About Converged Sensor Network

Smart Processing
Fusion & Exploitation
CSN Capabilities

Embedded Smart Processing

The Challenge Created by Persistent Surveillance

ISR sensors are surveying ever wider areas with ever greater sophistication. Mounted on long endurance UAVs, these sensors can deliver the Persistent Surveillance necessary to find and fix an elusive, insurgent enemy.

But we know that identifying any specific insurgent activity takes hundreds of hours of Persistent Surveillance. The volume of data collected over that time crushes existing Tasking, Processing, Exploitation, and Dissemination (TPED) architectures: An estimated 90% of the received data is not even examined – it just “falls on the floor.”

A New Approach is Needed

TPED architecture on next generation technology

Persistent Surveillance generates an unrelenting stream of data. To manage and exploit it we need a new TPED architecture built on next-generation computing technology.

We need a flexible network of sensor and computing assets, designed to meet two objectives:

	multiply the effectiveness of human analysts
	reduce the delivery time for actionable intelligence to tactical users

To meet these objectives, we need an architecture that performs the initial stages of processing and exploitation directly on the tactical sensor platforms, not on ground stations. We need Embedded Smart Processing™.

Embedded Smart Processing – See Clearly, Strike Quickly

Embedded Smart Processing Supports Cross-Cueing

Embedded Smart Processing streams multi-sensor data directly to an on-platform real-time computing system. This embedded computer uses advanced signal and image processing algorithms to make a first pass through the incoming data,

	prioritizing the data for downstream analysis
	tasking the collection of additional data to more rapidly find and fix targets.

Embedded Smart Processing Supports Cross-Cueing from Multiple Sensors

Example - An aided target recognition (AiTR) algorithm can examine large numbers of SAR images and/or motion imagery in milliseconds and provide an alert to any potential threats identified across the fields of view of multiple platforms.

Example - Data fusion synthesizes information from multiple types of sensors to improve situational analysis.

Example - Real-time, cross-cueing, using sensors on multiple platforms to detect, track, and engage threats with a higher degree of precision.

Embedded Smart Processing also makes all sensors part of a flexible, intelligent network. These sensors can then be tasked to support multiple simultaneous users, delivering mission-specific, tailored sensing to both analysts and warfighters on the tactical edge.

Making Embedded Smart Processing Happen

The deployment of Embedded Smart Processing requires embedded, real-time computers that are small, powerful, and rugged. Mercury offers a range of choices, so you can select the best possible solution to your specific needs.

PowerBlock 50

PowerBlock 15

Ensemble 3000 3U OpenVPX

CSNA Vision Encompasses a Flexible Network of Networks

Embedded Smart Processing also requires that these computer systems can be configured and re-configured dynamically into flexible, mission-focused networks. Mercury has created a flexible, standards-based approach to support Smart Processing. The Converged Sensor Network™ (CSN™) Architecture. It combines the power of information management – fusion, exploitation, and dissemination – with signal and image processing to deliver transformational access to information at the tactical edge.

The CSN Architecture Vision Encompasses a Flexible Network of Networks

Admiral mike Johnson networked approach video

A Networked Approach to Exploiting Imagery

The volume of data collected over hundreds of hours of Persistent Surveillance results in processor and analyst-intense tasks, often occurring on the ground rather than on platform, where it is collected.The CSN Architecture vision encompasses a flexible, mission-focused network - combining the power of information management – fusion, exploitation, and dissemination – with signal and image processing to deliver transformational access to information at the tactical edge.

Fusion is the process of the process of synthesizing information from multiple sensors to improve situational analysis. Since each type of sensor data, for example, Electro-optical ISR, SAR, and SIGINT offers unique characteristics, the resulting fused image reveals between 2 and 5X more detail than any input image, detail, depending on the input sensor data. Some types of fusion operations that can be developed with Mercury’s CSNA solutions includes:

Component analysis

Regularization

Signal registration

Unmixing/Decluttering

Wavelet or fuzzy logic analysis

Exploitation is the process of extracting information of interest of an image for identification and classification. This data , processor, and analyst-intense task often occurs on the ground (rather than on platform, where it is collected). The data must be transmitted over bandwidth-limited datalinks - a slow process which frequently mandates the need for data compression.

On Board Exploitation Mercury’s CSN Architecture features automatic on-board exploitation closest to the sensor – directly on the platform. On board exploitation mitigates the need for lossy data compression, and obviates the need for bandwidth limited datalinks because the exploited, processed data can be transmitted to other elements of the network, precluding the bandwidth limitation and compression requirements – and delivering sensor information in a timely manner to those who need it.

Some types of fusion operations that can be developed with CSNA solutions include:

Stabilization

Segmentation

Ortho-rectification

Mosaicing

Form Factor Matching & Comparison

Activity & Coherent Change detection

Flexible, Standards-Based Approach to Sensor Networking

Timely and Accurate Connectivity to Networked Sensor Data

Connectivity in the battlefield has become a requirement. Networked, tactical exploitation in the sky and on the ground solve the data overabundance issue by providing timely and accurate connectivity to networked sensor data.

To enable timely and accurate connectivity to networked sensor data, Mercury offers SigmaNET™ , a software and hardware CSNA-based solution which promotes fusion, exploitation, and dissemination capabilities closer to the sensor. SigmaNet delivers improved efficiencies in R&D and lowers risk to improve product velocity and return on investment because it:

	makes the sensor and related embedded computing system interoperate with net-centric control and data
	makes the sensor and embedded computing assets look like a compute cluster on an IP-based network

Thus, the system communication becomes a network-based service; yielding easy prototyping in the lab and migration to field.

Converged Sensor Networking Architecture

Converged Sensor Networking Architecture

Admiral Mike Johnson Video

Admiral Mike Johnson discusses imagery intelligence and the need for a networked approach to exploiting imagery.

Frequently Asked Questions

Q Tell me again, what does ‘CSNA’ stand for?
A Converged Sensor Network™ Architecture

Q What is meant by the word ‘converged’?
A With CSNA, the sensor, the network, and the embedded computer all come together, or converge, out on the tactical edge, out where the military is either trying to get a look at something or watching for something.

Q Can you give me a short definition of CSNA?
A CSNA merges the flexibility of industry-standard networking with signal processing for sensor data. Sensors, embedded computing systems, data storage, and image display units are all connected by the network.

Q Why extend the network out to ‘the tactical edge’?
A For several reasons:

  • To flexibly and rapidly get information from the sensor to the person who needs it, often a war fighter out on the tactical edge. If that war fighter is connected to the network, then he can get information quickly, when he needs it.

  • If the embedded image processing computer is on the tactical edge, then processed images are what moves over the network, not colossal amounts of raw data.

  • If multiple sensors are network connected, out on the tactical edge, then their information can be more easily and quickly synthesized, that is, combined in a fashion that adds value to the information. For example, an infrared image overlayed on a radar image of a parking lot can determine which vehicles were recently used (still hot).

Q What do you mean by a ‘sensor’?
A An electronic device that captures raw information – examples include a radar antenna, a video camera, and a sonar hydrophone.

Q Is there a software component to CSNA?
A One of the additional values of CSNA is that it can accelerate the movement of new application software – new algorithms – from the laboratory to deployment out on the tactical edge. Using the current technology in the labs, new software is created on clusters of computers, connected by standard IP-based networks. After a new algorithm is created, debugged, tested, and proven to be valuable, it then has to be re-written to run on a deployed, embedded computer. But with CSNA, that deployed computer is not only connected to the network, to the software it looks just like a cluster of lab computers on an IP-based network. The new algorithm software does not have to be re-worked to run on the deployed computer.