The 21st century is witnessing the emergence of a wide variety of wireless sensor network (WSN) applications ranging from simple environmental monitoring to complex satellite monitoring applications. The advent of complex WSN applications has led to a massive transition in the development, functioning, and capabilities of wireless sensor nodes. The contemporary nodes have multi-functional capabilities enabling the heterogeneous WSN applications. The future of WSN task assignment envisions WSN to be heterogeneous network with minimal human interaction. This led to the investigative model of a deep learning-based task assignment algorithm. The algorithm employs a multilayer feed forward neural network (MLFFNN) trained by particle swarm optimization (PSO) for solving task assignment problem in a dynamic centralized heterogeneous WSN. The analyses include the study of hidden layers and effectiveness of the task assignment algorithms. The chapter would be highly beneficial to a wide range of audiences employing the machine and deep learning in WSN.
TopIntroduction
The smart computing era is massively led by the inception of smart devices in everyday life. The smart computing era infers to a world, powered by a collective set of semi-autonomous or autonomous devices, with the ability to function with minimal or no human intervention(Friday, Wah, Al-garadi, & Rita, 2018). The smart devices compared to its legacy counterparts inherently have sensing, processing, and communication capabilities(Issac, Silas, & Rajsingh, 2019a). Nowadays, contemporary, smart devices are mounted with dedicated artificial intelligence engines to provide decision making functionality without major human interventions. The transfiguration of the everyday device into a smart device is pictorially represented in figure 1.
Figure 1. The evolution from every-day device to a smart device
A wide range of autonomous smart devices has been visualized from time to time based on human requirements(Sohraby, Minoli, & Znati, 2007). Since the inception of smart devices, sensors have been playing a critical role in everyday devices(Sohraby et al., 2007). The inception of a wide variety of sensors along with its functionality and applications has given rise to multiple research domains such as the Wireless Sensor Network (WSN), Wireless Sensor and Actuator Network (WSAN), Body Area Network (BAN), Vehicular Ad-hoc Network (VANETS), etc(Munir, Gordon-Ross, & Ranka, 2014). However, this chapter concentrates on WSN, for the following set of reasons.
- 1.
WSN continues as one of the prime and evolving research domain(Bhushan & Sahoo, 2019c, 2020).
- 2.
The contemporary nodes have better capability compared to their legacy counterparts(Munir et al., 2014; Roozeboom et al., 2013).
- 3.
The envisioned next-generation WSN applications require a set of heterogeneous wireless sensor nodes(Misra & Vaish, 2011; Tkach & Edan, 2020).
Motivation
The primary notion assuming a WSN to be a resource-constrained network still prevails, however, the contemporary nodes in the WSN have comparatively gained higher processing, storage, and energy capabilities, as well as, multi-sensing capabilities, in comparison to the legacy wireless sensor nodes(Munir et al., 2014; Roozeboom et al., 2013). To this end, the need to consider the contemporary nodes’ additional capabilities during the task assignment arises, as the majority of the existing algorithms are confined to homogeneous WSN(Yu & Prasanna, 2005). The primary objective of such next-generation task assignment algorithms would be to maximize the performance and minimize the energy utilization. In a nutshell, the systematic overview of the heterogeneous WSN along with the investigation on the various existing WSN task assignment methods and the future directions of the task assignment algorithms is needed.
Major Contributions
The contributions of the chapter are manifold and are enumerated below;
- 1.
The genesis of computing and Wireless Sensor Network has been presented.
- 2.
Various types of task assignment in WSN along with factors influencing the task assignment were identified and presented.
- 3.
The past and recent task assignment methods have been identified and investigated.
- 4.
Future perspective of task assignment using deep learning methods has been investigated.
- 5.
A PSO trained Multi-layer Feed Forward Neural Network (MLFFNN) based task assignment algorithm has been proposed and simulated.