1
BBUCAF: A Biometric-Based User Clustering Authentication Framework in Wireless Sensor Network

Rinesh, S.1*, Thamaraiselvi, K.2, Mahdi Ismael Omar1 and Abdulfetah Abdulahi Ahmed1

1 Department of Computer Science, Jigjiga University, Jijiga, Ethiopia

2 Department of Computer Science, Malla Reddy College of Engineering, Hyderabad, Telangana, India

Abstract

Wireless Sensor Networks (WSN) have made much progress in the last few years, so data transmission must be more secure. Cryptographic keys keep information private, authenticate people, and keep data safe. Several research projects were done to interact with important management issues in WSNs. Prime statistics are used to make collective keys. It would then be able to accurately check the security of nodes. A new network way is modeled for sending data between nodes without restriction. A strong authentication system is needed to maintain network safety and allow people to use a network service freely. But the limited supplies of sensor nodes make it tough to authenticate people. To overcome the security-based issues, a biometric-based user clustering authentication framework (BBUCAF) has been introduced to increase the level of security and the network's speed among the nodes. A biometric-based model is created by taking features from the fingerprint. Securely, feature vectors create a private key for the user. Such a key is sent to every sensor node. Then, private keys between sensor nodes are made by combining a randomly generated count and the user's key, which is sent to each sensor node. C- means Clustering is used to group nodes based on their range and unique identification. A collective key is made here using a fuzzy registration component that considers prime numbers. Fuzzy membership and biometric-based secret keys send data between groups and sensor nodes. Each cluster has group keys that differ from one cluster to the next. The network's speed improves the network's effectiveness by cutting down on network traffic, protecting against DoS attacks, and extending the battery capacity of a node's battery with less energy usage.

Keywords: Wireless sensor network, nodes, clustering, network traffic, authentication

1.1 Introduction to Wireless Sensor Network

Several sensors can be used together in a single WSN. Nodes in the network that sense their surroundings are known as sensor nodes [1]. A wide range of applications, such as structural health monitoring, environmental control, and combat observation, can benefit from such connections [2]. A node can perform computing, identify itself, and communicate with other devices [3]. Those nodes can be dispersed in a situation where they can identify each other and work together to accomplish the task in a large region [4, 5]. Sensor nodes in WSNs are used for specific tasks [6]. Small sensor nodes in the network model their surroundings' information after spotting it [7]. Due to their wide range of applications, WSNs are becoming increasingly popular in education and the market [8]. WSNs are primarily designed to gather and send environmental information to a home or remote location via a network of sensing devices located in an isolated community [9]. The original data are then processed online or offline as per application standards for a full evaluation in a remote location [10]. If a patient is not in the hospital, for example, remote patient tracking is important for doctors.

These systems can benefit from numerous applications, including structural health monitoring, environmental control, and combat monitoring [11]. Most apps allow users to obtain data immediately from a gateway node because queries are handled on this node in most cases [12]. The information from a gateway node is very hard to receive on rare occasions. Therefore, sensor nodes collect information directly [13]. By sending the request to a sensor node, unauthorized users can quickly obtain sensitive information [14]. As a result of sensor nodes' inability to verify query messages may leak sensitive data, and network resources, such as node power and bandwidth, could be wastefully depleted [15]. Any or all of the associated issues could impact the network's lifespan and effectiveness, making the system inaccessible to genuine people [16]. Since network data and resources can be illegally accessed, authentication is necessary [17]. To achieve this, sensor nodes must validate users' identities [18]. All of the following issues can be solved with user authentication, which enables authorized users to join a system [19]. As a result of the resource limits of WSNs' small sensor devices, namely their power and storage, along with their processing and transmission capabilities, providing authentication in these networks is a very difficult issue [20]. Even though several standards have been presented, the authentication procedure is still vulnerable. In the end, a more robust and intelligent process is needed to assure the security of a WSN [21]. Maintaining a safe network requires a robust authentication system that allows users to access network services without restriction. Authentication is difficult due to the restricted supply of sensor nodes [22, 23]. To overcome all the above-mentioned security-based issues, BBUCAF has been developed. The main contribution of BBUCAF is

• To build a biometric model, enhance the network's security and performance using fingerprints' unique characteristics.
• The user's private key is generated securely using feature vectors. Every sensor node receives a key. Then, each sensor node receives a random count, and the user's private key is combined with each sensor node.
• Numerous benefits of a faster network include reducing network traffic, preventing denial-of-service (DDoS) assaults, and increasing node battery life.

1.2 Background Study

Many researchers have carried out research works. Tsu-Yang Wu et al. [24] developed Three-Factor Authentication Protocol (TAP), in which the logical study and informal analysis confirm safety, Burross-Abadii-Needham (BAN) logic, and ProVerif tools. The evaluation of security and performance reveals that the method offers stronger security and reduced computational burden.

P.P. Devi et al. [25] proposed SDN-Enabled Hybrid Clone Node Detection Mechanisms (SDN-HCN). An SDN-based methodology performs a network path evaluation and time-based research methodologies to identify and reduce duplicate nodes produced by cloning attacks. To identify clone nodes in a wireless network, one must use the HCN technique. The simulation results reveal that several metrics are analyzed in the experiment.

M. Rakesh Kumar et al. [26] introduced a Secure Fuzzy Extractor-based Biometric Key Authentication (SFE-BKA) Scheme. The hash function is critical to the system's security. In SFE-BKA, the hash parameter value is irrespective of hash functions in an attempt to improve information security. The proposed method is not affected by this variance in hashing in terms of latency or delay. The outcome of SFE-BKA yielded 40% less data loss, 20% less energy usage, and less latency than earlier encoding systems.

S. Ashraf et al. [27] developed a Depuration-based Efficient Coverage Mechanism (DECM). Two rounds of deployment are required to complete the process. When a node is to be moved to new locations, the Dissimilitude Enhancement Scheme (DES) is used to find it. The Depuration mechanism in the second cycle reduces the separation between prior and new places by controlling the needless migration of the sensor nodes. By analyzing the simulation findings and computing in 0.016 seconds, the DECM has attained more than 98% protection.

Fan Wu et al. [28] described the Authentication Protocol for Wireless Sensor Networks (AP-WSN). Proverif 's formal verification shows that the new system retains its security features. AP-WSN is feasible and meets general demands in a way that counters various threats and meets security properties. The proposed approach outperforms previous schemes in terms of security and is suitable for use. The simulation findings indicate that the plan may be successfully implemented in an IoT system and have a practical use.

Diksha Rangwani et al. [29] discussed improved privacy-preserving remote user authentication (PP-RUA). The suggested system is formally analyzed using the probabilistic Random-Oracle-Model to show the resilience of the scheme. Further, the system is simulated using a well-accepted AVISPA tool to show its security strength. The performance assessment of the system demonstrates that along with its consistency in aspects of privacy, the suggested scheme is more effective in computing and networking overheads than other current schemes.

SungJin Yu et al. [30] discussed Secure and Lightweight Three-Factor-Based User Authentication (SLUA). Secure, untraceable, and mutually authenticated communications are possible with the SLUA. Informal and formal methods are used to assess the safety of SLUA, along with the logic of Burrows-Abadi-Needham (BAN), the Real-or-Random (ROR) model, and the AVISPA simulation. SLUA-performance of WSNs is compared to other existing systems. Security and efficiency are more protected and more efficient in the proposed SLUA than in the prior...