Radar System Design Using MATLAB and Simulink

김용정 부장([email protected])
Application Engineer
MathWorks

© 2016 The MathWorks, Inc.

1
Outline

§ Introduction
§ Radar System Design and Analysis
§ ADAS Radar Systems
§ Increasing the fidelity of RF and Antenna models
§ Summary

2
 Radar System Design: From Antenna to Algorithms
Mixed-Signal
Continuous & discrete time Algorithms
Antenna, Antenna arrays beamforming, beamsteering, MIMO
type of element, # elements, configuration
• Simulink (Simscape) • Phased Array System Toolbox
• Antenna Toolbox • DSP System Toolbox • Communications System Toolbox
• Phased Array System Toolbox • Control System Toolbox • DSP System Toolbox

RX Signal and Data

LNA ADC
Processing

PA DAC
TX

• Phased Array System Toolbox

• Communications System Toolbox • Signal Processing Toolbox
• SimRF
• Phased Array System Toolbox • Instrument Control Toolbox
• RF Toolbox
• SimEvents
Channel RF Impairments
interference, clutter, noise frequency dependency, non-linearity, noise, mismatches Waveforms and Task Scheduling
3
Radar System Design with Phased Array System Toolbox
- Specify radar requirements
Waveforms Transmitter Tx Antenna Arrays - Research and design new algorithms
Pulse, LFM, FMCW, etc. Monostatic and Bistatic ULA, URA, UCA, etc.
- Test algorithms with realistic data
- Simulate end-to-end radar systems

Waveform Transmit
Transmitter
Generator Array

Environment,
Targets, and
Interference

Signal Receive
Receiver
Processing Array
Environment effects,
impairments, interferences

Beamforming, Matched Receiver Rx Antenna Arrays

Filtering, Detection, CFAR, Monostatic and Bistatic ULA, URA, UCA, etc.
STAP, etc.

4
 Benefits of Flexible Modeling and Simulation Framework
§ Rapidly model and simulate phased array systems in the MATLAB and Simulink environments
– Interactive development with algorithms and tools specifically for phased array systems
– Explore alternative system architectures and make system level trade-offs
– Access to MATLAB’s visualization and analysis tools
– Capture system requirements in an executable model
– Common “language” and interface across teams and projects

§ Re-use and extend existing code and IP

– Use existing C, MATLAB, and other code
– Open API lets you include proprietary target models and environment models

§ Process offline data in the same environment

– Re-use the same algorithms and test benches

5
What’s New in Radar Modeling?

Waveform Transmit
Transmitter
Generator Array

Signal Receive
Receiver
Processing Array

Array orientation & Phase shift quantization

Wideband support

Range-time Intensity
Doppler-time Intensity 6
 What’s New in Targets, Platform and Environment?
Platform acceleration Rain, fog and gas
Two-ray

Angle dependent backscatter

Fractional Delay Wideband free space

Targets & Platforms Propagation Path Environment Scenario Viewer

7
End-to-End Wideband Radar System – Example Summary

Waveform Transmit
Transmitter
Generator Array

Signal Receive
Receiver
Processing Array

Wideband Radar

– Tx/Rx Array
§ Transmitter, Receiver, Antenna array
Built with blocks from Phased
– Propagation channel Array System Toolbox
§ Path loss, environment
– Signal Processing
§ Beamformer, Matched filter, Stretch processing and Integration
– Target model User-defined MATLAB
§ Custom function with angle and frequency inputs function
8
Pulse Compression in the Wideband Example
§ Pulse Compression § Stretch Processing
– Maximizes peak SNR – Apply dechirp processing to a linear
– Convolution in time domain with FM pulse over a given range span.
matched filter § Resulting sinusoid translates directly to
range
– Multiply in frequency domain with
– Alternative for LFM pulses with large
matched filter
bandwidth signals

Rx (uncompressed) Compressed Rx
Convolution
X FFT

Reference (Tx)

Ref Chirp
or equivalent operation in frequency domain
f Ref Rx

Rx (uncompressed)
FFT X IFFT

Matched filter
FFT
Range
t
9
Wideband Simulink Models

§ End-to-End Wideband System Demo

10
Visualize Radar and Target Trajectories

§ Demo

11
Modeling a Radar Task Scheduler
Beam
DETERMINISTIC [theta,phi,dwell, …]

§ Demo
Health Field of view
over which
tasks are
scheduled
Surveillance

DYNAMIC
Scheduler
Tracking Front-End
& Signal
Processing
(Detection) Data
Processing
Threat (Tracking)

Radar Scheduler Using SimEvents

12
.

ADAS Radar

§ Modeling an ADAS Radar System

§ Long Range Radar Model - Adaptive Cruise Control (ACC)

§ Short Range Radar Model - Blind Spot Detection

13
 Background and Motivation

§ Consumer demand and legislation drives safety improvements every model year
§ Sensor accuracy requirements increase as “passive” safety evolves to “active” system control
§ Momentum of autonomous vehicles accelerates ADAS technology

§ Reasons for using radar in ADAS:

– Range and velocity precision
– Angular resolution and width of view
– Conditions that impact camera vision do not impact radar view

§ Short range radar characteristics

– Wide bandwidth enables high range resolution at short ranges.
§ Long-range radar characteristics
– Range distance and beam width

14
Modeling ADAS Features in Simulink

Integrate sensing (radar, vision, etc…) and control algorithms

Sensor Fusion
Radar-
Algorithm
Control-
Algorithm
Vision-
Algorithm

15
 Designing Radar Systems with Simulink

§ Design a radar component or system

– Mix of models with different levels of fidelity
– Multipath propagation
– Multiple objects
– Develop optimal detection algorithms
Radar Model

§ Integrate a radar component or system

– Validate radar performance and examine what-if scenarios
– Simplify system level debug for anomalous data from road-test
– Evaluate the use of off-the-shelf components to reduce system cost System Model
– Integrate the radar in a larger system and facilitate sensor integration
16
Long Range Radar Model - ACC

§ FMCW radar model

– Waveform generation
– Waveform processing
– Transmitter and receiver
§ Environment model
– Propagation channel
– Obstacles
§ Vehicle model
– ACC speed control
– Vehicle dynamics
§ 3D visualization

17
 Blind Spot Detection Model
§ Built using Long Range Radar example as a starting point

§ Change orientation of the radar

§ Modify positioning and velocity of the vehicles
§ Update detection algorithm

§ Lower transmit power for shorter distance

§ Higher range resolution for distance accuracy

18
.

Increasing the Fidelity of the RF and Antenna Models

19
Antenna Toolbox

§ Easy design
– Library of parameterized antenna elements
– Functionality for the design of linear and rectangular antenna arrays
– No need for full CAD design

§ Rapid simulation setup

– Method of Moments field solver for port, field, and surface analysis
– No need to be an EM expert

§ Seamless integration
– Model the antenna together with signal processing algorithms
– Rapid iteration of different antenna scenarios for radar and
communication systems design
20
Integrating an Antenna Array in a Radar System Model

§ You can integrate your antenna into the radar model built in Phased Array System Toolbox

% Import antenna element in Phased Array

>> myantenna = dipole; Antenna element
>> myURA = phased.URA;
>> myURA.Element = myantenna; Phased Array System Toolbox array

21
Antenna Array, Phased Array, and Radiation Pattern

§ Phased Array System Toolbox arrays use pattern superposition

– ULA, URA, UCA and conformal arrays use the same pattern for all elements
– Heterogeneous arrays have different patterns for different elements
§ Antenna Toolbox arrays perform full wave EM analysis
– Isolated element vs embedded element vs full array

Isolated element Embedded element Full array

pattern(p, 10e9); pattern(l, 10e9, ... pattern(l, 10e9);

'ElementNumber',2); 22
Wideband Antenna Integration Example
§ Demo
With ideal element

With Blade Dipole element

23
Model-Based Design for Radar Systems

RESEARCH REQUIREMENTS • Complete system model

• Requirements traceability
DESIGN
• System behavior exploration
Environment Models
• Cross team collaboration
Timing and Control Logic

TEST & VERIFICATION

• Deploy on desktop
Digital Models Analog Models
• Generate code and HDL
Algorithms
• Integrate into larger systems
• Explore design tradeoffs
IMPLEMENTATION
• Configuration management
C,C++ VHDL, Verilog SPICE
• Automated regression testing
Analog
MCU DSP FPGA ASIC
Hardware
• Report generation
• V&V and security analysis for code
INTEGRATION • Support for certification & standards
24
Summary

§ MATLAB, Simulink, Phased Array System Toolbox, Antenna Toolbox and

SimRF provide flexible platform for radar system design and simulation
– Large number of examples to get started with

§ MathWorks products also enable design and implementation of radar systems

across the workflow

§ Thank you very much for you time today

25