Lowest-ID with Adaptive ID Reassignment: A Novel Mobile Ad-Hoc Networks Clustering Algorithm

Clustering is a promising approach for building hierarchies and simplifying the routing process in mobile ad-hoc network environments. The main objective of clustering is to identify suitable node representatives, i.e. cluster heads (CHs), to store routing and topology information and maximize clusters stability. Traditional clustering algorithms suggest CH election exclusively based on node IDs or location information and involve frequent broadcasting of control packets, even when network topology remains unchanged. More recent works take into account additional metrics (such as energy and mobility) and optimize initial clustering. However, in many situations (e.g. in relatively static topologies) re-clustering procedure is hardly ever invoked; hence initially elected CHs soon reach battery exhaustion. Herein, we introduce an efficient distributed clustering algorithm that uses both mobility and energy metrics to provide stable cluster formations. CHs are initially elected based on the time and cost-efficient lowest-ID method. During clustering maintenance phase though, node IDs are re-assigned according to nodes mobility and energy status, ensuring that nodes with low-mobility and sufficient energy supply are assigned low IDs and, hence, are elected as CHs. Our algorithm also reduces control traffic volume since broadcast period is adjusted according to nodes mobility pattern: we employ infrequent broadcasting for relative static network topologies, and increase broadcast frequency for highly mobile network configurations. Simulation results verify that energy consumption is uniformly distributed among network nodes and that signaling overhead is significantly decreased.

💡 Research Summary

The paper addresses two persistent challenges in mobile ad‑hoc networks (MANETs): rapid energy depletion of cluster heads (CHs) and excessive control traffic caused by frequent reclustering. Traditional clustering schemes such as Lowest‑ID (LID) are attractive for their simplicity and fast initial cluster formation, but they suffer from a bias toward low‑ID nodes, which become CHs repeatedly and exhaust their batteries. More sophisticated methods like Highest‑Degree (HD) and Weighted Clustering Algorithm (WCA) incorporate mobility and residual energy metrics, yet they introduce higher computational complexity and often trigger reclustering without clear criteria, leading to increased signaling overhead.

The authors propose a novel algorithm called LID‑AR (Lowest‑ID with Adaptive ID Reassignment). LID‑AR retains the fast, low‑cost initial clustering of the classic LID approach: each node broadcasts a “Hello” packet during a Hello Period (HP), and the node with the smallest ID among its neighbors is elected CH. This step requires only two HPs and minimal control packets.

During the maintenance phase, every node (v) computes a weighted score (W_v) that reflects both its remaining battery (B_v) and its recent average mobility (M_{v,P}) over the last (k) HPs:

\