni message, the server validates the credentials with the Blockchain. If the credentials are correct, then it returns one-time passcode with the time stamp as follows, OT Pt→R⊕MI D (i) _t The OT Pt is a session passcode that is valid only for the period t. That is, the node Mi can submit the passcode to the CH to join the network. It can be seen that, after the session is completed the OTP will become invalid which ensures no adversary can reuse the code after sometime. In this manner, the nodes are authenticated in the network. The above procedure explains the steps involved in the proposed DOT protocol. The proposed authentication protocol allows cluster formation with only valid nodes which improves the security level. https://doi.org/10.15837/ijccc.2023.2.5144 8 4.3 Secure Cluster Formation Once the nodes are authenticated, the next step is to form various clusters to make the data transmission simple and effective. We proposed a WDC method that uses weight value similarity for cluster formation. The major issue in cluster-based MANET is the stability of the clusters is always questionable since the nodes move away adequately which leads to collapse in the network. Thus, we presented the WDC method to improve cluster stability. First, each node Mi computes weight value based on the following three metrics, Mobility – It is represented asµ in terms of the moving node velocities in the network. Connectivity – It is represented as Ci which denotes the current node connectivity(i.e.) the no. of nodes connected with the particular node. Centroid Distance – It is represented as CDi. It is computed as the distance between the network’s central point and the current position of the node as Euclidean distance. Altogether, the weight value for the node Mi (Wi) is computed as follows, Wi=Σ µ,Ci,CDi Each node computes the Wi, then initiates the cluster formation process. The WDC-based cluster formation involves the following steps, Handshake Session – First, each node transmits the CluFor message as follows, CluFor→MI D (i),Wi,OT Pt Decision-Making Session – On receiving CluFor message, each node makesa decisionbased on two conditions as follows, Condition 1:Validate OT Pt Condition 2:Compute weight difference WD W D(i, j)=Difference[Wi,Wj ] The decision is True only if both conditions are satisfied. Cluster formation – If Mi receives CluFor from node Mj , it first checks condition 1. If the condition is satisfied (i.e.) Mj is valid, then it moves to the next condition 2. If the W D(i, j) is low, then the node accepts the formation message. Similarly, node Mj validates the Mi. After decision-making, each node transmits CluCon message to the optimum nodes to form clusters. This mutual session authentication improves the security level in the network. Also, consideration of W D(i, j) ensures that all nodes in the cluster have the same level of mobility and distance. So that, the clusters formed by the WDC method will be stable. Optimal CH selection At the end of the WDC method, the overall network is segregated into v number of clusters denoted as G1,G2,..,Gv . The next step is to select the optimal CH in each cluster. In each cluster, CH is responsible to manage the nodes within the group. Thus, the CH must be optimal and capable of handling the cluster management process. For optimal CH selection, we proposed an SBO algorithm with multiple objectives. The SBO algorithm is inspiredby the growth procedure of strawberry plants. The strawberry plant’s growth depends upon runners and roots. Here, the runner represents the global search while a local search is represented by roots. The overall algorithm deals with the following factors, The problem domain is spread out with the runners The mother strawberry plant is developed and improved through the roots and hair. This growth is random in the problem domain. The daughter plants which are developed from the mother plants improve the growth through roots and runners. The daughter plants exist only if the growth is healthy otherwise it dies The above facts are considered in the algorithm. Initialization: Initialize the number of solutions as the strawberry plants. Here, the solutions are the nodes presented in each group Gj . Problem Formulation: The objective function f(x) is the combination of functions that needs to be considered in the CH selection process. In the proposed SBO, the objective is to select optimal CH and the problem ofoptimization is formulated as follows, Min f(S)→ K O F if Sl