SUMMARY 1)FOR “COMPREHENSIVE OPTIMIZATION OF DECLARATIVE SENSOR NETWORK QUERIES” To be very precise the research paper deals with the issues of achieving optimization of declarative sensor network queries.As represented by the paper,sensor networks like ad-hoc and wireless networks are energy constrained and because of these restrictions it is very difficult to achieve the aim stated above.The approach or the technique used to solve the problem is adapting the DQP(distributed query processing) i.e TWO PHASE OPTIMIZATION and theories related to it. ANALYSIS Basically the fact that sensor networks have many exceptions have led to many problems when it comes to optimizing queries.Pertaining to this fact its always been one of the most discussed topic in database management system.Initially the concept which somehow provided the solution for this problem is TINYDB.The TINYDB query language is basically a declarative SQL-like query language supporting selection ,join,projection and aggregation.It doesn’t discriminates between different sensors and counts the entire collection as a single relation with different attributes like tempreatutre,pressur etc. and for that for that purpose there is a different sensor.But the problem lies is that optimization is limited to operator reordering and the use of cost models to determine an appropriate acquisition rate.so a new solution was required.So the technical approach used was the implementation of two-phase optimization approach i.e DQP.It is centralized,distributed as well as hybrid and the result were quite convincing.Tthe issues and challenges which arrived while implementing the DQP approach for sensor networks were the issues which were violated in sensor networks.Those issues were related to sensors major ones being that sensors are energy constraint,they act in a way as to preserve energy so as to maintain the network longevity and code is implemented in a way so the higher energy tasks are discarded or delayed.Another issue which comes into highlight is that the communication links are distance sensitive and an altogether a new topology is followed for the transfer of the data and also the need to run sensor nodes according to data-dependent duty cycles.Now about the technical approach which has been deviced to deal with the issues and challenges which have arrived in the paper.firstly there have been a few technical assumptions before implementing the real technique like the sensor networks are self organizing and they tend to identify their neighbours nodes and establish a link between them on their own.They are implemented so as to be risk free and hence can be re-established in case of failure.
Implementing a traditional ABC system is a substantial project requiring a lot of resources. It works well in a smaller setting such as a single department, plant or location; however, it becomes quite cumbersome on a large scale for use on an ongoing basis. Once implemented, ABC is costly to maintain and update since it consists of data concerning many activity measures that are periodically collected, checked and entered into the system.
As the technology developing, everything becomes computable. And when people realizing the importance of the Internet of Things, more and more data is collected. Analyzing such amount of data becomes a big challenge for modern people. As a very important component of our life, internet becomes indispensable. Data sharing between multiple users becomes more popular. It seems our life will stop if without the internet. The user devices becomes much lighter, most computing and data storage are separated with remote operations. Distributed system becomes more and more useful for our life.
problem and it will be used in this case to build the databases. The databases
As defined by Kroenke Database is an integrated, self-describing collection of related data. Data is stored in a uniform way, typically all in one place- for example, a single physical computer. A database maintains a description of the data it contains and the data has some relationship to other data in the databa...
Sensor data, social networking, personal settings, photos, information based on location, online activity, metric usage, etc., can be included in unstructured data from the web. The attempt to store, process, and analyze this entire unstructured data became the cause of development of schema-less alternatives to SQL. Considered together, these alternatives are referred to as NoSQL, denoting “not only SQL”. Whereas, the term NoSQL surrounds a large range of alternatives to relational databases, and the feature that they have in common is that they enable you to treat data more
Large amounts of data streams are generated in resource-constrained environments. Sensor networks represent a typical example. These devices have short life batteries. The design of techniques that are energy efficient is a crucial issue given that sending all the generated stream to a central site is energy inefficient in addition to its lack of scalability problem.
6) Michele Zorzi, “Wireless Sensor and networks: Recent trends and Research Issues’’, University of Padova, Italy, IEEE communication society, Poland, April 28-30, 2008.
The cost model for distributed data mining in [1] gives the apriori estimates of the response time for the given task considering a specific architectural model. The distributed data mining response time T is given as
[7] Elmasri & Navathe. Fundamentals of database systems, 4th edition. Addison-Wesley, Redwood City, CA. 2004.
Many recent advancements which took place in the wireless communications and electronics is the reason which enabled development of cost effective sensors, and those sensors are used for various applications like in household, military and other areas. The networks still have some technical problems which still need to be fixed and are currently being researched.
Sensor Network Topology: The topology of the sensor network must be maintained even with very high node densities.
In the past, most of the databases were centralized, protected, and kept in a one location using a complicated database system known as centralized database. Nowadays, with the new technology of personal computers and cell phones, a new sort of database has appeared, and it seems that majority of people are pleasant with it, even if their private data is split everywhere. Many enterprises had changed their databases from the centralized databases, into the distributed database system, since it meets the demand of accessing and processing the data in the organization. Distributed database technology is considered as one of the most remarkable developments in this century (Ozsu, 1991; Rahimi & Haug, 2010; Cain, 2012). Distributed databases are basically a collection of databases that are divided on multiple computers which are connected logically but located in different physical locations, and each site manages its own local data. In contrast, centralized database is a database that is located in a one location and it is considered as a big single database (Connolly & Begg, 2010).
[8]Y. Tirta, Z. Li, Y. Lu, and S. Bagchi, “Efficient Collection of Sensor Data in Remote Fields Using Mobile Collectors,” Proc. 13th Int’l Conf. Computer Comm. And Networks (ICCN ’04), Oct. 2004.
A Wireless Sensor Network (WSN) consists of hundreds and thousands of the sensor nodes. The sensors either communicate directly to an external base station (BS) or among each other. Larger the number of sensors deployed to sense a large geographical area, greater is its accuracy. Figure 1 shows the components of sensor nodes in its sch...
De Poorter, E., Troubleyn, E., Moerman, I., & Demeester, P. (2011). IDRA: A flexible system architecture for next generation wireless sensor networks. Wireless Networks, 17(6), 1423-1440. doi:http://dx.doi.org/10.1007/s11276-011-0356-5