A Linear Wireless Sensor Network (LWSN) is a grouping of sensors arranged in a linear topology with or without junction zones. LWSNs are applied in several areas involving, among others, gas pipelines, road or rail infrastructure, border surveillance infrastructure, waterways, oil, etc. The (LWSN) are subjected to the same constraints as those subjected by the traditional networks (non-linear), energy saving, overload, delays, latency etc. To optimize the energy consumption of the network, among other things, one technique consists in organizing the network into several groups of nodes called clusters. Several algorithms for the construction of clusters have been proposed in the research. The majority of these propositions concern classical networks (non-linear). Recent proposals for clustering algorithms relating exclusively to linear networks have been made in the current research. These last proposals concerned k-redundant networks (number of nodes within transmission range) with a k previously assumed as an assumption (k = 2). On the other hand, these proposals do not take into account the addition, loss and removal of nodes in the sensor network. Therefore, faced with these situations, the clustering algorithms already proposed may be unsuitable. In this article we propose a generalization (extension) of a clustering mechanism for linear wireless sensor networks. The algorithms that we propose in this work will allow the linear sensor network to be able to react to the addition of new nodes as well as to the loss and deletion that the network nodes may experience. The solution that we present also works for all k-redundant networks regardless of the value of k, it is not fixed on a constant value of k as was the case for the solutions previously proposed in current research. The solution was evaluated with the Castalia / Omnet ++ simulator, on a topology with several junction zones. The results obtained showed a complete reconstitution of the clusters with a relatively very fast reaction time, and this as well for the addition of nodes as in the face of the losses and deletions of nodes of the linear wireless sensor network.
