A simulation study published in Peer-to-Peer Networking and Applications demonstrates that an adaptive bio-inspired heuristic model can significantly improve the accuracy of unknown node localization while simultaneously routing communication paths around physical barriers in wireless sensor networks. Authored by Sathya Prakash Racharla and Kalaivani Jeyaraj, the investigation introduces the Alpha and Beta Horse Herd Optimization algorithm, designated as ABHHO, to address persistent spatial coordination challenges that arise in dense or obstructed sensor deployments. Wireless sensor networks serve critical operational roles across healthcare monitoring, industrial automation, agricultural surveillance, environmental tracking, and defense systems. Despite the widespread utility of these sensor systems, determining the precise geographic coordinates of distributed nodes remains a substantial technical bottleneck, particularly when high node density, limited battery reserves, multipath signal propagation, and non-line-of-sight obstructions degrade conventional positional calculations. The researchers developed their framework to balance multi-objective demands without relying on single-objective trade-offs that often impair network lifespan or transmission reliability.
Sensor node localization typically relies on establishing spatial relationships between a small subset of anchor nodes possessing predetermined coordinates and numerous unknown nodes whose locations must be estimated mathematically. While satellite-based systems such as the Global Positioning System offer dependable positioning in open outdoor environments, their substantial hardware installation costs and high operational energy draw make them unsuitable for energy-constrained wireless sensor nodes. Prior research methodologies have explored geometric principles, machine learning models, and standard metaheuristic optimization techniques to resolve unknown coordinates, but these conventional approaches frequently struggle with severe computational overhead, poor scalability across large geographic surfaces, coplanar anchor arrangements, and high sensitivity to physical interference. Furthermore, classic optimization routines routinely become trapped in local optima when searching irregular three-dimensional terrains where physical obstacles impede line-of-sight signal propagation. Racharla and Jeyaraj formulated their approach to overcome these specific performance boundaries by integrating node coordinate estimation with an adaptive node segmentation and obstacle-aware routing scheme.
The mathematical foundation of the ABHHO model builds upon the standard Horse Herd Optimization Algorithm, a metaheuristic framework that models the collective herding, grazing, and social behaviors observed among horses across distinct age brackets. In the baseline optimization design, simulated agents adjust their positional trajectories based on six fundamental natural behaviors: grazing, hierarchical structure, sociability, imitation, roaming, and defensive mechanisms. While standard horse herding algorithms can handle high-dimensional problem domains by sorting candidate solutions, they frequently exhibit premature convergence and prove incapable of resolving complex multi-objective optimization equations. To resolve these operational constraints, the authors re-engineered the algorithmic architecture by introducing dual-leadership dynamics wherein an Alpha horse leads global exploration across the broader search environment while a Beta horse governs local exploitation through targeted iterative adjustments. This dynamic leadership structure prevents search agents from stagnating in suboptimal candidate coordinates when traversing complex spatial layouts.
To formally implement these behavioral adaptations, the researchers introduced a refined mathematical definition for the sociability factor within the algorithm’s position-updating equations. In the modified formula, sociability is governed by the ratio of defense mechanism parameters to the combined values of grazing and hierarchical behaviors, operating within designated bounded ranges including a defense limit of 0.5, a grazing span of 1.50, and a hierarchy ceiling of 0.9. Additionally, the authors incorporated an Obstruction-aware Correction Factor into the distance estimation module to dynamically compensate for physical wave degradation. In operational wireless deployments, radio signals experience severe diffraction, reflection, and shadowing when traveling past solid structures, causing measured received signal strength indicators to deviate markedly from theoretical free-space expectations. By calculating the real-time deviation between predicted and observed signal values under shifting environmental factors, the correction factor systematically recalibrates intermediate distance calculations prior to final coordinate optimization.
The system evaluates candidate solutions through a specialized multi-objective fitness function that simultaneously tracks four vital operational parameters: network coverage rate, average localization error, quality of service, and hop count. The mathematical objective function employs a deliberate fractional architecture that balances competing optimization goals without requiring arbitrary weighting coefficients or explicit normalization steps. In this fractional construction, the numerator aggregates the average localization error and cumulative hop count, designating them as values that must be minimized to preserve energy and precision. Conversely, the denominator aggregates the estimated convergence rate and the composite quality-of-service score, treating them as system objectives that must be maximized. The composite quality-of-service metric is calculated through weighted coefficients tracking packet delivery ratios, transmission delays, throughput volumes, and sensor energy consumption derived from a standardized first-order radio model.
Alongside the primary coordinate optimization engine, the framework executes an adaptive anchor node segmentation protocol to structure the physical network environment and identify obstacle-free transmission corridors. Anchor nodes move along defined paths to gather spatial reference samples, after which the deployment region is partitioned into uniform geometric cells using a grid-based spatial approach. The algorithm evaluates variations between actual geometric distances and estimated distances across neighboring anchor pairs, identifying boundary conditions where discrepancy thresholds are exceeded. Following spatial segmentation, the system systematically removes suboptimal anchor nodes according to strict elimination criteria: anchors exhibiting redundant geographic coverage overlap are purged to reduce duplicate communication overhead; nodes positioned beyond designated communication radii are excised due to diminished reference utility; and anchors contributing negligible improvements to overall coverage or quality of service are deactivated to limit computational load.
The computational performance and localization fidelity of the ABHHO framework were verified through extensive numerical simulations conducted within the MATLAB 2020a software environment. The simulation architecture configured the optimization routine with an agent population size of 10, a maximum operational threshold of 100 iterations, and a chromosome structure consisting of 13 specific genes. The physical layer parameters were standardized using an ambient environmental temperature of 290 Kelvin to maintain uniform thermal noise, a communication channel bandwidth of 2 megahertz, an initial node energy reserve of 1.5 joules, a radio transmission radius fixed at 25 meters, and a reference deployment comprising 15 active anchor nodes. Within the 13-gene chromosome representation, specific gene sequences encoded the spatial coordinates of unknown nodes, indexing metrics for the three nearest anchors, calculated distance values, initial routing trajectories, obstacle flags, coverage factors, and weight parameters for quality of service and hop counts.
Across iterative testing runs, the ABHHO algorithm demonstrated pronounced improvements over standard metaheuristic routines, including the Dragonfly Algorithm, the Hippopotamus Optimization Algorithm, the Water Strider Algorithm, and the baseline Horse Herd Optimization Algorithm. At an operational midpoint of 50 iterations, the cost function value achieved by the ABHHO framework was reduced by 22 percent compared to the Dragonfly Algorithm, 33 percent relative to the Hippopotamus Optimization Algorithm, 27 percent against the Water Strider Algorithm, and 11 percent compared to standard horse herd optimization. Concurrently, the network coverage ratio recorded at iteration 50 increased by 2.1 percent over the Dragonfly benchmark, 4.2 percent over the Hippopotamus algorithm, 4.2 percent over the Water Strider method, and 5.3 percent compared to the unenhanced horse herd framework, confirming superior spatial monitoring under simulated signal attenuation, interference, and surface irregularities.
When examining localization accuracy across varying network densities, the proposed model maintained consistent advantages over existing techniques. For an experimental setup comprising 20 sensor nodes, the average localization error recorded by the ABHHO system decreased by 32 percent compared to the Dragonfly Algorithm, 38 percent against the Hippopotamus model, 44 percent versus the Water Strider approach, and 50 percent relative to baseline horse herd optimization. In broader comparative trials against recent metaheuristics including the Slime Mould Algorithm, the Aquila Optimizer, Hybrid Pelican-Particle Swarm Optimization, and the Dynamic Arithmetic Optimization Algorithm, the full ABHHO implementation recorded an average localization error of 0.82 ± 0.04 meters while completing optimization execution in 8.9 ± 0.7 seconds. Furthermore, in true-position benchmarking trials, ABHHO attained an average localization error of 1.98 meters, outperforming Particle Swarm Optimization at 3.12 meters, Grey Wolf Optimizer at 2.84 meters, and standard horse herd optimization at 2.53 meters, representing an overall 36.54 percent improvement over baseline performance.
The authors evaluated the internal algorithmic contributions through an ablation study that systematically isolated the individual behavioral components of the optimization model. When the algorithm was executed without the alpha leadership strategy, the average localization error increased to 0.97 ± 0.06 meters; similarly, removing the beta leadership strategy caused the localization error to rise to 0.95 ± 0.05 meters, compared to the 0.82 ± 0.04 meters achieved by the complete model with an associated coverage rate of 95.6 ± 1.3 percent. Computational complexity analysis demonstrated that the ABHHO model operates within an asymptotic complexity bound of O[Mi + Nl + Cz], where Mi represents maximum iterations, Nl denotes population size, and Cz indicates chromosome length. Across extensive simulation benchmarks, the proposed algorithm recorded an overall execution runtime of 39.97 minutes, contrasting with runtimes of 72.4 minutes for the Dragonfly Algorithm, 60.7 minutes for the Hippopotamus Optimizer, 57.4 minutes for Dijkstra’s algorithm, and 43.9 minutes for the A* search procedure.
Statistical evaluation further confirmed the operational stability and distinct performance characteristics of the proposed architecture. Across sample configurations scaled between 20 and 100 nodes, the ABHHO model achieved parametric test metrics including a two-sample t-test value of 0.39 and an analysis of variance outcome yielding a p-value of 0.0033, indicating statistically measurable differences when compared against competing algorithms. In obstacle avoidance assessments where simulated barrier density was increased from 10 percent to 40 percent of the deployment environment, average path lengths expanded from 72.4 meters to 88.7 meters, while the algorithm maintained stable collision avoidance and low routing failure rates. Scalability trials involving 100 distributed sensor nodes documented an overall task success rate of 99.8 percent alongside a localized computation time of 7.2 ± 0.4 seconds, validating the algorithmic efficiency when managing higher node volumes in multi-hop configurations.
Despite the positive performance reported across the MATLAB simulations, Racharla and Jeyaraj noted several technical limitations that restrict immediate real-world deployment. The current experimental findings rely exclusively on numerical simulations and lack empirical validation against physical hardware testbeds, sensor prototypes, or dynamic real-world environmental datasets. The authors observed that while the system functions effectively within structured and simulated bounds, positional accuracy degrades when nodes are placed across highly irregular surfaces, and network performance decreases when subjected to rapid topological shifts or unpredictable link disruptions characteristic of highly heterogeneous sensor deployments. Future research will focus on integrating physical hardware-in-the-loop testing, evaluating real-world propagation measurements, and formulating hybrid optimization mechanisms to enhance tracking reliability under dynamic field conditions.
Subject of Research: Obstacle-aware node localization and multi-objective path planning in wireless sensor networks using an enhanced optimization algorithm
Article Title: Development of obstacle-aware localization with node segmentation model performing multi-objective optimal path planning in wsn using dynamic optimization algorithm
Article References: Racharla, S. P., & Jeyaraj, K. (2026). Development of obstacle-aware localization with node segmentation model performing multi-objective optimal path planning in wsn using dynamic optimization algorithm. Peer-to-Peer Networking and Applications, 20(1), Article 1. https://doi.org/10.1007/s12083-026-02308-5
Image Credits: AI Generated
DOI: 10.1007/s12083-026-02308-5
Keywords: wireless sensor networks, node localization, path planning, swarm intelligence, horse herd optimization, obstacle avoidance, node segmentation, multi-objective optimization, quality of service, received signal strength, energy efficiency, first-order radio model, matlab simulation, metaheuristic algorithms, hop count, computational complexity, coverage ratio, packet delivery ratio
Cite Scienmag News
William Thompson. (November 19, 2026). Optimized Horse Herd Algorithm Cuts Sensor Localization Error in Complex Networks. Scienmag. https://scienmag.com/optimized-horse-herd-algorithm-cuts-sensor-localization-error-in-complex-networks/
William Thompson. "Optimized Horse Herd Algorithm Cuts Sensor Localization Error in Complex Networks." Scienmag, 19 November 2026, https://scienmag.com/optimized-horse-herd-algorithm-cuts-sensor-localization-error-in-complex-networks/. Accessed 19 November 2026.
William Thompson. "Optimized Horse Herd Algorithm Cuts Sensor Localization Error in Complex Networks." Scienmag. November 19, 2026. https://scienmag.com/optimized-horse-herd-algorithm-cuts-sensor-localization-error-in-complex-networks/








