The Node Placement of Large-scale Industrial Wireless Sensor Networks Based on Binary Differential Evolution Harmony Search Algorithm

This paper discusses the optimal node placement problem of Industrial Wireless Sensor Networks (IWSNs). Compared with the non-industrial Wireless Sensor Networks (WSNs), IWSNs have a critical requirement on the reliability of networks. Thus, a new model of the IWSNs node placement is formulated in which the reliability, cost, load constraint and scalability are taken into account. Considering its NP-hard characteristic, a novel hybrid Binary Differential Evolution Harmony Search Algorithm (HBDEHS) is presented to tackle the optimal sensor deployment problem. Four large-scale node deployment problems are generated as the benchmarks to verify the proposed model and the optimization algorithm. Furthermore, the other five binary optimization algorithms, i.e., Global Harmony Search Algorithm (NGHS), Modified Binary Differential Evolution Algorithm (MBDE), Discrete Binary Harmony Search Algorithm (DBHS), Discrete Binary PSO algorithm (DBPSO) and Simple Genetic Algorithm (SGA) are also adopted to solve the problems for a comparison. The experimental results indicate that all algorithms can find out the feasible solutions, which demonstrates that the proposed model is valid and can be effectively used to tackle the optimal node placement problem. Moreover, the comparison results also illustrate that HBDEHS has the best global search ability and outperforms DBHS, NGHS, MBDE, DBPSO and SGA in terms of search accuracy and convergence speed.