Description
Objective: Develop an auto-focus signal processing capability to optimize detection of quiet contacts by arrays of hydrophones. Description: Arrays of hydrophones are used to detect, classify, and localize contacts in the ocean environment. Finding a contact, especially a quiet contact, is extremely challenging due to the large volume of data that needs to be searched as well as the large number of other noise sources (e.g., shipping, fishing, whales, etc.) that generate clutter on the displays. Array signal processing, also known as beamforming, steers many beams to spatially filter the noise environment and generate a 3-D data volume that is a function of time, frequency, and bearing (i.e., steered beam) that are processed to generate several detection surfaces. Several parameters can be adjusted to optimize the detection of a signal on an array. One of these parameters is focus range. (Other parameters are more sensitive and will be provided to Phase II awardees). However, only a limited number of display surfaces are typically generated due to processing constraints, and this may not provide the best opportunity to detect all signals. Furthermore, the operators typically have a large workload and are only able to search for a limited number of the available display surfaces. Automation approaches have been developed for decades to help reduce operator workload. However, a well-trained operator can still detect lower Signal to Noise Ratio (SNR) signals than the state-of-the-art automation. The main reason for this is if the automation detection threshold is adjusted to detect lower SNR signals, it will cause an increase in the number of false alerts that detracts from the search process. Another approach that is used to reduce the operator workload is ORing, which combines multiple Passive Narrow Band (PNB) displays by taking the maximum value at each time/frequency bin and then combines all contacts found on any of the displays onto a single display; however, it also takes the maximum of the noise bins. This results in ORing loss by increasing the noise floor and reducing the overall SNR. As a result, automation has not yet solved the operator workload problem and operators are still required to conduct manual search on a limited number of detection surfaces. This leads to system losses that can at times be significant and offers an opportunity to mitigate those losses with a new processing paradigm. The objective of this SBIR topic is to develop a signal processing approach that will auto-focus on the signal processing (much like a digital camera does) with respect to parameters such as focus range. There is currently nothing available commercially. The easiest example to understand is range focusing. Let’s assume we are trying to track whales and there are several of them at different ranges. If we process a single far field (i.e., distant) focus range, then the close-range whales may barely be detected. Instead, if we process several focus ranges, let’s say 10, from close to far, there will be one focus range where each of the whales displays the clearest signal with the highest SNR. Over time, the whales will swim closer and farther, and the best detection range will change. The problem is that the operator doesn’t have time to look at the detection surfaces for all 10 focus ranges so instead we need to combine them into a single display that contains the higher SNR instance of each whale regardless of the range where they are. Different whales will also have different broadband signatures and would be more detectable when averaging over different frequency bands. The optimal frequency band may also vary as the ambient noise environment (such as nearby shipping and weather conditions) changes. If the processing generates a large number of detections in multiple frequency bands, then a user will be able to find the most detectable instance of each whale over time. Processing multiple focus ranges is relatively straightforward and is largely just brute force processing. The innovative part of this SBIR topic is the use of this larger data volume to build a combined display that contains the best representation of every available signal. This combined display would be the primary search space for the operators and would also be provided with other automation algorithms. One of the keys to success will be developing an alternative to standard ORing that takes the maximum value at each pixel across the beams being ORed. It is speculated that improvements are possible since the SNR of the signals will be well behaved across the ORing dimension. For example, if multiple focus ranges are combined, there will be one focus range where the signal is strongest, but the signal will gradually degrade as the difference between the focus range and the actual range increases. For pixels that contain noise instead of signal, it is expected that the levels will be more random and that this could be exploited to enhance the signal without increasing the background noise. Overall, it is expected that this auto focus approach will allow system gains that are currently not being realized with the current signal processing and automation approach. This would significantly improve system performance by providing earlier detections and longer holding times of contact without increasing the operator workload or requiring a complete overhaul of the signal processing and automation framework. And although this does come at an increased computational cost, it would allow us to squeeze every dB out of the signal processing. Work produced in Phase II may become classified. Note: The prospective contractor(s) must be U.S. owned and operated with no foreign influence as defined by 32 U.S.C. § 2004.20 et seq., National Industrial Security Program Executive Agent and Operating Manual, unless acceptable mitigating procedures can and have been implemented and approved by the Defense Counterintelligence and Security Agency (DCSA) formerly Defense Security Service (DSS). The selected contractor must be able to acquire and maintain a secret level facility and Personnel Security Clearances. This will allow contractor personnel to perform on advanced phases of this project as set forth by DCSA and NAVSEA in order to gain access to classified information pertaining to the national defense of the United States and its allies; this will be an inherent requirement. The selected company will be required to safeguard classified material during the advanced phases of this contract IAW the National Industrial Security Program Operating Manual (NISPOM), which can be found at Title 32, Part 2004.20 of the Code of Federal Regulations. Keywords: SONAR; Array Signal Processing; Surveillance Automation; Acoustic Detection; Acoustic Target Classification; Sonar Operator Workload CMMC Level: Level 2 (Self)