ECE 477 Digital Systems Senior Design Project Fall 2010

Homework 3: Design Constraint Analysis and Component Selection Rationale

Team Code Name: Purdue Airbus Group No. 1

Team Member Completing This Homework: Chandler Wall

E-mail Address of Team Member: cswall@purdue.edu

Evaluation:

SCORE / DESCRIPTION
10 / Excellent – among the best papers submitted for this assignment. Very few corrections needed for version submitted in Final Report.
9 / Very good – all requirements aptly met. Minor additions/corrections needed for version submitted in Final Report.
8 / Good – all requirements considered and addressed. Several noteworthy additions/corrections needed for version submitted in Final Report.
7 / Average – all requirements basically met, but some revisions in content should be made for the version submitted in the Final Report.
6 / Marginal – all requirements met at a nominal level. Significant revisions in content should be made for the version submitted in the Final Report.
* / Below the passing threshold – major revisions required to meet report requirements at a nominal level. Revise and resubmit.

* Resubmissions are due within one week of the date of return, and will be awarded a score of “6” provided all report requirements have been met at a nominal level.

Comments:
1.0 Introduction

The Purdue Airbus is an unmanned, autonomous flying vehicle (UAV). It will be equipped with an autopilot system that will operate the plane’s spatial controls in order to maintain an appropriate flight path. The autopilot system will also have the ability to autonomously navigate the plane between provided GPS coordinates. The autopilot system will be augmented by a digital signal processor that will encode and compress image data captured by the on-board camera. Once the image data has been compressed, it will be stored locally and transmitted through the air in real-time.

The second half of the Purdue Airbus project is a compact ground station. The ground station will provide an interface for modifying the autopilot’s on-board GPS coordinates, monitoring the plane’s telemetry, and viewing real-time image data transmitted by the plane. Wireless communication between the plane and a ground station will be maintained using a pair of RF modules. This document is meant to outline the various design constraints associated with the project as well as explain the rationale behind the selection of any major components.

Updated Project Specific Success Criteria

1.  An ability to embed/overlay real-time telemetry data with image data using the ground station.

2.  An ability to transmit real-time image data from the UAV to the ground station using wireless modules.

3.  An ability to store encoded image data on a flash memory module interfaced with the digital signal processor for later viewing.

4.  An ability to to control the UAV and upload new navigation coordinates from the ground station.

5.  An ability to monitor battery status and enter into a power save mode, halting video capture and transmission in order to preserve power for autopilot operation.

2.0 Design Constraint Analysis

This section will be used to enumerate, quantify, and describe the various design constraints associated with the project. The most critical of these design constraints pertain to the computation ability and on-chip peripheral availability of the digital signal processor. Other important factors include interface requirements, off-chip peripheral requirements, and power characteristics. Lastly, packaging and cost constraints will be discussed, too.

2.1 Computation Requirements

Encoding and compressing image data in real-time is a computationally intense process. In order to encode and compress image data supplied by the on-board camera, a high-performance digital signal processor is necessary. Since image data will be stored and transmitted in real-time, the encoding, compression, and memory management algorithms must be efficient and optimized for the target data. These algorithms will benefit from the enhanced instruction set and dedicated video ALU available on certain digital signal processors. Additionally, the extensibility of most digital signal processors’ memory architecture will allow external memory devices to serve as high speed, large capacity image buffers.

Depending upon the digital signal processor’s computing capabilities, the target image resolution will be QVGA (320*240 pixels) or VGA (640*480 pixels). At these resolutions, a single frame of 16-bit RGB bitmap data will require 1.2 Mbit or 4.9 Mbit of memory for storage, respectively. Greater complexity is introduced with the desire to transmit up to five frames of compressed image data per second and store up to fifteen frames of compressed image data per second. These high bandwidths will require memory access with minimal overhead.

2.2 Interface Requirements

When viewed at a macro level, the Purdue Airbus will be comprised of three independent electrical devices and a ground station. While each of these devices is capable of operating independently from the others, the overall success of the system depends upon a fluid, reliable, and lightweight interface that binds the devices together. A fluid interface will allow data to be transferred between devices with minimal overhead. A reliable interface will reduce the probability of system failure and data corruption. A lightweight interface will reduce power consumption and simplify the system’s layout.

The three independent electrical devices previously mentioned include an autopilot module, a wireless module, and a camera module. The autopilot module and wireless module can be interfaced using standard serial buses, specifically SPI and UART. The camera module, however, utilizes a combination of serial and parallel interfaces. The camera’s serial interface use an I2C bus, while the 8-bit parallel interface will occupy general purpose I/O pins. Additional general purpose I/O pins will be required for the camera’s five strobe lines. The combination of UART and I2C interfaces is not provided on many digital signal processors. Accordingly, a PLD will be incorporated for seamless I2C bus support. This PLD will require eight general purpose I/O pins. Therefore, a minimum of twenty three general purpose I/O pins will be required.

Since the autopilot module, wireless module, and camera module do not share a common operating voltage, voltage translators will likely be required to ensure a fluid interface. The lack of non-CMOS loads reduces the need for extraordinary current sinking/sourcing capabilities.

2.3 On-Chip Peripheral Requirements

As mentioned above, the autopilot module and wireless module require standard serial buses. For this reason, the digital signal processor must support SPI and UART. Additionally, in order to support an extended memory architecture, the digital signal processor must contain peripheral DMA. Since these requirements are standard on most general purpose digital signal processors, they will not drastically impact the hardware selection.

In order for the ground station to communicate with the plane, it must support a serial bus for the wireless module. Since the ground station will be powered by an Intel Atom board, USB connectivity will be available for serial communication with the wireless module. In order to utilize the USB port on the Intel Atom board, the wireless module will require an adapter that transforms the UART bus to USB.

2.4 Off-Chip Peripheral Requirements

The design will rely heavily on off-chip peripherals for correct operation. The most complex of these peripherals is the autopilot system which will allow the plane to fly autonomously. The autopilot system is comprised of two microcontrollers and a variety of sensors which provide telemetry data. The microcontrollers interpret the telemetry data from the sensors and modify the flight controls in order to maintain a stable flight path.

In addition to the autopilot system, the design will require a pair of wireless modules that allow communication between the plane and ground station. The wireless modules must be able to maintain transmission speeds capable of transporting real-time image and telemetry data. Additionally, the wireless modules will need to resist interference to ensure uninterrupted operation.

In order to capture image data, the design will also require a camera. The camera must output color image data at a minimum frame rate of ten frames per second. To ensure image clarity, a minimum of QVGA (320*240 pixels) resolution is desired.

Completing the off-chip peripheral requirements for the Purdue Airbus is the need for a compact platform that can be used as the basis for the ground station. The compact platform must be capable of decoding and displaying the image and telemetry data provided by the wireless module.

2.5 Power Constraints

The majority of the components required in order for the Purdue Airbus to function correctly will operate solely on battery power. Additionally, all of the plane’s components will need to fit within the airframe. This will prevent the design from utilizing multiple large batteries. Due to this restriction, efficient components must be selected. Ideally, the selected components will be able to enter into a sleep or reduced power mode when they are not needed.

2.6 Packaging Constraints

The size of the airframe is the primary packaging constraint. In order to ensure proper placement of the system’s components, care will have to be taken when deciding upon the orientation of each component. Additionally, proper placement of the wireless module and its antenna is critical to quality signal reception. Also worth considering is the ideal position for the camera. Depending upon where the camera is mounted, the perspective of the image data will change drastically.

2.7 Cost Constraints

Although the market for commercially available UAVs appears to be expanding, there are still few options available to the general public. The primary limiting factor is the cost of such devices. The DraganFlyer X4 is a quadrocopter that uses on-board telemetry sensing devices to assist the driver during remote operation. The telemetry information is transmitted from the device to a compatible ground station. The base model has a hefty price tag of $8,495. [1] Also worth mentioning is that the base model does not include GPS navigation, an on-board camera, or the ability to fly autonomously.

While this device does not provide a direct comparison to the Purdue Airbus, the two devices provide similar functionality. In order to ensure usability and availability of a device based on the Purdue Airbus, the project’s budget will be limited to $1,000. At this price, the Purdue Airbus should be affordable to general consumers, hobbyists or enthusiasts with a niche to fill.

3.0 Component Selection Rationale

The component selection process is influenced by a variety of factors. The design constraints listed in this document describes the primary criteria; however, other factors also influence component selection, including price, availability, documentation, and support. The major components which require the greatest consideration are the digital signal processor, wireless module, and camera module. In this section, a digital signal processor and a powerful microcontroller are compared: ADSP-BF533 from Analog Devices and PIC32MX440F512H from Microchip. Additionally, three cameras are compared: TCM8230MD from Toshiba, TCM8240MD from Toshiba, and CM-26N/P from Hanse. Lastly, two wireless modules are compared: XBee-PRO 802.15.4 from Digi and XBee-PRO XSC from Digi.

Although both processing devices meet the interface requirements, the ADSP-BF533 from Analog Devices more closely meets the computational constraints. The ADSP-BF533 is a high-performance, 32-bit digital signal processor capable of operating at 600 MHz. It has 148 KB of on-chip SRAM and supports a large index space for external Flash memory. There are two 16-bit MACs, two 40-bit ALUs, and four 8-bit video ALUs. It also has numerous peripheral interfaces, including SPI, UART, eight peripheral DMAs, and general purpose I/O pins. [2] These various interfaces provide a great deal of flexibility. The PIC32MX440F512H from Microchip is a high-performance, 32-bit microcontroller capable of operating at 80 MHz. It has 32 KB of on-chip RAM and 512 KB of on-chip Flash memory. Similar to the digital signal processor from Analog Devices, there are numerous peripheral interfaces, including two I2C, two UART, and two SPI. [3] Aside from outpacing the microcontroller from Microchip, the digital signal processor from Analog Devices has more SRAM. Although the microcontroller from Microchip has more Flash memory, this isn’t a deciding factor since additional Flash memory would still be required for image data storage.

The three color, CMOS based cameras each provide similar functionality. The two cameras from Toshiba provide a parallel data bus for capturing image data, while the third camera from Hanse provides an analog video output that conforms to NTSC or PAL standard. [6] Since the TCM8230MD from Toshiba is capable of capturing image data at thirty frames per second, it is a strong match for the design constraints. [4] The TCM8240MD from Toshiba, however, supports native image compression and higher resolutions. [5] Overall, the TCM8230MD from Toshiba is the best match for the project’s needs.

The two wireless modules from Digi provide similar functionality and features. The XBee-PRO XSC is a long range, high-performance, 900 MHz wireless module. It has an outdoor range of up to fifteen miles with line-of-sight and data rates up to 10 kbps. [8] The XBee-PRO 802.15.4 is a long range, high-performance, 2.4 GHz wireless module. It has an outdoor range of up to 1 mile with line-of-sight and data rates up to 250 kbps. [7] Since the XBee-PRO 802.15.4 can sustain data rates twenty five times faster than the XBee-PRO XSC, it is the best match for the design constraints. Without the additional bandwidth provided by the XBee-PRO 802.15.4, real-time image data transmission would not be possible.

4.0 Summary

This report has enumerated the major design constraints necessary for project success. Additionally, qualified components have been compared, and the best suited components have been selected for use. The critical components that were selected were the ADSP-BF533 from Analog Devices, the TCM8230MD from Toshiba, and the XBee-PRO 802.15.4 from Digi. These components were selected for their superior performance, peripheral support, interface options, and availability.


List of References

[1] Dragan Fly Innovations Inc., DraganFlyer X4 Product Page. [Online]. Available: http://www.draganfly.com/quote/quote-df-x4.php

[2] Analog Devices Inc., “Blackfin Embedded

Processor,”ADSP-BF531/ADSP-BF532/ADSP-BF533 Datasheet Rev. G, May 2010

[3] Microchip Technology Inc., PIC32MX3XX/4XX Datasheet, Apr. 2010

[4] Toshiba Corporation, “VGA Camera Module,” TCM8230MD Datasheet, Jun. 2003

[Revised Jan. 2004]

[5] Toshiba Corporation, “1.3 Mega pixel sensor chip,” TCM8240MD Datasheet, Dec. 2003