Stream Computing Framework for Outage Detection ...
Recommend Documents
which is when one sees with the mind's eye. Likewise, the motor network produces movement and it is active during imagined movements. However, although ...
HTC Desire. Motorola Atrix. Samsung Galaxy Tab. (a). 1. 2. 3. 4. 5. 30. 40. 50. 60. 70. 80 .... calculates how much two time series, x and y, can be 'ex- plained' by ...
stress monitoring of an individual via his/her mobile device .... ture, and respiratory data among others) and nearby rugged .... smartphones via Bluetooth.
services according to a pay-per-use business model [1]. In general, Cloud service .... Cloud applications that offer specific security features at application level. A specific .... what are the effects on the monitored subsystem. It is responsible t
tool for data mining and statistical analysis with the intent that researchers will ...
implementation of stream, an R package that provides an intuitive interface for ...
Master of Engineering in Computer Science and Engineering at the ... The author hereby grants to MIT permission to reproduce and distribute publicly paper and ...
4.1 Cloud Based Developed System/Service .... IBM's M2 Project Taps Hadoop for Massive Mashups.www.eweek.com, February 2010. [7] ... %20reuse.pdf. [19].
by Timothy J. Knight[15]. If functional ..... [3] Matan Erez, Jung Ho Ahn, Ankit Gang, William J. Dally. ... [10] Ujval J. Kapasi, William J. Dally, Scott Rixner, John D.
Jan 14, 2016 - (16 lb/ft-s; at least 10% of the investigated sites experienced substantial channel ...... Bosley, and David Levinson is also greatly appreciated.
In this dissertation, we explore the concept of stream computing with GPUs. ......
Programmable graphics hardware is an example of a data parallel architecture[?]
...
Department of Electrical and Computer Engineering,. The University of Texas .... communicate with a BS x in the i-th tier only if its downlink. SINR with respect to ...
to gather large scale minimize drive testing (MDT) reports data, and detects outage by applying machine learning and anomaly detection techniques. To improve ...
Sep 25, 2012 - Firstly, high-dimensional KPI databases of network measurements ... cells are degradation in handover activity, low call setup rates and low cell loading. ..... The 7 center sites are normal cells where the UEs are placed to ...
Apr 3, 2018 - management functions of the entire network and improve the ... and maintenance system alarm, which make that self-healing function unable to ...
May 23, 2013 ... The Mobile Computing Decision Framework (MCDF) provides a holistic decision-
making process that assists organizations in determining ...
Jun 27, 2012 - information (CSI) error, and the asymptotic behavior of the outage ... decreases exponentially with respect to the quality of CSI measured by the ...
enable researchers to more easily build volunteer computing projects. ..... website, we made a qualitative comparison of SLINC to. BOINC. .... The Project Builder is a graphical tool that ... free software and is released under the GNU General.
Humanities. Shaqra University. Saudi Arabia [email protected]. Anthony Atkins. 2. 2. School of Computing. Staffordshire University. Stafford, UK.
Computing Language (OpenCL), the developer still needs to manage the ...... It was initially developed by Apple, and then submitted to Khronos Group to develop its .... account the data transfer overhead when making scheduling decisions.
10000. SSSP-s. SSSP-m. SSSP-l tim e (s). MapReduce. iMapReduce. 23.2%. 37.0%. 38.6% ..... [9] J. Leskovec, K. J. Lang, A. Dasgupta, and M. W. Mahoney,.
stored in safe storage to maintain security and privacy. In. HTCC, cloud storages are secured by SEDs full-disk encryption hardware that moves the processing ...
A Design Framework For Pervasive Computing Systems ... An initial framework. ...... Thank you for spending so much time with me and being the best supervisor ..... with some user customisation, much as current desktop software applications.
application deployment, shorter time to market, and lower cost of operation. ... component based development named Cloud Computing Reusability Model ... allows the reuse of the know-how hidden inside already implemented ... What if the ... the import
Dec 18, 2010 - The network service is more abstract and less clearly .... service topology service yes no network. Fig. 1. The organigram or flow chart of the ...
Stream Computing Framework for Outage Detection ...
grow to the level of big data by connecting millions of consumers to the smart grid. .... A. Martin et.al pointed out that data processing technology used in smart ...
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) < detection of outages. IEC 62056 DLMS/COSEM is a standard defined for smart meters and is widely adopted across the globe including India. DLMS/COSEM meters represent data and events using OBIS codes. These can be used for electricity, gas and water metering. The OBject Identification System (OBIS) defines the identification codes for commonly used data items in metering equipment. This is a six field code. OBIS codes identify data items used in energy metering equipment, in a hierarchical structure using six value groups A to F[][] . Event status word with OBIS code 0.0.94.91.18.255 is used to report the latest event. In case of event ids the odd number is for occurrence and even number is for restoration unless specified differently. Event ID 1, 3, 5 represent voltage missing for the phases R,Y,B respectively and 2, 4, 6 for restoration[] . A. Design Apache Storm is a widely used platform to design stream computing applications. Spouts and Bolts are the building blocks of the program. A Spout is a generator of data which can filter and transform the tuples received from other process or files and send to appropriate bolts. Bolts implement the business logic of the application and can pass the tuples to next bolts. The interconnection of Spouts and Bolts form an acyclic graph in the program and is termed as topology . The number of Spouts and Bolts of different kind determines the speed up of the program and can be configured to utilize maximum parallelism available in the machine or cluster. Here we designed a topology to filter outage events based on OBIS codes and update the customer status on processing each tuple. Topology consists of one Spout which will read events stored on a file and send to all parallel Bolts which is designed to filter outage events. Next level Bolts are grouped by meter ID and change the status of the corresponding customer. Sketch of the topology is shown in the figure 2. B. Implementation Topology is implemented using Apache Storm. Eclipse IDE was used to code the program and make a jar file of the topology. Tuple structure used is (meterID, time, OBISCODE, value). Zookeeper was installed and started before running the topology which act as a job scheduler. Also Storm nimbus, supervisor and UI are need to start to run the topology. Topology tested on a Ubuntu 14.04 OS loaded Intel Xeon 3.70 GHz processor machine with 8 cores. System has 16 GB RAM and 300 GB hard disk.
web browser to get the status of the topology running on a Storm cluster. Main section displayed on a Storm UI are cluster summary, topology summary and supervisor summary. Topology summary gives details about the names of topologies running on a cluster. Under each topology Spouts and Bolts on the topology with a detailed list of statistics are displayed. Storm UI can also give a visual representation of the topology []. Screen shot of the running topology is shown in fig 3.
Fig. 3 Topology from UI
IV. EMPIRICAL RESULTS We tested the outage detection topology with 2 lakhs events stored on a file. Total processing time taken is 32 seconds as reported by Storm UI by using 2 cores of the system. This shows that 160 µs was taken to process each event and able to achieve a speed of 6250 messages per second. It takes more than 2.5 minutes to process 1 million messages which is generated in one second. One Gbps Ethernet cards are available now and can handle 1 GB data per second which will be generated under a feeder . The status of the meter is saved to a file and displayed using web server. The screen shot of the OMS is shown in figure 4. Topology status from the UI is shown in figure 5.
Fig 4 Meter Status
Fig.2 Outage Detection Topology Storm UI is a user interface which can be displayed on a
3
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT)
