Amine Bchaiker

Amine Bchaiker

Auto-generated profile, based on publicly available information

Amine Bchaiker


Research Focus & Metrics

Academic Career Length

1Year (2026-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
25 June 2026 in SSRN Electronic Journal
Verified
Preprint
Open access
ABAmine Bchaiker
Amine Bchaiker
AEAlami Essaadoui
Alami Essaadoui
AAAziz EL Afia
Aziz EL Afia
PMPrince Marfo
Prince Marfo
Wireless Sensor Networks (WSNs) deployed in real-world settings face heterogeneous operatingconditions in which conventional cluster-head (CH) selection based on instantaneousmetrics cannot anticipate short-term reliability degradation, leading to prematureCH failures and reduced network lifetime. We propose PSO-LSTM, a proactive reliabilitygatedextension of PSO-EECS, in which a lightweight Long Short-Term Memory (LSTM)model at the base station predicts short-horizon node failure probabilities from per-nodetemporal signals. These predictions form a set-level reliability score that multiplicativelygates the energy-centric term of the PSO fitness, recovering PSO-EECS behavior understable conditions. Experiments are conducted over 20 paired seeds on a 200-node heterogeneousWSN incorporating six co-occurring heterogeneity factors: energy variability, spatialhotspots, stochastic traffic, retransmission faults, energy harvesting, and asymmetric sinkplacement. Compared with seven bio-inspired baselines, namely PSO with GravitationalSearch Algorithm (PSOGSA), Grey Wolf Optimizer (GWO), Whale Optimization Algorithm(WOA), Ant Lion Optimizer (ALO), Harris Hawks Optimization (HHO), PSO-EECS, and anablated no-gate variant, PSO-LSTM achieves statistically significant improvements on FirstNode Dead (FND) (Δ ≥ +4.68%, padj ≤ 0.023, δ ≥ +0.761), Last Node Dead (LND)(Δ ≥ +6.21%, padj ≤ 0.001, δ = +1.000), and cumulative throughput (Δ ≥ +2.50%,padj ≤ 0.001, δ = +1.000), while introducing no statistically significant Half Node Dead(HND) degradation (padj ≥ 0.088), confirming the gate’s non-disruptiveness during stablenetwork operation. Ablation analysis confirms that the reliability gate is the key mechanismdriving these improvements.