Proposal:



Networked UAV C3

Program Description

October 4, 2004

Ken Davey

L-3 Communications Corporation

ComCept Division

2800 Discovery Blvd.

Rockwall, Texas 75032

Phone: 972-772-7501

Fax: 972-772-7510

ken.davey@L-

Timothy X Brown,

Interdisciplinary Telecommunications

Electrical and Computer Engineering

University of Colorado, Boulder

80309-0530

Phone: 303-492-1630

Fax: 303-492-1112

timxb@colorado.edu

Introduction

The ComCept division of L-3 Communications, along with the University of Colorado (CU), has designed, installed and operated a wireless communications test bed involving an 802.11 based Mobile Ad Hoc Network (MANET) and Unmanned Aerial Vehicles (UAVs) which will interact with it. The Networked UAV C3 program builds upon this unique platform and experience base in order to develop advanced Communication, Command, and Control (C3) for networks of small UAVs. It is viewed that this effort will take place in stages. The next section will provide an overview of the multistage effort.

Overview of Networked UAV C3

Small (10kg) UAVs are low-cost low-risk assets that have unique capabilities in military, scientific, and commercial applications. The main benefit of these UAVs is that they can be deployed in groups. Group deployment increases operational success even with UAV losses; enables simultaneous sensing/searching over larger areas than a single UAV; and supports longer range communication with low-power radios through multi-plane networking. This section presents the basic architecture we are developing and the three-stage program for achieving a full-scale demonstration of multiple UAVs networked to collaboratively solve a task.

1 Architecture

A key concept behind Networked UAV C3 is UAV Flocking. The main idea in UAV Flocking is that groups of UAVs are controlled with simple-to-specify high-level mission commands while communication and intelligence on and among UAVs reduces these commands to low-level actions. To achieve this goal operational commands are defined in a three level hierarchy:

• Level One: UAV Control Systems: The lowest level is the physical actions taken by the planes to operate control surfaces, manipulate payload sensors, and establish communication. At this level the surfaces, sensors, and radio separately act to achieve individual goals set by the middle level.

• Level Two: System Command & Communication: The middle level sets the short term goals for the lower layers such as fly the plane to a given waypoint, image a point on the ground, or send collected data to a ground gateway. The middle level integrates commands from the higher level and information collected at the lower level to set and update goals for the lower level. At this level, the planes act separately to achieve individual goals set by the highest level.

• Level Three: Mission Tasking & Network Intelligence: The highest level addresses a mission task concept that must be completed by a group of UAVs. Examples include, locate a radio with a given profile and track it while maintaining connectivity with the ground gateway. Choose UAV positions that maintain network connectivity between radios within a group of moving vehicles. These goals are set by a UAV operator or possibly a higher level of control. This level integrates information collected at the lower level with mission task assessment and sets goals for the individual planes. A key element of this level is that the decisions are made through distributed decision making among the UAV nodes based on individual node capabilities and the state of the flock at any given time. The distributed approach is more robust to variations in individual node capabilities and spreads the command load across multiple nodes.

This separation into hierarchy allows the problem to be broken into solvable pieces that build to a complete solution. By shielding the operator from low-level C3 details, the operator can focus on specifying and achieving tasks that are easily understood by humans. For instance a specific task might be specified as: (1) deploy 4 UAVs to location X, (2) measure radio activity around this location, and (3) return to start. The operator sends simple low-bandwidth commands and can easily modify the mission as data is collected. It can be imagined that this eventually develops into a simple graphical point and click interface.

The flocking concept allows non-UAV specialists to command multiple UAVs to achieve a desired outcome. The UAVs themselves will require specialists to outfit, launch, recover, maintain, and store the vehicles. To complete the flocking concept, multi-UAV operations need to be developed to minimize the number of UAV specialist associated with a group of UAVs. For instance, three UAV operators could launch a group of UAVs into a station keeping position and then hand control to the mission operator.

The flocking concept is best demonstrated by developing specific flocking applications that provide measurable performance benchmarks. Two applications are relevant given current capabilities and expertise:

• Communication Leashing: In this application a single UAV positions itself as a relay between a single mobile ground node and a ground network gateway. As the ground node moves away from and around the gateway the UAV must choose the position that provides the best throughput. Because of terrain, this is more complex than simply flying near the midpoint of the two nodes. The plane will integrate radio signal strength measurements from each node and geographic data to decide the best position. Further constraints can be placed such as no-fly zones. While this is only a single UAV flock, it emphasizes the communication and control interactions between the different layers in the hierarchy and so is a good starting application. The concept can be extended to multiple UAVs who collectively maintain connectivity between more widespread mobile ground nodes.

• Radio Direction Finding: In this application a group of UAVs collaborates to find the source of a radio transmitter. The UAV has a GPS to know its own position and a sensor that can measure properties of a radio signal (e.g. signal strength, angle of arrival, time of arrival). Generally, the problem is to solve for the location, antenna pattern, and transmit power of the unknown transmitter. It is complicated by the variability of radio signals and possible target movement. The UAV flock has unique capabilities for solving this problem efficiently since it can share information to develop location estimates and then actively choose actions that refines these estimates quickly. For instance, if the estimated location is to the North of the flock, then the signal strength should increase if the flock flies north. Further, the flock can spread out in an East-West axis to lengthen the measurement baseline. This application emphasizes cooperation and distributed decision making of the UAVs.

2 Networked UAV C3 Program:

It is anticipated that there will be 3 stages associated with the Networked UAV C3 program. The 3 stages are described below along with approximate efforts.

• Stage 1: Integration of sensors, inter-plane communication equipment, flight control systems, and software algorithms for tying network intelligence and mission-level tasking information into automatic UAV flight controls. This addresses Level 1 in the Three-Level Hierarchy of Networked UAV C3. (6-7 months)

• Stage 2: Integration of commands from the higher level and information collected at the lower level to set and update goals for the lower level. At this stage, the planes act separately to achieve individual goals set by the highest level. Communication Leashing is to be demonstrated. This addresses Level 2 in the Three-Level Hierarchy of Networked UAV C3. (12 months)

• Stage 3: Implementation of Direction Finding for demonstration, and demonstration of C3 via external interfaces – including graphical interfaces. This addresses Level 3 in the Three-Level Hierarchy of Networked UAV C3. Improved UAV operation techniques will be recommended. (12 months)

The effort is not necessarily sequential and some overlap between stages is possible. Figure 1 shows the relationship between the main activities in each stage.

[pic]

Technical Approach and Detailed Activities

The Stage 1 activity will implement autonomous control of flight, communication, and sensor systems. The next section describes the technical approach followed by the specific tasks to be performed under this contract.

1 Technical Approach

The UAV can be functionally composed of an Autonomous Flight Control, Communication Networking, and Sensors. The AFC operates the control surfaces, measures position and orientation information, and guides the plane to reach waypoint or flight pattern goals. The networking communicates with other UAVs and ground nodes to route traffic to or from the UAV. It also acts as a relay node in an ad hoc (aka mesh) network formed with other network nodes. The sensors might be video cameras or simple temperature, chemical, or other probes. The main tasks in this stage are to develop the hardware for each subsystem and to define interfaces between these subsystems and a central coordinating controller. The architecture is shown in Figure 2.

The AFC effort will integrate a commercial autonomous flight control system into the current UAV vehicles. This requires acquiring, installing, integrating, and flight testing the system. The control loops will be modified to incorporate flight patterns necessary for our applications. The communication networking effort will modify current ad hoc routing software to optimize its performance for UAV nodes and to introduce traffic differentiation based on quality of service goals. The sensors effort will incorporate a camera and a simple probe into the airframe.

The interface development will leverage existing COTS communication technologies such as CAN, TTL UART, I2C Bus, SPI Bus, and Ethernet. More importantly, software interfaces will be developed that provide a layer of abstraction in the Central Controller that shields it from the details of which technologies are being used to assess a subsystem. This abstraction layer would consist of standard drivers that would initialize and communicate with the subsystem side and would provided a uniform API for service discovery, command, and communication on the Central Controller side.

[pic]

-----------------------

Figure 1: Networked UAV Concept Development

Stage 3

Stage 2

Stage 1

C3 via external interfaces

Multi-UAV operations

Direction finding algorithms

Leashing algorithms

Direction finding flight demo

Leashing flight demo

Base Demo: UAV flies to target. Sends back images

Level 3: Multi UAV planning and decision making

Level 2: Single UAV planning, inter plane control interface

Autonomous Flight Control

Sensors

Central Controller

Communication Networking

UAV

Level 1: Intraplane subsystem com-munication and control

Figure 2: Level 1 Architecture

................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download