Open AccessOptimal detector design for molecular communication systems using an improved swarm intelligence algorithm
Author(s) -
Ntouni Georgia D.,
Paschos Alexandros E.,
Kapinas Vasileios M.,
Karagiannidis George K.,
Hadjileontiadis Leontios J.
Publication year - 2018
Publication title -
micro and nano letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.25
H-Index - 31
ISSN - 1750-0443
DOI - 10.1049/mnl.2017.0489
Subject(s) - particle swarm optimization , algorithm , detector , swarm intelligence , swarm behaviour , computer science , acceleration , focus (optics) , mathematical optimization , variable (mathematics) , process (computing) , mathematics , artificial intelligence , telecommunications , mathematical analysis , physics , classical mechanics , optics , operating system
The authors optimise the detection process of diffusion‐based molecular communication systems utilising the weighted sum detector with appropriate weight values. Interestingly, no optimisation technique has ever been proposed for the calculation of the weights. To this end, they build on the standard particle swarm optimisation (PSO) technique and propose a robust iterative optimisation algorithm, called acceleration‐aided PSO (A‐APSO). While modified swarm‐based optimisation algorithms focus on slight variations of the standard mathematical formulas, in A‐APSO, the acceleration variable of the particles in the swarm is also involved in the search space of the optimisation problem. Particularly, they implement the A‐APSO algorithm to evaluate the detector's weights that minimise the closed‐form expression of the error probability. Their findings reveal that, when employing the A‐APSO weights, the error performance is superior to that achieved by utilising the weight values already existing in the literature or those evaluated with the standard PSO algorithm.