03-10-2016, 12:44 PM
1457501033-FinalPaper.doc (Size: 86.5 KB / Downloads: 5)
Abstract-The Internet of Things is a concept that is fast gaining recognition in the world today. It allows the various entities to be connected to each other through a network preferably the internet. Such is the popularity of the technology that by 2020, close to 100 billion devices will be connected to the internet. One of the most prominent and perhaps the main driver for the internet of thing is sensor data computing. Sensor data computing is a cloud-based utility that is provisioned as SaaS, IaaS, and PaaS. This paper addresses the main concerns that relate to the internet of things and sensor data computing. Focus is placed on how the two entities work together to deliver seamless services. Challenges and applicability of the two techniques are also discussed
. INTRODUCTION
The concept of the internet of things is gaining worldwide recognition whereby millions of devices are now being connected to each other using the internet. The nature and range of products and services create the opportunities for the establishment of an infrastructure that supports the IoT technology. An important element that characterizes IoT concept is that there is no human interaction involved in the communication processes. Devices communicate with each other through a communication channel and sensor networks. Every device is identified by a unique IP address. Devices also communicate with each other using machine-to-machine communication because they portray what is known as smart characteristics [1]. The Figure 1 depicts the idea of the Internet of Things.
The communication infrastructure in an IoT environment is enhanced using sensors attached to every device. Each sensor node is connected to each other using a wireless sensor network. A sensor node is used to detect various parameters such as temperature, pressure, motion, and light among others [2]. The number of sensor-enabled devices is estimated to exceed 24 million by the end of 2016. The majority of these applications will be found in industrial sector especially in power transmission, natural gas, agriculture, transport, and manufacturing plants [2].
The operation and functionality of sensor networks work in collaboration with the use of sensor data that is normally stored in servers. In this case, sensor data encompass the data obtained from sensor nodes and transmitted through the wireless sensor network [3]. These devices use information, services, and architecture that is stored in cloud servers.
Consequently, the majority of sensor data entities are supplied as services either through an on-demand or subscription basis. The concept of sensor data computing as a service in IoT forms the bulk of discussion in this manuscript.
LITERATURE REVIEW
Zaslavsky offers the view that the internet of things is the foundational pillar for the future of the internet [4]. IoT is a concept that involves the interconnection of various devices using a network channel preferably the internet. Notable devices include smartphones, handheld devices, cameras, sensors, and TVs among others [5]. IoT technology is characterized by the connectivity of devices through the Internet with the help of sensors [6]. The concept of IoT implies that almost every device is required to have a sensor to provide room for efficient transmission of data signals.
The basic architecture that characterizes IoT technology is made possible by the utilization of cloud computing methodology. This implies that majority of services and architectures are availed through the cloud [6]. In IoT, sensor data computing is normally provisioned as a cloud service with the aim of providing access to sensor data among various consumers [7]. Sensor data as a service is founded on the same working principles such as those employed by SaaS (software-as-a-service), IaaS (Infrastructure-as-a-service), and PaaS (platform-as-a-service) [7]. Sensor data computing creates an enabling environment for the IoT infrastructure. Sensor data computing and IoT technology have found use in a variety of application areas such as meteorology, agriculture, astronomy, manufacturing, transport, communication, environmental science, and other related industrial processes [4].
Billions of objects are capable of sensing and communicating from anywhere are connecting online, where these objects are connected to a network so that data can be shared. There are more connected objects than people on the planet. Smartphones are just the inception of the tech era.
We will carry sensors in future that will measure our health and how we move around the environment. These sensors will help us to navigate and socialize the world in ways that we can barely imagine. The society shall have a greater impact than the first digital revolution.
However, as with any new technology, there are possibilities for significant challenges too. In the case of the Internet of Things, breaches of security and privacy have the greatest potential for causing harm. Despite these challenges, sensory data computing continues to be an important service that enhances service delivery in an IoT environment.
III. SENSOR DATA COMPUTING ISSUES IN INTERNET OF THINGS
Sensor data computing in an IoT environment incorporates various issues’ that determine the level and quality of services provided. An important element associated with IoT and sensor networks is the ability of handle large volumes of data [8]. Data management in cloud computing environment is a sophisticated task especially considering that clients need to update and access the data on a regular basis. In an IoT environment, data should b available seamlessly otherwise service delivery will be affected. This can be catastrophic for critical services such as in transport and the health sector.
The data generated by sensor devices portrays multiple characteristics, which makes it difficult to manage. Even though most of the data represents time series constraints, it is difficult to manage any form of manipulations that apply to this data [9]. The quality of data generated by sensor nodes is also an issue of concern. In most cases the data generated has a lot of noise thus making it difficult to represent accurate measurements [9]. It is also difficult to manage data interoperability, especially if the data comes from different architectures.
A sensory network is designed in such a way that it is supposed to exhibit self-organizing features. In other words, a WSN is supposed to resume normal operations even after reconfiguration of major settings. Also, a WSN is supposed to establish automatic communication with the WSNs in the surrounding environment. The problem is that not many WSN harbors the ability to self-organize. This not only affects service delivery but also the security of the WSN [8].
IV. CHALLENGES IN IOT AND SENSOR DATA COMPUTING
Sensor data computing relies on the efficiency of the underlying cloud-computing model. As such, one of the major challenges is associated with security and privacy both for data and for the underlying architecture [8]. Sensor nodes and sensor networks are placed in an uncontrollable environment where there is little or no surveillance. Consequently, there is no guarantee as to whether these devices can be stolen, destroyed, or reconfigured. Another security challenge emanates from the fact that sensor data is normally transmitted using wireless networks. The security and privacy of the transmission medium are subject to malicious attacks such as Denial of Service and Sybil Attack [8].
Sensory data computing is a data-driven service that depends on the reliability and efficiency of both data and the transmission medium. [8] States that data organization is an issue that continues to affect sensor data computing. Data availability cannot always be guaranteed. Also, data confidentiality and authentication are also a subject of concern because some WSNs are not secure enough to guarantee it. This also affects the integrity of the data stored in cloud servers thereby making it difficult to trust the efficiency of the WSN.
Sensor networks rely on an efficient power supply to deliver core functionality. However, it is sometimes difficult to guarantee a reliable source of power supply. Power interruption is often a challenge that affects service delivery. Also, the majority of modern sensor networks rely on optimized power sources, which is difficult to attain [8].
Another important challenge in sensor data computing is related with the ability to handle real-time operations. Critical operations such as those in the military and the health sector require the use of real-time data. Many constraints hinder the delivery and processing of real-time data. Some wireless sensor networks do not have the capacity to handle real-time execution, which culminates in disruption of service delivery and quality assurance [8].
V. APPLICATIONS OF SENSOR DATA COMPUTING
A. Transport
High-performance computing, data collection, analytics and open data are powering progress in transportation. Consumers can book services and plan trips using various range of applications via mobile computing. The internet of things could significantly improve the way we travel by connecting up the different strands of multimode journeys.
B. Healthcare
The Internet of Things will help to shift from cure to prevention and give people greater control over decisions affecting their wellbeing. In turn, this could integrate the delivery of care, improve clinical outcomes and yield considerable cost efficiencies for the NHS.
Telehealth is the delivery of remote health related services is increasingly feasible, leading to the rise of connected smart devices.
C. Smart Gas Meters
The £10.9 billion smart meters program is the biggest government investment to date in Internet of Things technologies. By 2020, 53 million electricity and gas smart meters in homes and small businesses.
D. Agriculture
As a field of human activity that was one of the very first to be transformed by technology, agriculture is well versed in taking the benefits of new opportunities. The Internet of Things potentially offers another leap forward.
Field-based sensors can already measure soil moisture and communicate with weather stations for the latest forecasts. This data is used by large farming operations to determine how much water to apply to crops and when to apply it. Other sensors can collect data on temperature, light, soil acidity and fertilizer content.
Animal tracking allows livestock to be monitored for disease and accidents – as well as providing opportunities for better husbandry.
VI. CONCLUSION
The concept of Internet of Things is an important aspect that allows the interconnection of numerous devices using the internet. It has revolutionized the way of doing things and introduced the smart city concept in the modern world. The efficiency of IoT concept is made possible by the applicability of sensory data computing. This is an important aspect because sensory data obtained from wireless sensor networks is used to execute various functions within the smart city environment. Sensory data computing is thus treated as service whereby it is involved with the generation, transmission, and processing of sensory data. Because these services are mostly delivered in a cloud-computing environment, various issues of data management emerge. Data availability, authentication, aggregation, and integrity are major issues in sensory data computing. Also, power management and the aspect of self-organization among WSNs are other issues that affect service delivery.
Even though sensory data computing is a core service in IoT, the issue of security continues to be a major challenge. The security and privacy of data and the underlying transmission medium affects service delivery. Power supply and the autonomous nature of WSNs are other notable challenges eminent in sensory data computing. Finally, the ability to handle and process real time data is a challenge that affects service delivery and quality assurance in crucial operations like health, transport, and military.
Despite these challenges, sensory data computing continues to be an important service that enhances service delivery in an IoT environment. Proper management and efficient design of WSNs should be encouraged in an attempt to address these challenges. Also, sensory data computing should be structured in such a way that it adopts the use of modern technology for service delivery.