Lead Investigators:

K. Yao, C.E. Taylor, D. Blumstein, M. Srivastava, and L. Girod

Overview

This portion of the Annual Report covers three separate but related parts of acoustic signal and array processing activities in CENS.  The first part, entitled “Acoustic Array for Source Detection, Localization, and Tracking,” is supported under the CORE program of CENS, with K. Yao as the P.I. and R.E. Hudson as a research participant.  The second part, entitle, “Adaptive Communication in Acoustic Sensor Arrays,” supported by a NSF grant, with C.E. Taylor as the P.I., and E. Stable, K. Yao, M. Cody, and E. Vallejo as co-PIs, and D. Blumstein, L. Girod, and R.E. Hudson as faculty research participants.  The third part, entitled, “Cooperative Acoustic Vehicle Localization,” funded by Toyota, with M. Srivastava as the P.I., with L. Girod and K. Yao as co-PIs.

In the first part, the research goals in the past year have been: 1) to analysis, design, and verify a newly formulated near-optimum 3-dimensional acoustical array for sensor network applications;  2) to implement the previously proven Approximate-Maximum-Likelihood (AML) source detection/localization algorithm with L. Girod’s wireless sensor self-localization scheme and hardware nodes for self-contained animal acoustic source detection and localization; 3) to verify the operations of the wireless sensor platform system described in 2) in order to study bio-complexity issues of birds and marmots in the jungles of Mexico and meadows of Colorado. 

In the second part, the research goals have been: 1) to conduct basic research in learning theory by using distributed sensing, adaptive and communication among animals (e.g., birds and marmots) utilizing field measurements collected by acoustical arrays; 2) to collect and process extensive field data of various birds in the jungles of Mexico and marmots in the meadows of Colorado in order to formulate and verify various bio-complexity conjectures of these animals.

In the third part, the research goals have been: 1) to formulate, implement, and conduct field test of a newly formulated cooperative vehicle detection and localization scheme using an acoustical array system with wireless nodes; 2) to understand and relate the acoustical array system to other modality of sensing (e.g., RF; image; etc.) for localization in realistic situations.

Approach

In the first part: 1) an isotropic 3-D array was formulated to have a uniform performance over all azimuth and elevation angles that asymptotically approaches the optimum Cramer-Rao Bound (CRB) for large signal-to-noise ratio;  limited preliminary field tests verified some of the expected theoretical and simulation results of the array; 2)  the L. Girod platform system with wireless sensor self-localization capability plus the source detection/localization AML algorithm has been implemented and shown to work efficiently in the meadows of Colorado in the Fall of 2006; 3) the AML algorithm implemented on two acoustic arrays using a commercially available Firepod acoustic processing system yielded limited useful field data due to uncertain array locations in the Summer of 2006.

In the second part: 1) Based on the development of a recent successful line of research in learning theory (Angluin, Kanazawa, Vapnik), we are studying distributed sensing, adaptation and communication in two quite different domains.  First, primarily to further our theoretical understanding of how such systems can operate and evolve in simple, controlled settings, we are developing a environment in which distributed computing machines can communicate with each other about network events, with particular attention to irregular activities that could signify intrusions or malfunctions of various kinds. And second, primarily to further our understanding of how such systems work in natural settings, we are beginning the study of a social bird, the Acorn Woodpecker (Melanerpes formicivorus), in its natural habitat. and on several species of antbirds in their native rairforest habitat at Estacion Chajul in Chiapas, Mexico; 2)  In the Fall of 2006,  extensive marmots field data were collected at the Rocky Mountain Research Laboratory in Colorado, using the L. Girod platform system with wireless sensor self-localization capability plus the source detection/localization AML algorithm.  Even under demanding scenarios (with considerable foliage, reverberation, windy conditions, etc.), accurate locations of marmots were found. 

In the third part: 1) In a period of approximately only three months, a full working system of a cooperative vehicular detection/localization system using an acoustical array with wireless nodes was implemented and field tested.  The vehicle locations were found to be quite accurate under several different maneuvering scenarios for speeds of a moving vehicle up to 30-40 mph; 2) Various limiting conditions on the performance of the existing algorithm/system (e.g., length of the PN sequence codes using for ranging; loudness of the acoustic sensing pulses; etc.) were determined.  Qualitative discussions on possible non-acoustic sensing modalities for vehicular detection/localization have been considered. 

Systems/Experiments

In the first and second parts of our research: we have conducted field measurements of the 3-D isotropic and non-isotropic array with the L. Girod wireless nodes (at Rocky Mountain Research Laboratory) and at UCLA Science Court as well as with the wired Firepod nodes at  Chiapas, Mexico and at UCLA Science Court.

In the third part of our research: we have conducted field measurements of the cooperative acoustic vehicular detection/localization system at UCLA Parking Structures 3 and 9. 

Accomplishments

First and second parts of the research: A newly formulated isotropic 3-D acoustic array was proven by analysis, simulation, and from field measurements to have near uniform performance approaching the optimum CR Bound; the combined AML algorithm with the wireless sensor self-localization hardware nodes of L. Girod yielded an effective self-contained easily field-deployable acoustic source detection/localization system suitable for field biologists to collect data in diverse conditions; a basic learning theory based on distributed sensing, adaptation and communication and motivated by field measured data has been formulated.

Third part of the research:  In a period of only third month, a working vehicular detection/location system based on a cooperative acoustical array with wireless sensing nodes was presented to Toyota Motors. 

Future Directions

First part of research: We want to reduce the computational complexity of the isotropic 3-D AML algorithm for real-time applications;  we need to conduct more field measurements to verify the uniformity of the isotropic 3-D array for high elevation angles; we plan to conduct 3-D beamforming field measurements to demonstrate the enhancement of a desired source signal and the attenuation of some other unwanted interfering signals; we want to study the optimum AML algorithm when different sensors encountered non-uniform background noises.

Second part of research:  At Hastings Reserve, with acorn woodpeckers, we plan to explore localization in 3D and to determine if woodpeckers distinguish among the source of calls;  At RMBL we plan to construct a “landscape of fear”, by which we mean determine which animals and where they are is the source of calls;  At Estacion Chajul, Chiapas,  we plan to identify individual antbirds, of 3-4 species, and construct a map of which territories they are found it.;  At UCLA we plan to improve and harden our sensor arrays.

Third part of the research:  After an informal discussion the Toyota sponsor, we have submitted a list of possible topics for future efforts.  The long term interests of Toyota on vehicular detection/localization is unclear at this point.

People

Faculty:

• Kung Yao, Electrical Engineering, UCLA
• Charles Taylor, Ecology and Evolutionary Biology, UCLA
• Martin Cody, Ecology and Evolutionary Biology, UCLA
• Dan Blumstein, Ecology and Evolution Biology, UCLA<
• Edward Stabler, Linguistics, UCLA
• Edwardo Vallego, Monterrey Inst. Technology, Mexico (ITESM);
• Mani Srivastava, Electrical Engineering, UCLA

Researchers:

• Lewis Girod, M.I.T.
  
• R. E. Hudson, UCLA.

Graduate Students:

• Shadnaz Asgari, UCLA
  
• Andreas Ali, UCLA
  
• C.E. Chen, UCLA
  
• C.K. Chen, UCLA
  
• Travis Collier, UCLA

External Research Partnerships

Toyota Motors, (current)