Adaptive Wireless Sensor Network and Internet of Things

Just within last 5-10 years, we are having intelligent systems in every part of our world ranging from intelligent transportation, smart homes, smart manufacturing, and intelligent network with wearable sensors, etc. These networks are making our home much easier than expected. They are powered with different technologies such as Wireless Sensor network [1], intelligent system and Internet of Things commonly known as IoT [2]. It was assumed that by the end of 2020, we will be connected with billions of devices that makes whole world connected. Another Promising technology is wireless sensor network. Generally speaking, wireless sensor network is the network of the distributed wireless nodes that are connected to the gateway called sink where nodes collect data and send data to the server using gateway. Sensors are what makes wireless sensor network special. These days, general computing devices such as mobiles are equipped with the computation power and sensors which are used in general life.

The common vision of using IoT and wireless sensor networks is to incorporate two major technologies i.e. communication and information technology where managing and monitoring physical systems are connected by network devices so the whole technology can be operated and managed with the tip of the hand. With this revolutionary age of computing and data, massive data are collected from internet, smartphones, etc. With the massive use of IoT technologies, many data are yet to be gained. By embedding computational capabilities in such kind of objects, it will provide a leap in every way possible. With the advent of IPV6 network, billions of objects can get networking ID which is very easy to incorporate with the current networking system. Not Just any device, one can incorporate devices with sensors in own body improving private health care [3]. Body Area Network focuses on the use of the devices with sensors that can track the physical state of their body. There are more demands of sensors that are embedded in systems that can be worn in body and it will find physical state of the person who wears. Section 2 deals with the wireless sensor network (WSN) [1] and IoT, their basic characteristics and general services provided by WSN and IoT. Section 3 deals with the approaches by integrating WSN and IoT to become self-organizing adaptive system. Section 4 deals with the challenges to be faced adapting merging system.

Wireless Sensor Network or commonly known as WSN [1] is the network of nodes which is the combination of Sensors, Microcontroller, radio transceiver and power source which are placed or scattered spatially to acquire data from sensors which senses different phenomenon like temperature, light, etc., make computation to sample the data and finally sending to the servers or base station through gateways. The uses of the WSN have proliferated in few years to support many applications such as Environmental sensing, Healthcare systems, Industrial monitoring and many more. With such success, progress and development of WSN, there are yet other areas that can be automatized. With the advent of new technologies, mobile systems have gained more attention. More people are using mobile systems specially “Smart Phones” as tons of money and intelligence are spent on the development of next generation mobile devices that will simplify our task. The areas that WSN have impacted in our daily life have increased immensely. Some of the mostly used area that are personal are as follows.

Personal health monitoring

This is the area that is really hitting the most. The concept of smart wear or wearable sensors or fairly modern concept of “Body Sensor Network” is really hitting it off. With the help of sensor network, there will be ease in monitoring and assisting patient so there will not be any troubles for doctors to get in touch with patients or assisting them in any kind [4,5]. With ubiquitous health care monitoring system, doctor can be abreast of health status and make real-time decision [6] E.g. Diet Sense [7] helps people to track the change in the body weight through sound and images.

Environmental monitoring

Sensor Networks have affected the way we take environmental concerns. Recording data from weathers, transportations, road traffic flows, different emission of gases, etc. and assisting with the data can help user to make smart choices. Some of the evolving technology is PEIR (Personal Environmental Impact Report) [8] which assist users to find out the exposure to pollutants through mobile phones.

Enriching social media

Sensors have enriched the content in social media using smart phones. Some of the multimodal information generated are acceleration, audio samples, location, neighboring devices, etc. Some of the upcoming technology is the affective computing which senses the feelings of the user along with ambient intelligence [3].

Smart home

The most promising technology is the concept of smart home where the concept of WSN is immersed with the concept of IoT. Many small networking devices connected with each other to make networking at home. With the help of Smartphone, meaningful information is shown to users who later can keep track of things. The concept of ambient intelligence is booming where it refers to electronic environment that are sensitive to the presence of the people. On the other hand, IOT is networking of embedded devices which are “things” with the other electronics, sensor which is to enable exchanging data between different parties. The main goal of IoT is to control different devices remotely rather than being physically present. The backbone of internet of things is internet where devices connect with other devices through internet and control remotely. Application of IoT is to make everything smarter and control everything from the tip of user’s hand [9].

Today, everything around us is electrified or digitized. Almost any manufactured good now includes an embedded processor (typically a microcontroller, or MCU), along with user interfaces, that can add programmability and deterministic “command and control” functionality. The electrification of the world and the pervasiveness of embedded processing are the keys to making objects “smart. Some of the application areas of IoT are Smart Watches, location tracking, fitness, Light and temperature control, traffic control, real time inventory, remote monitoring, etc.

One of the most important elements in the IoT paradigm is wireless sensor networks (WSN). The benefits of connecting both WSN and other IoT elements go beyond remote access, as heterogeneous information systems can be able to collaborate and provide common services [10].

IoT is all about integrating different component (mostly electrical) where WSN is one the major component. There are different factor leading to the success of the integration such as the physical layer of integrating all device such as networking. It is necessary to find out on what basis the integration is being done. Is it on the basis of topology [11] or different method such as stack-based approach [10]? With different heterogeneous devices connecting with each other, heterogeneous data are created and managing such big data is still a subject of concern. There are lots of layers on integrating level: Starting from network layer to making use of the data produced using different techniques. The goal of this paper is to make IoT adaptive, so we will exclude other factors and focus just on the machine learning part and data.

Figure 1:Framework for the integration of IoT and WSN to make adaptive system.

As we can see in (Figure 1), it is the general framework for making adaptive system which use IoT where WSN is an integral part. In the system, system takes input from different sensors and devices such as Door Sensor, Infrared Sensor, etc. from House, RFID, Environment sensors, sensors from health care, etc. Middleware makes easy interface to take input as sensors from different devices. With such data, multi-modal sensors are accumulated and saved into repository and stored into the database or any storing mechanism. System is provided with the personal preference mechanism and benchmarks. Benchmarks algorithms are used. We can use both structured and unstructured learning mechanism to learn data sets gathered. From the learning, we get the valuable information. For the adaptive IoT, we have used the feedback mechanism where the system learns from the historical data and keeps on learning repeatedly.

Making adaptive systems is a good thing as we get everything in control but there are different challenges and personal preference are used to invoke the event. To analyze and make use of data, different machine learning which are on the way to implement it. Some of the challenges are:

Security challenges

There are different challenges bounded with the security issues. Some of the challenges are integration of security mechanisms. There are different types of security mechanisms provided by the wireless node but as they are included with the RFIDs, different kind of integration is required. Different security mechanisms are needed to integrate otherwise minimum security can be obtained.

Preserving user’s privacy

Main issue while integrating IoT and WSN is dropping privacy concern of user as every data from each sensor is passed to the repository to be analyzed. Every location detail, user activity and other activities can create ample amount of data. Without the secure mechanism to protect user privacy, data generating device can be source to the source of spies where every detail are sent to the wrong hand. Many users are aware of this scenario which makes then reluctant of using services. The authors in [9] present a comprehensive application independent architecture for anonymous tasking and reporting. The infrastructure enables applications to task a mobile device using a new tasking language, anonymously distribute tasks to mobile devices and collect anonymous yet verifiable reports from the devices.

Energy problem

Energy is the main problem using wireless sensor network as node that sense data and send to the server uses 2A batteries which have limited energy to sustain. There must be some sort of energy conservation solutions that help systems to conserve less energy and increase productivity.

Merging two systems, IOT and WSN to product intelligence and productivity is a great thing but with that there are different nonfunctional challenges that should be overcome. The way innovating approaches are on way to make this successful, there are still some lags behind the integration of different heterogeneous devices mining data to make affective and social decisions to help our daily life.

1. Wireless sensor networks.
2. Internet of things.
3. Ambient intelligence.
4. Movassaghi S, Mehran A, Justin L, David S, Jamalipour A, et al. (2013) Wireless body area networks: A survey. IEEE Communications Surveys & Tutorials 16(3): 1658-1686.
5. Joonyoung J, Kiryong H, Jeonwoo L, Youngsung K, Daeyoung K (2008) Wireless body area network in a ubiquitous healthcare system for physiological signal monitoring and health consulting. International Journal of Signal Processing, Image Processing and Pattern Recognition, pp.1-8.
6. Roozbeh J, Encarnacao A, Zahoory A, Dabiri F, Noshadi H, et al. (2004) Wireless sensor network for health monitoring, University of California, Los Angeles Department of Computer Science, Los Angeles, California, USA.
7. Reddy S, Parker A, Hyman J, Burke J, Estin D, et al. (2007) Image browsing, processing and clustering for participatory sensing: Lessons from a Diet Sense Prototype. In: Proceedings of the Workshop on Embedded Networked Sensors (Em-NetS), Cork, Ireland.
8. Mun M, Reddy S (2009) PEIR, the personal environmental impact report, as a platform for participatory sensing systems research. In: Proceedings of ACM MobiSys, Krakow, Poland, pp.1-14.
9. Kapadia A, Triandopoulos N, Cornelius C, Peebles D, Kotz D (2008) Anony sense: opportunistic and privacy-preserving context collection. In: Indulska J, Patterson DJ, Rodden T, Ott M (Eds.), Pervasive, LNCS, Heidelberg, Germany, 5013: 280-297.
10. Roman R, Lopez J (2009) Integrating wireless sensor networks and the internet: A security analysis. Internet Research, 19(2): 246-259.
11. Christin D, Reinhardt A, Mogre PS, Steinmetz R (2009) Wireless sensor networks and the internet of things: Selected challenges. 8th GI/ITG KuVS Fachgesprch, Drahtlose Sensornetze, pp.1-3.