AMSEC: Adaptive Multi-Score, Unequal Clustering for WSNs

Adaptive Multi-Score, Unequal Clustering (AMSEC) algorithm is a novel energy-efficient clustering method for wireless sensor networks..

The proposed Adaptive Multi-Score, Unequal Clustering (AMSEC) algorithm is a novel energy-efficient clustering method for wireless sensor networks with 500 randomly deployed sensor nodes. The base station (BS) is located outside the sensing field, and communication follows the first-order radio model. In each round, the algorithm first identifies the set of alive nodes and evaluates each candidate’s suitability for becoming a cluster head (CH) based on a multi-score metric that considers four weighted factors: residual energy (to prioritize healthier nodes), local neighbor density (to promote well-connected CHs), inverse distance to the BS (to encourage far nodes to become CHs and balance load), and a recent-CH penalty (to avoid repeated selection of the same nodes). A target cluster head percentage is enforced using an adaptive percentile gating mechanism. From the top candidates, spatial contention thinning is applied: each CH must be a minimum guard distance away from others, with the guard radius adapted according to the node’s distance to the BS (near-BS CHs get larger guard zones to prevent excessive clustering in the center). The final CH set forms clusters using an unequal cluster radius strategy, where nodes closer to the BS have smaller cluster ranges and far nodes have larger ones, reducing the burden on near-BS CHs. Non-CH nodes join the nearest CH within its cluster radius; otherwise, they transmit directly to the BS. Energy consumption is calculated for all transmissions, receptions, and data aggregation, and node energies are updated accordingly. Over multiple rounds, this adaptive process balances energy usage, delays the first node death, and extends the network lifetime, which is validated by plotting the number of alive nodes and residual energy versus rounds.