On Big Data Management in the Internet of Things
Chapter One
The Objective of the Research
This work aims to propose and demonstrate a generic, efficient, scalable, and robust approach to Big Data management approach in IoT which extracts real-time value from data and demonstrates its operation in an application area. Using existing and emerging computing paradigms, we seek to develop an approach to significantly reduce latency in streaming data from a network of connected devices and thus capture events that trigger actions in real-time.
CHAPTER TWO
LITERATURE REVIEW
Internet of Things (IoT)
The Internet of Things (IoT) has become a popular topic in both industry and academia. The term was originally coined by the British technologist Kevin Ashton in 1999, to describe a system where the internet is connected to the physical world via ubiquitous sensors [37]. It can simply be described as an interconnection of a massive number of objects/sensors/devices (things) through a communication network to provide value-added services [1]. IoT is a concept and a paradigm that considers pervasive presence in the environment of a variety of things/objects that through a communication medium and unique addressing schemes are able to interact with each other and with other things/objects to create new applications/services and reach common goals [26].
“Things”, in this context, are anything that can connect to any other thing over a communication medium or to the internet. They typically are recognisable and generate and communicate data to one another. IoT’s growth has been rapid over the years as more and more devices are becoming connected and application areas are increasing by the day. With an estimated figure of nearly 50 billion devices to be connected by 2020, and the advances in data handling capabilities, the possibilities in the IoT world are endless.
Why Internet of Things?
The main aim of the IoT concept is to have a smart and interconnected world and to have “things” capture data without help from humans, process the data and make intelligent decisions, all by themselves. This will engineer a new automated world where we can reduce cost and waste [10]. Analysing data from so many sources can give invaluable insights about human behaviour and decision-making patterns, which is very useful in modern marketing and business intelligence. Also, there are environments that are either unsafe or impossible for humans to go into (for example an oil well) and readings need to be taken and communicated to other nodes. IoT fits well into these environments as data can be obtained analysed and instructions can also be communicated to the devices in these environments.
Applications of IoT
Constant events monitoring is one main application area of IoT. An example is smart cement, cement equipped with sensors to monitor stresses, cracks and warpages on a bridge continuously for maintenance purposes [27]. Other applications areas include health, where patients’ movements, heart rate, blood pressure and all other medical readings can be constantly monitored and actions taken whenever there is a need; transport, to monitor vehicular movements, fuel consumption and operational efficiency; and manufacturing, for safety monitoring and maintenance scheduling [9], [27], [28], [29].
CHAPTER THREE
ANALYSIS
Latency
The question of how real is “real time” is often answered by how long getting a response can take before data become obsolete, start depreciating in value or even become useless. Latency refers to the time to get a computation/communication request satisfied. Examples are the time to confirm that a debit card transaction is not fraudulent or the time to recommend another article to a user on a website. Detecting a fraud after the transaction has been completed is of no use, and also taking 50 seconds to confirm that a transaction is not fraudulent will put off most customers. Also, detecting what a visitor to a website might like to read after he/she might have left the website is useless and taking too much time to load a page because of that can also discourage customers from visiting. There are more critical systems like stock trading where milliseconds latency improvements can mean gains in millions of Dollars for companies. Latency requirements of IoT systems are typically strict. Constant monitoring systems especially need to react to actionable events at near zero latency.
CHAPTER FOUR
USE CASES AND IMPLEMENTATION
Introduction
In this chapter, some use case scenarios of the proposed approach are presented. One of the use cases is also implemented. The use case, tools and programming languages used are discussed. The implementation and screen-shots from the latency benchmark readings are also presented.
CHAPTER FIVE
SUMMARY, CONCLUSION AND RECOMMENDATION
Summary
This work focuses on the reducing the time it takes to react to actionable events in IoT applications. It investigates how real-time value can be obtained from the Big Data typically emitted by IoT application. The work presents and extensive literature review of IoT and Big Data, as well as the relationship between the two concepts. It also discussed data streams and existing stream processing frameworks. The work also looked into related work by other authors as well as the existing approaches in reacting to actionable events from IoT applications. This work finally presents a new latency-reducing approach to processing Big Data and actionable events in IoT applications; illustrates use case scenarios and also implements a particular use case.
Conclusion
A new latency-reducing approach of processing Big Data in a way to react to actionable events under strict latency requirements has been proposed in this work. Fog computing paradigm, which has been widely advocated for IoT applications, is leveraged to reduce action latency in IoT applications. The work proposed an “intelli-Fog” layer which caches mined intelligence from Big Data and makes it available for decision making at the Fog layer. This makes intelligent response to actionable events faster and closer to the devices. Immediate decision making and Big Data analytics can also be taking place concurrently in this new approach with just a single data entry point unlike the Lambda architecture.
Recommendations
The proposed approach is still very new and had only been tested on just one of the use cases presented in this work. Future work should look into testing the approach by implementing in use cases and comparing the benchmark with existing approaches. While testing the approach against existing approaches, however, careful attention should be paid to the environments to make user the programs run on the same hardware and using the same communication facilities.
Also this work investigates the architectural part of IoT Big Data management. A second phase of the work should focus on the Big Data processing part. Data ingestion, message queues and stream processing methods there are most suitable for IoT or can be optimized for IoT applications should be investigated.
References
- Perera C., Ranjan R., Wang L., Khan S., and Zomaya y., “Big Data Privacy in the Internet of Things Era” Issue No.03 – May-June (2015 vol.17) pp: 32-39
- Perera, A. Zaslavsky, P. Christen and D. Georgakopoulos, “Context Aware Computing for The Internet of Things: A Survey,” Communications Surveys Tutorials, IEEE, vol. 16, no. 1, pp. 414- 454,2013
- Eaton, D. Deroos, T. Deutsch, G. Lapis and P. Zikopoulos, “Understanding Big Data” , McGraw- Hill Companies,2012.
- Mohak Shah “Big Data and the Internet of Things” Research and Technology Center – North America, Palo Alto,
- Cramer I., “Big Data” Business Technology magazine, 2.2012 Issue
- Baesens, “Analytics in a Big Data World: The Essential Guide to Data Science and Applications.” John Wiley & Sons,2014.
- Ortiz, A.M., Hussein, D. ; Soochang Park ; Han, S.N. ; Crespi, N., “The ClusterBetween Internet of Things and Social Networks: Review and Research Challenges”. IEEE IoT Journal, Vol 1, Issue 3, 2015.
- Khodadadi, F. ; Calheiros, R.N. ; Buyya, R. “A data-centric framework for development and deployment of Internet of Things applications in clouds ” (ISSNIP) 2015 IEEE Tenth International Conference,2015
- MikeBarlow “Real-Time Big Data Analytics: Emerging Architecture” O’Reilly Media, ISBN: 978-1-449-36421-2 2013
