Distributed Signal Processing for Distributed Sensor Networks

Lead Research Organisation: University of Edinburgh
Department Name: Digital Communications

Abstract

The world is moving towards an environment of pervasive interconnected sensing. This is driven by the falling cost of hardware and significant advances in communications networks. In the defence sector many companies have realized the significance of this network enabled capability and have instigated their own programmes. Ideally the signal processing algorithms for such networks should be: computationally efficient - to minimize power and/or cooling requirements; communications aware - to minimize the costs of communications between sensors;non-hierarchical - so the network is robust to the removal and addition of sensors; globally convergent - so solutions are as accurate as those that would be obtained if each node had direct access to all the information across the network.In this proposal we offer a truly distributed signal processing solution to the distributed sensor problem of source localization. This is in contrast to the recent body of work on distributed adaptive processing algorithms which offer only parallel processing solutions to single input/ single output temporal filltering problems. Distributed sensor networks, by design and intent, provide a spatial processing capability, enabling source location. The distributed signal processing algorithms we seek facilitate this spatial processing capability. While our starting point is framed in terms of parallel processing adaptive filtering techniques, the theoretical framework we develop facilitates extensions into spatial signal processing and source location.

Publications

10 25 50
publication icon
Abd-Elrady E (2013) Filtering approaches to accelerated consensus in diffusion sensor networks in International Journal of Communication Systems

publication icon
Cavalcante R (2010) Adaptive Filter Algorithms for Accelerated Discrete-Time Consensus in IEEE Transactions on Signal Processing

publication icon
Emad Abd-Elrady (2010) Direct Learning Architectures for Digital Predistortion of Nonlinear Volterra Systems in Sensor Signal Processing for Defence 2010 (SSPD 2010)

publication icon
Wan, S (2010) Array shape self-calibration using particle swarm optimization and decaying diagonal loading in Sensor Signal Processing for Defence 2010 (SSPD 2010)

 
Description In this study we have developed distributed signal processing solutions to the distributed sensor problem of acoustic source localization. In many such distributed systems, the objective is to reach agreement on values acquired by the nodes in a network. A common approach to solving such problems is the iterative, weighted linear combination of the neighbouring values to which each node has access. Methods to compute appropriate weights have been extensively studied, but the resulting iterative algorithms still require many iterations to provide a fairly good estimate of the consensus value. In this project we have shown that a good estimate of the consensus value can be obtained within a few iterations of conventional consensus algorithms by filtering the output of each node with an adaptive filter, i.e. a filter that learns from the data. This appears to be a new application for adaptive filters. The resultant algorithms do not require knowledge of the network topology and can handle networks that change with time.
Exploitation Route In the design of sensor networks, particularly acoustic ones.
Sectors Digital/Communication/Information Technologies (including Software)

Electronics

Environment

 
Description University of Edinburgh
Amount £40,000 (GBP)
Funding ID Institute for Digital Communications PhD Studentship 
Organisation University of Edinburgh 
Sector Academic/University
Country United Kingdom
Start