The pervasive computing is also known as ubiquitous systems which are mainly focused on the context awareness techniques. In fact, the pervasive information devices are distinguished by various data and the interrogation of these systems are passed out at anytime, anywhere and by any type of devices. Pervasive computing devices are mainly depends on the NNN paradigm (aNywhere, aNytime, with aNything) [1]. Most characterizing from conventional computers will be the capability of a digital artifact to communicate to their physical location. Usually computers have been detached from the circumstances, in which they are used, and only recently location and context-awareness has established improved interest. In pervasive computing the location and context are considered as a keywords. The interaction between system-human and systemmultiple system was increased a lot. The latest internet technologies drive a more general interest in contextawareness such as smart phone applications (Google maps, Gmail and yahoo mail). Context-aware applications and services nee`d equipment from the foundation of wireless network, user infrastructure, and middleware. Furthermore, application and technology associated with network, middleware and user communications are based on concept and research activities. In pervasive computing the context aware is an important aspect. Context awareness is belongings to mobile computing or cloud computing. This is used to complementarily for location awareness. While location is determining that how certain process operates in a system, context might be added for more flexible to mobile users (eg. smart phones). Context aware devices may also try to build guess about the user's current situation. Context is any information which can be used to represent the location of an entity. But the context awareness is affected by some factors in terms of prediction,

Ruksana Maniyar1*, Shantesh G. Gollagi2, Sumaiya M. Saleem Kakkeri3 1

addressed in the project we report are the Context Modeling, recognition, reasoning, Quality of Contexts (QoC), unexpected adaptations, Power management, Security and privacy. Context prediction allows a system to turn into truly proactive. For industrial applications, this can involve actual economic cost. Sometimes, in pervasive computing the context aware systems have to manage conflicting situations while collecting data and extracting high level context data information. Hence to reduce this problem, use quality of context parameters [2]. Context aware systems are computer devices that can makes available the relevant services and information to customers by developing context. However everyone needs to protect their personal information. People don’t like to disclose outside who they are and what services they want [3]. Smart phones become as a pervasive computing application, their power consumption remains one of the major prominent issues. In the following survey, there are different type of techniques are recommended and the issues, challenges and performance are analyzed. Web ontology language approach such as Context Broker Architecture (CoBrA) is better for handling the context modeling and reasoning. Context based prediction method is recommended for power reduction. To maintain the privacy and security of sensitive context data introduce few authentication policies such as Gaia Authentication, Mist – Privacy Communication, Dynamic Security Policies and Access Control.

There are different types of technologies suggested for handling the context-awareness issues and met few solutions to avoid those problems. In this literature survey few methods and policies are considered how they tackle the problem and decrease the issues in pervasive computing.

Context modeling is described as an enclosing environment of some devices and context is surrounding element for the devices [4]. Context modeling is following an ontology based approaches. Context recognition is identifying the situation of user. This includes what and how the situations are recognized from the context. And reasoning is managing the uncertainty situations. Ontology web language is used to provide semantic based context for particular users. Context modeling framework is focused to provide clarity in information and representation, meaning and usage. It takes into description the common behavior of context information such as its temporal nature, uncertainty, impreciseness, incompleteness and privacy. Also context model should address special requirements of pervasive computing environments like distribution, mobility, heterogeneity of context devices. Hence, have to improve the support for automatic context reasoning. Some of the existing approach of context modeling is Key-Value Models, Markup Scheme Models, Graphical Models, Object Oriented Models, Logic Based and Ontology Based Models. All above approaches are based on the ontology concept (CoBrA). This system makes available a set of ontological idea to describe entities such as people, places or several other kinds of substance within their contexts. The CoBrA method uses broker-centric agent architecture to provide runtime support for context-aware systems, particularly in Intelligent Meeting Rooms, a prevalent scenario of a pervasive computing environment. However these methods are not applicable for heterogeneous pervasive computing environment. To overcome this problem used MUSIC model which contains three layers are conceptual, exchange and functional layer. Model Driven Development is used in conceptual layer which is used to enable the description of context such as objects, scope and illustration. Also it is depends on the standard specification language like UML and OWL (Web Ontology Language) [5]. The method makes easy the context reasoning by given that structure for contexts, rules and their semantics. In context management system using context data or element, the semantic concept have to illustrate. Context awareness performs reasoning about the context and passes decisions about the actions to be triggered. Ontology based approaches are used to support the context reasoning efficiently [6][7]. The goal of exchange layer is to be utilized for interoperability between computing systems. In this layer the context information is represented as a sufficient depiction like XML. Functional layer is used for actual implementation of context model demonstration and the interior mechanism which will be used for various modes. With the concept of ontology based approach, to establish a common vocabulary of concepts and to explicitly address heterogeneous representations of context information in pervasive computing. However it makes a serious problem with devices in case of limited sources and arising a problem from plugging mechanism.

Ruksana Maniyar1*, Shantesh G. Gollagi2, Sumaiya M. Saleem Kakkeri3 1

Based on the context recognition we can avoid critical based situation in pervasive computing. Recognize and predict high level context information from low level sensor data is presented. Using a distributed context-aware system with context prediction, we can generate a finegrained dynamic prediction of power consumption, thus improving the accuracy of approximation, and reducing loss through faster, automated reaction [6][8]. However this is failed in case of decentralized architecture. Also context prediction provides mostly informal description of the problem. Also these methods can not able to handle the conflict situation.

In pervasive computing and mobile devices context aware systems are facing lot more conflicting situations during the collection and extraction of high level context information. Quality of context is useful to give solution for uncertain and conflicting situations involving context information. Quality of context parameters is helping to face with these issues. Why this issue arise means, first have to collect raw data then from this data have to extract high level context information. After removing repeated and inconsistent context have to combine and save the context information and finally delivering this context data to specific users. Several conflict situations are occurring due to execution of above tasks. These conflicting conditions powerfully influence the potential of context-aware systems to adapt to the evolving situation in pervasive environments [9]. Even few existing methods were available to resolve with the conflict situations. But those strategies might slow down the process of decision making, divert user, or remove few significant context objects as well [10]. Hence to avoid this drawback new method Quality of Context parameters introduced. It is described as up-to-datedness, trustworthiness, completeness, and significance, are used to specify the quality of context information. So in this method have to analyze conflict situation of context aware system and propose the solving policies depends on QoC parameters. Up-to-datedness pointing the degree of rationalism to exploit a context objects at a specific instance of time. This metric can be helpful in determining the conflict in the context object that alters its value very speedily, e.g., location of a fast moving vehicle. Trustworthiness is the amount of the appropriateness of a sensor to collect the context of a particular type. We have to compute the trustworthiness of a context accuracy of sensor to measure that type of information. This concept is particularly useful in resolving the conflict when we have more than one sensor collecting the context of same entity or event. Completeness of a context object is mostly essential to get the comprehensive picture of the present circumstances of real world. According to this policy decision is made on the basis of that context object which has more complete information about current situation. Significance measures the worth or preciousness of context object. It is particularly important to mention this metric when there is a context object of high critical value. Based on the threshold value, users conflict situation handled by the QoC parameters [11] [12]. However this method also failed to handle the unexpected adoptions in pervasive computing.

Due to unsolved problems, sometimes unexpected situation occurred in pervasive computing. The most challenging issues are managing the loads of users and things, provided that interoperability across the heterogeneous things, and overcoming the unknown dynamic environment due to the mobility of things. Service-Oriented Architecture (SOA) is used to face the problem of unexpected situation. If there is a huge number of a request then complicated coordination occurred among millions of Things and services [13]. In such situation, performing even a simple service detection or composition may go beyond acceptable time, communication costs and resource consumption. To conquer the above challenges, we revisit the SOA itself along with the functionalities of traditional service-oriented middleware. Specifically, we present a Thing-based SOA that relieves service consumers from heavy communications and computations. All tasks in SOA revolve around a business logic that can be satisfied by one or several services. Thing-based SOA and a middleware designed to address the large scale, heterogeneity and unknown environment issues of the mobile IoT. However privacy concept is still missing in this method.

The main issue is dealing with security and privacy factor. Hence we have to improve the security mechanism of context sensing in pervasive computing. To deal with this issues introduce the terms Gaia Authentication, Mist – Privacy Communication, Dynamic Security Policies and Access Control. Authentication mechanisms in pervasive computing environments should strike a balance between

Ruksana Maniyar1*, Shantesh G. Gollagi2, Sumaiya M. Saleem Kakkeri3 1

to authenticate themselves to the system using a variety of means. These include the use of wearable devices, face recognition, smart badges, fingerprint identification, retinal scans, etc. The logic used includes temporal and fuzzy operators that allow the policies to capture context or temporal information, like revocation of authentication credentials under certain circumstances and so on. Because of the various authentication methods and their different strengths, it is sensible to accommodate different levels of confidence and incorporate context [3]. Dynamic security policy contains program modules correspond to the dynamic policy implementation, and can be enforced by executing them in a suitable software context. Security concerns need to be integrated into models of system behavior, and security policies have to form an integral part of system specifications. This research is crucial in the context of active spaces and in other dynamic environments where operational parameters are constantly changing. Dynamic policies enable the creation of customizable programs that can be deployed on-the-fly, to enforce and implement strong security policies that can adapt to a changing software environment. Another mechanism is mist privacy communication which allows certified individual to access services while protecting their location privacy. The following is overview of how Mist works. Mist contains a privacy preserving hierarchy of Mist Routers that figure an overlay network. This network makes easy confidential communication through routing packets via hop-by-hop and handle-based routing protocol. We use public key cryptography in the initial setup of these handles. These methods formulate message impossible to map out by snoopers and untrusted third parties. A handle based protocol is an identifier which is distinctive for single Mist Router. “Incoming handle” is contained by every received packet which is utilized by the Mist Router to recognize the subsequent hop to which to ahead the packet. The arriving handle is exchanged by an outgoing handle earlier than the packet is broadcast to the following hop. And hop-by-hop routing protocol will permit a Mist Router to promote the packet to the next hop, when hiding the original source and final destination. To deal with the new challenges and issues in describing and administering security policies in pervasive computing environments, we recommend an innovative class of policies named dynamic policies. and after the execution of dynamic policies. We consider that security problem need to be incorporated into models of system behavior, and security policies have to form an essential part of system provision. This investigation is critical in the context of active places and in other dynamic surroundings where operational parameters are continuously varying.

In case of mobile devices or computer devices the power consumption is a important issue. Also why energy factor issue arises in pervasive computing means they operate in a dynamically changing context [14]. This is in terms of both user needs as well as computational needs. Also this can be handled in the network communication and CPU processing of smart phones. But it is a bottleneck in terms of wide scale adoption of cloud based services. Because, more number of users needs different kind of information which is carried by several devices. In context aware system, the complexity arises due to the interaction gets increased between context consumers and providers. Hence to design software component for energy devices like those are somewhat difficult. Sometime it leads major impact, due to lack of important information in emergency case. Even more modern energy devices found, researchers also addressed power overhead problem in context aware. For example mobile context broker method reduced the overall power consumption in both users and providers, it is only limited for particular user and particular parameters. So further we can enhance for set of devices as well as complex queries.

Context modeling, reasoning and recognition Comprehensive context modeling, reasoning and context prediction is handled by different approaches and improves the reliability and accuracy of context. Conflict situation Quality of context parameter is used to face the conflict situation. It improves the completeness of context, trustworthy, significance and up to date of context.

Ruksana Maniyar1*, Shantesh G. Gollagi2, Sumaiya M. Saleem Kakkeri3 1

Service-Oriented Architecture is dealt with unexpected situation. It handled the overhead communication problem. Security and Privacy This dealt with issues of security and privacy of users and computing devices. Reduced the issues significantly yet still a problem due to dynamic changes in context. Power management It suggests few solutions but it is not helpful during huge number of users and complex queries.

In the above literature survey major types of issues in pervasive computing are discussed. Issues are in terms of context modeling, reasoning, recognition, conflict situation, unexpected adaptation, security and privacy, and power management. Also some more factors are performance of context aware, algorithms and data structures and reliability issues. Ontology based approaches are helpful with dealing the context modeling and reasoning of context awareness. It provides clear representation of context in pervasive computing [15]. Quality of context parameters describes that which is valuable information for particular application. This improves the performance and effectiveness of context aware systems. SOA can handle the decentralized approach and progress the scalability of context aware. Security and privacy of computing environments is increased by using few authentication schemes. Due to complex computing time and limited energy resources power consumption problem araised. Finally we conclude that still some more new techniques needed to improve and enhance further developments of efficient context awareness in pervasive computing.

Fatma Achour, Anis Jedidi and Faiez Gargouri” The Generic Model for Pervasive Information System” Proc journal of emerging technologies in web intelligence, vol. 4, no. 4, november 2012. Weishan Zhang, Klaus Marius Hansen, and Paolo Bellavista, “A Research Roadmap for Context-Awareness-Based Self-managed Systems,”

275–283, 2013.

Roy Campbell1, Jalal Al-Muhtadi, Prasad Naldurg, Geetanjali Sampemane, M. Dennis Mickunas “Towards Security and Privacy for Pervasive Computing” Proc National Science Foundation, NSF CCR 0086094 ITR and NSF 99-72884 EQ. Roland Reichle1, Michael Wagner, Mohammad Ullah Khan1, Kurt Geihs1, Jorge Lorenzo, Massimo Valla3, Cristina Fra, Nearchos Paspallis, and George A. Papadopoulos “A Comprehensive Context Modeling Framework for Pervasive Computing Systems” Proc IFIP International Federation for Information Processing pp. 281–295, 2008. Dejene Ejigu — Marian Scuturici — Lionel Brunie Laboratoire “An ontology-based approach to context modeling and reasoning in pervasive computing” Proc IEEE conference pp 4-19 ,2007. Horrocks, I.: DAML+OIL: a Reason-able Web Ontology Language. In: Chaudhri, A.B., Unland, R., Djeraba, C., Lindner, W. (eds.) EDBT 2002. LNCS, vol. 2490, Springer, Heidelberg (2002) Charith Perera, Student Member, IEEE, Arkady Zaslavsky, Member, IEEE, Peter Christen, and Dimitrios Georgakopoulos, Member, “Context Aware Computing for The Internet of Things: A Survey” ieee communications surveys & tutorials, vol. x, no. x, xxxx xxxx. Yong Ding, Hedda R. Schmidtke, Michael Beigl “Beyond Context-Awareness: Context Prediction in an Industrial Application” Proc ACM UbiComp’10, September 26–29, 2010. Atif Manzoor, Hong-Linh Truong, and Schahram Dustdar “Using Quality of Context to Resolve Conflicts inContext-Aware Systems” Proc FP6-2005-IST-5- 034749. ANIND K. DEY Intel Research, Berkeley and JENNIFER MANKOFF “Designing mediation for contextaware applications” Proc ACM Trans. Comput.-Hum. Interact., 12(1):53–80, 2005. Atif Manzoor, Hong-Linh Truong, and Schahram Dustdar “On the Evaluation of Quality of Context” Proc Springer-Verlag Berlin Heidelberg pp. 140–153, 2008.

Ruksana Maniyar1*, Shantesh G. Gollagi2, Sumaiya M. Saleem Kakkeri3 1

for groupaware ubiquitous computing environments. In COMPSAC ’05: Proceedings of the 29th Annual International Computer Software and Applications Conference (COMPSAC’05) Volume 1, pages 359–364. IEEE Computer Society, 2005. Thiago Teixeira, Sara Hachem, Val´erie Issarny, and Nikolaos Georgantas “Service Oriented Middleware for the Internet of Things” Proc Springer-Verlag Berlin Heidelberg pp. 220–229, 2011. Mohammad Moghimi, Jagannathan Venkatesh, Piero Zappi and Tajana Rosing “Context-Aware Mobile Power Management Using Fuzzy Inference as a Service” Proc Institute of Computer Science, Social Informatics and Telecommunications Engineering pp. 314-327, 2013. Harry Chen, Tim Finin, and Anupam Joshi Department of Computer Science and Electrical Engineering University of Maryland Baltimore County” An Ontology for Context-Aware Pervasive Computing Environ

Hirasugar Institute of Technoloyg Nidsasoshi