(IJACSA) International Journal of Advanced Computer Science and Applications, Vol. 9, No. 9, 2018 385 | Page www.ijacsa.thesai.org Development of a Novel Approach to Search Resources in IoT Nisar Hussain 1 , Dr. Tayyaba Anees 2 School of System and Technologies University of Management and Technology Lahore, Pakistan AzeemUllah 3 The University of Lahore Pakpattan Campus Pakistan Abstract—Internet of Things (IoT) referred to interconnected the world of things like physical devices, cars, sensors, home appliances, actuators and machines embedded with software at any time, any location. The increasing number of IoT devices facing challenges which are registration, integration, describing sensor, interoperability, semantics, security, discovery and searching. The current systems are suitable for limited number of devices. Our ecosystem change day by day which means we have billions and trillions of devices connecting to the Internet in future. One major challenge in current system is searching of suitable Smart Things from a millions or even billions number of devices in IoT. For the purpose of searching and indexing, some discovery methods and techniques are discussed and compared. Those techniques and methods are studied and find out the limitations and issued of the current system. Another challenge to searching the Smart Things is a variety of description models for describing the Smart Things. In this piece of work, a novel search engine is proposed to search the Smart Things with variety of description models. A web interface is implemented in this research with HTML, JSON and XML formats. The description models of Smart Things SensorML, SensorThings API and W3C JSON-LD are implemented in the current proposed system. Keywords—IoT; IoT resources; search engine for IoT; SensorM I. INTRODUCTION The physical objects have the ability to connect with Internet and share data in respect of Internet of Things. In 2020 the connected devices that are able to communicate with Internet are more than 50.2 billion according to CISCO prediction [1]. These physical objects are called Smart Things. IoT provide a range of connectivity protocols, applications and mechanisms to interoperability with existing infrastructure. In this regard, it„s essential to propose a search engine to search the devices with the context of the user requests. There are some standardized sensors description technologies which are currently used in many systems. These technologies are SensorML [2], SensorThings API[3] and W3C JSON-LD [4]. These technologies used different languages and formats. In this research, we use all these technologies and parse into a common format for supporting our system. Indexer helps us to store and retrieve the data quickly than rational database systems. Indexer also have the abilities to store structure and non-structure data like simple text, relation Database data, PDF format, JSON format, XML format and other structured data. Due to increasing significantly number of devices is a problem of find out a proper device which is fulfills the requirements of the user as well as machine transferable code. Only searching is not enough but also supports reliability and robustness with top ranking resources. It is also difficult as compare to searching traditional web documents because of traditional search on search limited attributes of web documents. Another problem is that there is not a signal standard of describing the devices meta-data. This problem creates another challenge which is lack of describing standard for describing functionality of the devices and resource. A. Problem Statements IoT platform is a scalable network that have huge amount of resources which is needed to be fast search and store. Traditional technique to store the resource description and meta-data using database system has some limitations like storing documents, non-structured data and rank up the search resources. These limitations can be solved by the indexer which can store resource description and provide facility to ranking the using scoring technique. In this research we use different sensor description formats to index descriptions of resources and also ranking up the search results for high performance and reliability. As compare to database systems indexer is more flexible to storing and searching of large data. Only indexer is not enough to search the resources another requirement is user interface which can help user to search resources. B. Objectives In this research different sensor description models are studied and find common descriptions. Different models have different properties and different technologies so we need to parse each model into a shared schema. The parsed description will used further process for a common model to store into indexer. Indexer store description of thing and supply interface to query and sorted its description. After storing the description we need an interface to query the stored information and apply different algorithms to rank the sensors.  This research compares different sensor description models.  Find similar fields between different models and descriptions of Smart Things.