Effective algorithm for optimizing compressive sensing in IoT and periodic monitoring applications. (15th January 2019)

• Record Type: Journal Article
• Authors: Aziz, Ahmed; Singh, Karan; Osamy, Walid; Khedr, Ahmed M.
• Is Part Of: Journal of network and computer applications. Volume 126(2019)
• Page Start: 12
• Page End: 28
• Publication Date: 2019-01-15
• Subjects: Chicken swarm optimization algorithm -- Cluster-based wireless sensor network -- Compressive sensing matrix -- Data gathering -- Hybrid-CS -- Plain-CS -- Internet of things -- Reconstruction process; Microcomputers -- Periodicals; Computer networks -- Periodicals; Application software -- Periodicals; Micro-ordinateurs -- Périodiques; Réseaux d'ordinateurs -- Périodiques; Logiciels d'application -- Périodiques; Application software; Computer networks; Microcomputers; Periodicals; 004.05; 004
• DOI: 10.1016/j.jnca.2018.10.013
• Languages: English
• ISSNs: 1084-8045

Abstract

Abstract: Compressive Sensing (CS) has succeeded in presenting itself as an adequate method for Internet of Things (IoT) mainly because CS can reduce the size of raw data transmission and achieve traffic load balancing throughout networks. A recent work of CS discussed integration of CS with clustering techniques to achieve additional power saving during the transmission process. For data aggregation, CS method can be combined with cluster-based algorithms in plain CS or hybrid CS. However, total energy consumption for data aggregation using plain CS is still very large. In additional, hybrid CS data aggregation is efficient only if the cluster head (CH) receives a large enough data vector, i.e., greater than or equal to the CS measurement size. Otherwise use of CS leads to larger amount of transmissions by the CHs. In this paper, we propose a new Cluster Size Load Balancing for CS algorithm (CSLB-CS) which could achieve optimal utilization of CS method in an IoT-based sensor network. CSLB-CS includes a cluster load balancing technique that reduces total number of transmissions and improves the reconstruction process by optimizing the CS matrix. Chicken Swarm Optimization Algorithm that outperforms the other swarm algorithms in terms of optimization accuracy and robustness is used to optimize the CS matrix. Detailed mathematical and comprehensive experimental analyses are provided to demonstrate efficiency of the proposed algorithm. Simulation results indicate that the proposed algorithm exceeds the performance of hybrid CS and plain CS in terms of network lifetime, overall energy consumption, total number of data transmitted, and reconstruction error.