42
Internet of Things (IoT)
3.1 Introduction
With the rapid development of Internet of Things (IoT), it has now become a buzzword for
everyone who works in this area of research. Further, it is seen that with the rapid development of sensors and devices with their connection to IoT become a treasure trove for
big data analytics. It has found numerous applications in developing smart cities where
predictions of accidents and traffic flow in the cities can be effectively monitored; smart
health care where the doctor is able to get useful information from the implant sensor chip
in the patient’s body; industrial production can also be enhanced manifolds by efficient
prediction of the working of machinery and smart metering in helping the electric distribution company to understand the individual household energy expenses and making
smart homes with connected appliances to name a few.
The 21st century is for IoT, where it is viewed as a network of physical devices coming together from electronics, sensors, and software. It is envisioned that the network of
approximately 27 billion of physical devices on IoT are presently available and the list
grows. These devices (Cars, Refrigerators, TVs etc.) can be uniquely identifiable through
embedded computing system and can be connected from anywhere through suitable
information and communication technology, to achieve greater service and value.
The “THING” in IoT means everything and anything around us that includes machines,
buildings, devices, animals, human beings, etc. Today’s, smart health care, smart homes,
smart traffic, and smart household devices use this technology for a better digital world.
3.1.1 Working Principles of IoT
IoT has a unique identification that is embedded relying on the RFID connections which
does not need any human or human–computer intervention for its working (Ashton, 2009).
The IoT devices use IPv6 addressing for a huge address space, which makes it operational
with active monitoring by computers with network connectivity and controlled by sensors
attached to that devices. This monitoring and control can be interestingly seen in smart
home applications where one can turn on the air conditioner while returning from office
on the way home.
3.6.5 Metrics Used ............................................................................................................. 51
3.6.5.1 Forecasts .................................................................................................... 51
3.6.6 Datasets Used ........................................................................................................... 51
3.6.6.1 UMass Smart* Microgrid Data Set ........................................................ 51
3.6.6.2 Residential Energy Consumption Survey Data ................................... 52
3.6.6.3 The Reference Energy Disaggregation Data ........................................ 52
3.7 Experimental Results and Discussion .............................................................................. 52
3.7.1 Experiment 1: Prediction and Forecasting Using Deep Convolutional
Neural Network with RECS Dataset ..................................................................... 52
3.7.2 Experiment 2: Prediction and forecasting using Deep Convolutional
Neural Network with Microgrid Dataset ............................................................. 53
3.7.3 Experiment 3: Prediction and forecasting using Deep Convolutional
Neural Network with REDD Dataset ................................................................... 55
3.8 Conclusions and Future Scope........................................................................................... 56
References ....................................................................................................................................... 57
Internet of Things (IoT)
3.1 Introduction
With the rapid development of Internet of Things (IoT), it has now become a buzzword for
everyone who works in this area of research. Further, it is seen that with the rapid development of sensors and devices with their connection to IoT become a treasure trove for
big data analytics. It has found numerous applications in developing smart cities where
predictions of accidents and traffic flow in the cities can be effectively monitored; smart
health care where the doctor is able to get useful information from the implant sensor chip
in the patient’s body; industrial production can also be enhanced manifolds by efficient
prediction of the working of machinery and smart metering in helping the electric distribution company to understand the individual household energy expenses and making
smart homes with connected appliances to name a few.
The 21st century is for IoT, where it is viewed as a network of physical devices coming together from electronics, sensors, and software. It is envisioned that the network of
approximately 27 billion of physical devices on IoT are presently available and the list
grows. These devices (Cars, Refrigerators, TVs etc.) can be uniquely identifiable through
embedded computing system and can be connected from anywhere through suitable
information and communication technology, to achieve greater service and value.
The “THING” in IoT means everything and anything around us that includes machines,
buildings, devices, animals, human beings, etc. Today’s, smart health care, smart homes,
smart traffic, and smart household devices use this technology for a better digital world.
3.1.1 Working Principles of IoT
IoT has a unique identification that is embedded relying on the RFID connections which
does not need any human or human–computer intervention for its working (Ashton, 2009).
The IoT devices use IPv6 addressing for a huge address space, which makes it operational
with active monitoring by computers with network connectivity and controlled by sensors
attached to that devices. This monitoring and control can be interestingly seen in smart
home applications where one can turn on the air conditioner while returning from office
on the way home.
3.6.5 Metrics Used ............................................................................................................. 51
3.6.5.1 Forecasts .................................................................................................... 51
3.6.6 Datasets Used ........................................................................................................... 51
3.6.6.1 UMass Smart* Microgrid Data Set ........................................................ 51
3.6.6.2 Residential Energy Consumption Survey Data ................................... 52
3.6.6.3 The Reference Energy Disaggregation Data ........................................ 52
3.7 Experimental Results and Discussion .............................................................................. 52
3.7.1 Experiment 1: Prediction and Forecasting Using Deep Convolutional
Neural Network with RECS Dataset ..................................................................... 52
3.7.2 Experiment 2: Prediction and forecasting using Deep Convolutional
Neural Network with Microgrid Dataset ............................................................. 53
3.7.3 Experiment 3: Prediction and forecasting using Deep Convolutional
Neural Network with REDD Dataset ................................................................... 55
3.8 Conclusions and Future Scope........................................................................................... 56
References ....................................................................................................................................... 57