RaMaN: A dynamic clustering protocol resilient against colluded role-manipulability and node replication attacks

Clustering in WSNs has been a topic of intense research as clustering increases scalability and lifetime of the sensor network where sensor nodes are limited in computation and battery power. Clustering process in WSNs is prone to attacks from inside and outside adversaries. Most of the secure clustering schemes provide solutions against outside adversaries but fail in presence of inside malicious nodes. Resilient cluster leader election [1] and SecDeach [2] are among recent schemes which provide resilience to manipulability of a cluster head role in presence of compromised nodes. But, Resilient cluster leader election scheme [1] does not support dynamic clustering and it is also prone to node replication attacks. Though, SecDeach provides secure dynamic clustering but it suffers from the attack in which compromised nodes collude to manipulate their roles as cluster heads after each round. SecDeach is also vulnerable to node replication attacks. In this paper, we present a secure clustering solution which supports dynamic clustering and eliminates the possibility of colluded role manipulability attack by malicious nodes which is possible in SecDeach. The proposed solution can also handle node replication attacks which exist in Resilient cluster leader election scheme and SecDeach.