Cluster-based routing keeps a farm-scale Wireless Sensor Network (WSN) alive by electing a small set of Cluster Heads (CHs) to relay data toward the Base Station (BS), yet most reported evaluations of such schemes stop at a few hundred nodes on a 100 m × 100 m plot– far smaller than a working commercial field. This paper re-examines the Fractional Dung Namib Optimizer (FDNO) routing framework of Lavanya and Rao [1] under a substantially heavier load: 1000 sensor nodes scattered over a 200 m × 200 m field, optimized with a swarm of 75 candidate agents for 1000 rounds. The pipeline retained from [1] first forecasts each node’s remaining energy with a recurrent neural predictor, groups nodes into clusters through a Dung Beetle / Namib Beetle hybrid search, elects a CH per cluster from a joint score over residual energy, forecast energy, trust, hop delay and inter-node distance, and finally threads a route between clusters using a fractional-order variant of the same hybrid search. To present the derivation in an independent notation rather than reproduce the original symbols verbatim, every quantity in Sections III and IV is renamed and re-derived under a notation table defined for this paper. Running the framework at the enlarged scale yields a route delay of 0.741 ms and hop distance of 33.958 m, alongside a retained per-node energy of 0.158 J and a trust score of 91.874 at round 1000– each better than the four benchmark schemes (SPSRN, EERPMS, FMCBER, ACO) evaluated alongside it, indicating that the clustering and routing search degrades gracefully as the deployment is pushed toward field sizes and node counts typical of real farms.

wireless sensor network, cluster routing, cluster head election, dung beetle optimizer, precision agriculture, fractional-order search