Since the data transmitted between the sensors in a sensor network are exposed to the external attacks especially that controlling these sensors are difficult and might be impossible. Therefore this proposed approach aims to help the sensor devices to transmit the data between them securely and guarantee the integrity and the privacy of these data. The proposed approach should also grant these sensors the ability to establish a secure communication with the neighboring nodes.
1.1 Research Motivation
The sensor network consists of a set of sensors which are distributed randomly to a specific area to collect readings from the surroundings and transfer the collected data between them until it reaches the base station. However, certain data has to be secret such as the readings from military deployments. Since the nature of the communication between sensor nodes in WSN is broadcast transmissions, the same key is used by a group of sensors within the same range. Therefore, these communications are vulnerable to external or internal attacks and should be kept protected especially if the attacker manages to compromise few sensor nodes or manage to intercept the communication channel.
The goal of proposed approach is to increase the resilience of the network by increasing the security of the communication with a polynomial key pre-distribution scheme to provide node–to– node mutual authentication.
1.2 Research Questions
The main condition and challenge that is faced by the proposed approach is the security characteristic, such that the proposed approach must give the desired level of security for the sensor network to protect its sensors from the Sybil nodes.
The security level of the proposed approach will be expressed by answering the following questions:
1) Is the LU-Block decomposition approach secure?
2) How many Sybil nodes the adversary need in order to break the entire network?
3) How many messages do the adversary need to intercept in order to break the used key?
1.3 Research Significance
Most of the known key pre-distribution approaches are based on pre-distributing the keying information to the sensors before they are deployed into the field. Thus, the base station does not need to transfer this information to the network. In this regards, the main characteristics of the proposed approach that will distinguish it from other existing approaches are:
1) The proposed approach should employ the polynomial-based key pre-distribution to ensure a good resilience to the network due to the security property of the -degree polynomials. The network is resistant against node captured if the number of compromised nodes is less than node, where users communicate can compute a common key but users cannot compute anything about a key of other nodes (Dai and Xu, 2010; Banaie et al., 2015).
2) The proposed approach is based on the Block LU-matrix based method to ensure that any pair of sensors can find a common key. Any matrix is a product of two matrices; a lower triangular matrix (L) and an upper triangular matrix (U), each node receives one row from the (L) matrix and one column from the (U) matrix. Any pair of nodes want to communicate they can exchange their columns, then each node multiplies its row with the other node’s column to obtain a shared key. Since the matrix that is built will be symmetric, then the multiplication of this matrix will ensure that the shared parties will get a shared key (Dai and Xu, 2010).
3) The proposed approach will break the public matrix into blocks to make it harder for the adversary to guess the right block that is used for establishing the key pair between the sensor nodes. The proposed approach depends on using one of the matrix blocks, and therefore there will be more options for the adversary. And this is the main advantage of the proposed approach.