Partner Organizations:

Other collaborators:

We have established a collaboration with IBM T.J. Watson
Research (Dr. Xiaohui Gu and Dr. Dimitrios Pendarakis).
Through these collaborations we expect to have impact on 
real-life applications of the techniques we propose.

Activities and findings:

Research and Education Activities: 

The purpose of this research is to develop novel techniques to provide
in-network storage capabilities for sensor network architectures.
At the hardware level, our goal is to provide support, in the current
sensor architectures, for flash card memory. 
At the software level, our goal is to implement software components
that allows us to access and index the data stored in the flash. 
These components have to address the special characteristics 
and constraints of the flash memories. Individual sensors, 
equipped with external memory (such as flash memory), provide 
local storage and also contribute to building distributed storage
across the sensor system. Our goal is to build a large-scale storage
capability in the sensor network by combining the storage capabilities
of the flash cards of the individual sensors.
The broader impact of this work will be storage systems that
are efficient and easy to manage. The outcome of the proposed
work is expected to have significant impact on several disciplines
including, industry manufacturing, environment monitoring,
seismic and structural monitoring.
The educational component of the project seeks to promote early
student involvement in research and teaching, to develop a strong
curriculum, and to increase education awareness in sensor networks.

Findings:

We have developed new external memory index structures for the
efficient storage and retrieval of records stored on the local flash
memory of flash-based sensor network devices (TOS 2006), techniques
for compressing historical multidimensional sensor data (IJSN 2006)
and distributed techniques for reliable coverage restoration 
of sensor fields (IPDPS 2007). We have also presented new 
real-time traffic management techniques to support real-time 
flows in sensor networks (RTSS 2006), techniques for distributed
real-time detection and tracking of homogeneous regions (RTSS 2006)
and dynamic rate allocation techniques for distributed stream
processing systems to enable stream processing on data generated 
by sensors or other data sources (IPDPS 2007).
We have also developed techniques for internet-scale content search
and retrieval (TPDS 2007), improving the execution of large-scale
programs (IS, to appear) and approximate query processing in
peer-to-peer networks (TKDE 2007).

We have developed techniques for reliable real-time data collection
in sensor networks (GIS 2008). We have developed a distributed
cluster-based framework for congestion control and for supporting
multiple classes of traffic in sensor networks (TOSN 2008).
We have also presented new techniques for facilitating
congestion avoidance in sensor networks under the presence
of a mobile sink (RTSS 2007).
We have presented new techniques for predicting and alleviating
hot-spots in distributed stream processing systems (DSN 2008)
and for migrating components to different nodes in response
to overloads (DBISP2P 2008). We have also developed techniques
for dependable and QoS-aware distributed stream processing (ISORC 2008)
and replica placement techniques to provide high availability
to distributed stream processing applications (DEBS 2008).
We have also developed resource management mechanisms using
a multiple feedback loop structure in soft real-time distributed
object systems (JSS 2008).

Training and Development:

The project supports 4 graduate students. Song Lin earned his PhD
degree in 2007 and has joined Yahoo. Kyriakos Karenos earned 
his PhD degree in 2008 and has joined IBM T.J. Watson Research. 
Thomas Repantis is expected to earn his PhD degree in September 2008.
Yannis Drougas is expected to complete his PhD in the 2008-2009
academic year.
The PI has also advised four undergraduate students, Bilson Campana, Ricardo
Perez, Mike Ballou and Roberto Rivas. 

Outreach Activities:

The PI was invited as a keynote speaker at the Databases, Information
Systems and Peer-to-Peer Computing Workshop (DBISP2P) in September 2006.

The PI received the best paper award at the SAINT 2008 conference
for the paper 'Dynamic relay node placement in wireless sensor networks'.

Journal Publications:

D. Zeinalipour-Yazti, V. Kalogeraki, D. Gunopulos, "pFusion: A P2P Architecture for Internet-Scale Content-Based Search and Retrieval", IEEE Transactions on Parallel and Distributed Systems (IEEE TPDS), vol. 18, (2007), p. 804., " " Published

Benjamin Arai, Gautam Das, Dimitrios Gunopulos, Vana Kalogeraki, "Efficient Approximate Query Processing in Peer-to-Peer Networks", IEEE Transactions on Knowledge and Data Engineering (IEEE TKDE), vol. 17, (2007), p. 919., " " Published

Sharmila Subramaniam, Vana Kalogeraki, Dimitrios Gunopulos, Fabio Casati, Umeshwar Dayal, Mehmet Sayal, Malu Castellanos, "Improving Business Process Models by Discovering Decision Points", Information Systems Journal, vol. 32, (2007), p. ., " " Published

Song Lin, Demetris Zeinalipour-Yazti, Vana Kalogeraki, Dimitrios Gunopulos, Walid Najjar, "Efficient Indexing Data Structures for Flash-Based Sensor Devices", ACM Transactions on Storage (ACM TOS), vol. 2, (2006), p. 468., " " Published

Song Lin, Vana Kalogeraki, Dimitrios Gunopulos, Stefano Lonardi, "Efficient Information Compression in Sensor Networks", International Journal of Sensor Networks, special issue on Wireless Ad Hoc and Sensor Networks, vol. 1, (2006), p. 229., " " Published

Yannis Drougas and Vana Kalogeraki, "RASC: Dynamic Rate Allocation for Distributed Stream Processing Applications", IEEE International Parallel & Distributed Processing Symposium (IPDPS 2007), vol. , (2007), p. 109., " " Published

Yannis Drougas and Vana Kalogeraki, "Distributed, Reliable Restoration Techniques using Wireless Sensor Devices", IEEE International Parallel & Distributed Processing Symposium (IPDPS 2007), vol. , (2007), p. 72., " " Published

Kyriakos Karenos and Vana Kalogeraki, "Real-time Traffic Management in Sensor Networks", 27th IEEE Real-Time Systems Symposium (RTSS 2006), vol. , (2006), p. 422., " " Published

Sharmila Subramaniam, Vana Kalogeraki and Themis Palpanas, "Distributed Real-Time Detection and Tracking of Homogeneous Regions in Sensor Networks", 27th IEEE Real-Time Systems Symposium (RTSS 2006), vol. , (2006), p. 401., " " Published

K. Karenos, V. Kalogeraki, S. Krishnamurthy, "Cluster-based Congestion Control for Sensor Networks", ACM Transactions on Sensor Networks (TOSN), vol. 4, (2008), p. ., " " Published

V. Kalogeraki, P. M. Melliar-Smith, L. E. Moser, Y. Drougas, "Resource Management Using Multiple Feedback Loops in Soft Real-Time Distributed Object Systems", Journal of Systems and Software, vol. 81, (2008), p. ., " " Published

V. Kalogeraki, A. Soheili, "Realiable Real-Time Data Collection", Encyclopedia of Geographical Information Science, vol. , (2008), p. 204., " " Published

Thomas Repantis and Vana Kalogeraki, "Alleviating Hot-Spots in Peer-to-Peer Stream Processing Environments", Fifth International Workshop on Databases, Information Systems and Peer-to-Peer Computing (DBISP2P 2007), Vienna, Austria, September 2007, vol. , (2007), p. ., " " Published

Kyriakos Karenos and Vana Kalogeraki, "Facilitating Congestion Avoidance in Sensor Networks with a Mobile Sink", 28th IEEE Real-Time Systems Symposium (RTSS 2007), Tucson, Arizona, December 2007, vol. , (2007), p. 3., " " Published

Vana Kalogeraki, Dimitrios Gunopulos, Ravi Sandu, Bhavani Thuraisingham, "QoS Aware Dependable Distributed Stream Processing", 11th IEEE International Symposium on Object- and Component-Oriented Real-Time Distributed Computing (ISORC 2008), Orlando, FL, May 2008, vol. , (2008), p. 69., " " Published

Jorge Mena and Vana Kalogeraki, "Dynamic Relay Node Placement in Wireless Sensor Networks", The 2008 International Symposium on Applications and the Internet (SAINT 2008) Turku, FINLAND, July - Aug 2008, vol. , (2008), p. ., " " Published

Thomas Repantis and Vana Kalogeraki, "Hot-Spot Prediction and Alleviation in Distributed Stream Processing Applications", The 38th Annual IEEE/IFIP International Conference on Dependable Systems and Networks (DSN 2008), Anchorage, Alaska, June 2008, vol. , (2008), p. ., " " Published

Thomas Repantis and Vana Kalogeraki, "Replica Placement for High Availability in Distributed Stream Processing Systems", The 2nd International Conference on Distributed Event-based Systems (DEBS 2008), Rome, Italy, July 2008, vol. , (2008), p. ., " " Published

S. Lin, B. Arai, D. Gunopulos, "Reliable Hierarchical Data Storage in Sensor Networks", 19th Int. Conference on Scientific and Statistical Database Management (SSDBM 2007), Banff, Canada, July 2007, vol. , (2007), p. 26., " " Published

A. J. Dou, V. Kalogeraki, "RG-EDF: An I/O Scheduling Policy for Flash Equipped Sensor Devices", 6th IFIP Workshop on Software Technologies for Future Embedded & Ubiquitous Systems (SEUS 2008), Capri Island, Italy, Oct 2008, vol. , (2008), p. ., " " Accepted

A. J. Dou, S. Lin, V. Kalogeraki, "Real-Time Querying of Historical Data in Flash-equipped Sensor Devices", The 29th IEEE Real-Time Systems Symposium (RTSS 2008), Barcelona, Spain, Nov-Dec 2008, vol. , (2008), p. ., " " Submitted

Book(s) of other one-time publications(s):

Other Specific Products:


Contributions:

Contributions within Discipline:

 The objective of this proposal is to develop new storage techniques
that will enable efficient flash memory support in sensor 
network architectures for long-term storage. Individual sensors,
equipped the flash memories, will provide local storage and also
contribute to building distributed storage across the sensor system.
Our techniques will address the special characteristics and
constraints of the flash memories and the energy constraints 
of the nodes. We will develop new techniques and the required 
data structures for inserting, deleting and indexing data stored
locally on the flash. We will design algorithms for supporting
equality, spatial and multi-dimensional queries in the sensor 
network. We will combine the index structures locally at the 
nodes to implement distributed storage. This will result 
in many Gigabytes of storage distributed across the system.