GNU Radio Installation: An Update

January 16, 2012

I tested the installation of GNU Radio (3.5.1) on Fedora 15 and 16 (32 and 64 bits) and Ubuntu 11.10 32-bit (the most recent distribution at this time). Using the build-gnuradio script, I found that Ubuntu is the system on which things are the easiest. The process runs entirely automatic, whereas certain things need to be fixed manually on Fedora during the installation.

Before starting the execution of the script, make sure that the current user has sudo privileges. On Ubuntu, the root account must be enabled with the Linux command sudo passwd. Afterwards, all users (on version 11.10) get the sudo execution privilege.

On my system, I have an USRP2 radio connected to the Ethernet port. The IP address of the radio is 192.168.10.2. Here is the matching Ubuntu network configuration:

The commands ping, uhd_find_devices and uhd_usrp_probe can be used to test and verify connectivity between the Ubuntu system and USRP2 radio (see the post How to setup your own development environment for GNU Radio?).

I run all of this above Oracle VM VirtualBox. You may wish to start by importing my virtual machine in your system (3.24 G file). It has a complete working installation of GNU Radio 3.5.1. The password, for the user liveuser and root, is password in both cases (I recommend changing it).


2011 in review

January 2, 2012

The WordPress.com stats helper monkeys prepared a 2011 annual report for this blog.

Here’s an excerpt:

The concert hall at the Syndey Opera House holds 2,700 people. This blog was viewed about 17,000 times in 2011. If it were a concert at Sydney Opera House, it would take about 6 sold-out performances for that many people to see it.

Click here to see the complete report.


How to setup your own development environment for GNU Radio?

October 30, 2011

This post discusses the setup of a development environment for GNU Radio. I will outline the installation process assuming a Fedora 15 (64 bits) Linux system.

OPTION A

On Fedora-Linux, I found that the approach that works best is to use the build-gnuradio script. It installs the latest version of GNU Radio. It is likely that you will encounter problems during the installation (I did!). Taken one-by-one, they are all solvable.

With the most recent versions, the use of SDRs from the USRP family requires installation of the USRP Hardware Driver (UHD). For a USRP2 unit, new firmware must be loaded into the SD card (follow the installation instructions). At boot time, the USRP2 loads its software from the inserted SD card. A correct installation is confirmed visually by the LEDs that blink according to a certain pattern:

The front panel LEDs A, C and E are synchronously blinking. LEDs D (firmware loaded) and F (CPLD loaded) remain solid. LEDs of the Ethernet connector are blinking and confirming data transfer.

When a SDR software is executed in the receive mode, LED C (receiving) on the USRP2 front panel should also be on:

OPTION B

This option is the easiest. I will install a complete working development environment, but not with the latest version of GNU Radio. Apparently, after version 3.2.2 GNU Radio is not updated in the repositories used to build the distributions of Linux.

However, an up-to-date Linux system is recommended. In particular, the list of available (RPM) software packages needs to be up-to-date otherwise some of the software packages may not be visible. Use the tool Software Update, available under the Applications/System Tools menu, to update your system.

A few software packages need to be installed. The best is to use the Add/Remove Software tool, available under the Applications/System Tools menu.

Launch a Find using the keyword gnuradio. A list of GNU Radio packages will be returned. Check each of them and click on the button Apply.

If you wish to use GNU Radio in combination with the Universal Software Radio Peripheral (USRP) family of radios, you must also install the related software packages. They can be obtained with a Find on the keyword usrp.


Getting PowerSDR, version HPSDR, Up and Running on Windows

October 24, 2011

This post goes through the installation of the High Performance Software Defined Radio (HPSDR) specific version of PowerSDR on Windows (XP is assumed).

Preliminaries: The Windows Environment

First, it is important to ensure that your Windows system is up-to-date. If it isn’t, then it is safer to go through a system update before doing anything to avoid problems latter in the installation. On Windows XP, if you are not certain about the state of your system, then from the Start menu pick Help and Support, then under Pick a task, select Keep your computer up-to-date with Windows Update.

Several iterations may be required before your system is entirely up-to-date.

Next, you need to install the Microsoft .NET Framework version 1.1 (you may use this version, for Windows XP, or download it directly from Microsoft). After, the installation of Microsoft .NET, I highly recommend running cycles of Windows Update again. All these updates may take from few minutes to several hours, but they are worth the investment and prevent problems latter in the installation process. Microsoft .NET specific updates are likely to be applicable. Several iterations may also be needed for that. Upon completion, Windows is all set for the installation of PowerSDR.

Assumed Hardware Configuration

The unit I used to build this post includes a Metis Ethernet interface, a Mercury receive module, a Penelope transmitter module (the first three vertical cards), the Alex RF bandpass filters (grey module placed right), an  Excalibur reference oscillator and a PennyWhistle power amplifier (placed behind the vertical cards).

PowerSDR

It is recommended installing the HPSDR version of PowerSDR in the subdirectory:

C:\Program Files\FlexRadio Systems\HPSDR\

The subdirectory FlexRadio Systems needs to be created under C:\Program Files. The subdirectory HPSDR with PowerSDR can be created by expanding this file, under the subdirectory C:\Program Files\FlexRadio Systems. The official distribution of the HPSDR version of PowerSDR is available here.

The PowerSDR distribution is not complete. To complete the installation, a Skins directory needs to be created. Hereafter, I’m assuming that the user name is Administrator (it needs to be changed with the actual username your are using for your installation). Open Windows Explorer and navigate to the subdirectory c:\Administrator. Right click of the icon representing the directory. The menu Administrator Properties will pop up. Uncheck the box for the Attribute Hidden. The Confirm Attribute Changes menu will pop up. Select Apply changes to this folder, subfolders and files and click OK. Under c:\Administrator, the directory Application Data will be listed. Under the subdirectory Application Data, create the branch FlexRadio Systems\PowerSDR. Expand this Skins file under that branch.

Next, a USB driver needs to be installed. Download and run the USB driver file.

On the desktop, create a shortcut to PowerSDR. The path should be like C:\Program Files\FlexRadio Systems\HPSDR\PowerSDR.exe.

PowerSDR is now ready to run. Start by clicking on the PowerSDR shortcut. The first time you run it, the PowerSDR Setup Wizard will pop up. In the first panel, HPSDR needs to be selected among the radio models supported by HPSDR.

On the next panel, for my specific hardware configuration, the Mercury, Penelope and Alex boxes need to be checked.

Finally, when PowerSDR is loaded enter the PowerSDR Setup menu. Under the Hardware Config panel, check the Alex box, listed under HPSDR Hardware Present, and Metis, under Connection type.

If everything went well, then PowerSDR should look like the following when started.

Visit the link PowerSDR page of the openSDR wiki for more details and information related to other HPSDR hardware configurations.

See the related post Digital Communications with the High Performance Software Defined Radio (HPSDR).


Digital Communications with the High Performance Software Defined Radio (HPSDR)

October 18, 2011

The High Performance Software Defined Radio (HPSDR) is a great platform for experimenting with the software defined radio concept and digital modes in general. The following is a picture of the HPSDR unit I used to build this post.

It includes the Metis Ethernet Interface, Mercury RX Module, Penelope TX Module, Excalibur Reference Oscillator, PennyWhistle PA and Alex Bandpass filters.

PowerSDR is the PC-based software used to control the HPSDR unit and process incoming and outgoing signals. Digital Master (DM) 780 together with Ham Radio Deluxe (HRD) are applications supporting several different modes of digital communications.  This post reviews a configuration of  HPSDR-PowerSDR, DM 780 and HRD that makes them work together. Following the installation of PowerSDR, HRD and DM 780, the main issue is the creation of three  software serial communications lines. Hereafter, we assume that PowerSDR (the HPSDR specific version), HRD and DM 780 have been installed.

One virtual serial line is used to control PowerSDR (and HPSDR) from HRD. A software is required to create the serial line. The null-modem emulator com0com is recommended. After downloading and installing the software, the following configuration needs to be done (accessible through the setup program):.

The pair of virtual ports COM7 and COM17 need to be created and are used to communicate the HRD to PowerSDR control commands.

Two additional virtual lines are needed for the data flow between DM 780 and PowerSDR. A second software needs to be installed to create the two virtual lines. The Virtual Audio Cable software is recommended.  Following the download and installation of the software, two virtual lines must be created as follows.

Note that the cable 2 may have to be created explicitly and that the SR range needs to be set to the values as indicated above.

Enter the CAT Control panel of the Setup menu of PowerSDR. The CAT Control must be configured as follows to enable control from the HRD software.

The Enable CAT box is checked together with port COM7 and Baud 9600 selected.

Enter the VAC subpanel of the Audio panel within the Setup menu of PowerSDR.

The Enable VAC box must be checked together with Input Virtual Cable 1 and Output Virtual Cable 2 selected. The Auto Enable box must be checked also to activate the virtual lines only when PowerSDR is in digital mode. PowerSDR is ready to operate in digital mode with HRD and DM 780. On 40M, PowerSDR may look like the following.

A connection between HRD and PowerSDR must be setup within HRD. This can be done at HRD startup (or through the Connect menu item). The parameters of the connection should be as follows.

The  HRD software should look like the following.

The displayed frequency matching the one on which PowerSDR is tuned is a positive feedback.

The last step is the configuration of DM 780. Enter the Soundcard section under the Tools menu item of DM 780.

The Input Device Virtual Cable 2 must be selected. Likewise, the Output Device Virtual Cable 1 must be selected. Just the reverse of what we did under PowerSDR. The DM 780 is now ready to operate. In the Phase Shift Keying (PSK) 31 mode, it should look like the following.

See this live operation video.


History of Cryptology by Richard Brisson

June 19, 2011

This winter, in my course COMP4203 on Wireless Networks and Security, Richard Brisson came and made a presentation on the History of Cryptology. Richard Brisson is an avid collector of cryptography and spy communications artifacts. Most impressive is his personal collection of rare Enigma machines used during the second World War by the Germans. To have a peek at Richard’s collection, check out the web site campx.ca.

Richard talking to the students.

A close up at one of Enigmas in Richard’s personal collection. See any difference with a normal computer keyboard? (Hint: look at the placement of the letters)

 

Matthew (a student in the course) and Richard. The machine on the left, is an electronic Enigma I fabricated a while ago.


Beware the Rogue Access Points and Web Traffic Hijacking

February 15, 2011

Two Carleton University 4th year computer science students, Michael Nemat and Matthew Ng, starred in a story shown February 14, 2011 on Global National, the national newscast of Canada’s Global Television Network. They were interviewed by their Ottawa correspondent Shirlee Engel on the rogue access point and web hijacking attacks. Click on the following image to wiew the video of the interview.

The participants and their relationships in the attacks demonstrated in the interview can be pictured as follows:

It involves two aspects: luring a victim with a free insecure WiFi access point and hijacking its web traffic when the victim starts using the rogue access point. Traffic may be hijacked by the adversary running an attack called SSL Strip. Credentials of the victim can be intercepted and saved, for future malicious uses, when the victim starts visiting theoretically secure web sites.

When the victim is about to enter its user name and password, its web browser is normally switched to a secure web connection. The adversary intercepts the procedure and breaks the connection into two segments. The first segment is a secure connection between the adversary and web server. This is required for normal operation with the web site. The second segment is an insecure connection between the adversary and victim. The adversary sits in the middle and bridges the two connections while copying in clear all the information typed by the victim.

Truly, the problem is not that the web server is insecure, but the failure to establish an end-to-end, browser-to-server, web connection. You may read the technical details about the SSL Strip attack by following this link.

Detection of the attacks by the victim requires recognizing a few differences in the presentation of the web pages. Look first at the following web page (you may double click on it to make the picture bigger):

Now look at the corresponding true web page:

The browser is Internet Explorer. With a secure end-to-end connection from the browser to the web server (second picture), the web site name is prefixed by “https:”. Moreover, the web site’s legitimate favicon (the company logo) is placed to the left of the name. A padlock is placed to the right. With an insecure connection (first picture), the name  is prefixed by “http:” with a fake padlock placed to the left. No padlock is displayed to the right. Here are two more examples.

Fake web site:

True web site:

This time, the browser is Chrome. With a secure end-to-end connection, the “https:” part of the web site name is highlighted in green with a green lock displayed to the left. In the absence of a secure end-to-end connection with the web site, there is no “https:” prefix shown in the name (not even a “http:”). There is no green padlock to the left of the name.

Fake web site:

True web site:

The browser is Safari. With a secure end-to-end connection from the browser to the web server (second picture), the web site name is prefixed by “https:”. This is it! Safari is the worst among all three browsers for security awareness.

The three bank web sites are compromised in a similar fashion. There are noticeable visual differences similar in nature for all of them. In the fake cases, the web site’s favicon may be replaced by a picture of a padlock that mimics the one the browser uses to indicate a secure connection with the web site. The prefix used for the web site name is “http:” instead of “https:”. The the absence of the “https:” prefix is the main indicator of an insecure connection.

Users can protect themselves from this attack by bookmarking the “https:” version of their bank web site name, as opposed to navigating their way from the unsecure page of the bank (the “http:” version) to the secure login page. The browser will initiate all communications through a secure connection and the adversary will have no chance to tap in. There are, however, web sites that do not redirect to a secure page before logging in (such as Facebook, Twitter, and many others). They are the most vulnerable as there are no visual clues that a secure connection is not present:

Fake web site:

True web site:

The browser is Internet Explorer. There are no visible differences. The user would have to look a the source code of the web page to observe replacement of “https:” (the true case) by “http:” (the false case) in the web site name.

To be fair with Facebook, if you fail to login from the main page, then you are redirected to the following login page with the security awareness features such as the padlock and “https:”

My advice, before you are about to enter any personal data on a web page, look for the true indicators of an end-to-end secure connection with the server such well-placed padlocks and web site names stating with the “https:” prefix. They should remain for the whole duration of your session with the web server. Avoid using the same password across different web sites. If your password for one web site has been exposed to an unauthorized person, then the door is opened to your other accounts because of an identical password. You may read the newspaper or check weather forecast using free public WiFi access points.

Acknowledgements: Thanks to Michael Nemat for providing me input and visual material for building this post.


Wireless in Movies: Windtalkers

February 15, 2011

The 2002 movie Windtalkers tells the story of 420 Native Americans who served as radio operators in US troops in the Pacific during Second World War. They were originating from the Navajo Indian reservation of Arizona. Because their language appeared so cryptic to people outside the reservation, Navajos were used a radio operators to secure wireless communications. Typically, a message in English was given to a radio operator, translated in Navajo, transmitted, received by another Navajo operator and than finally translated back to English.  You may view a segment from this movie here:

Japanese were confused by the Navajo code and they broke it only partially.


2010 in review

January 2, 2011

The stats helper monkeys at WordPress.com mulled over how this blog did in 2010, and here’s a high level summary of its overall blog health:

Healthy blog!

The Blog-Health-o-Meter™ reads This blog is doing awesome!.

Crunchy numbers

Featured image

A Boeing 747-400 passenger jet can hold 416 passengers. This blog was viewed about 2,100 times in 2010. That’s about 5 full 747s.

In 2010, there were 13 new posts, not bad for the first year! There were 39 pictures uploaded, taking up a total of 47mb. That’s about 3 pictures per month.

The busiest day of the year was December 1st with 61 views. The most popular post that day was D-Link DWA-125 on Linux Ubuntu 10.04.

Where did they come from?

The top referring sites in 2010 were people.scs.carleton.ca, alphainventions.com, mail.google.com, qrz.com, and cordless-homephone.info.

Some visitors came searching, mostly for usrp2, dwa-125 linux, kubuntu installing d link dwa 125 usb wireless, d-link dwa-125 ubuntu, and xbox 360 ad hoc.

Attractions in 2010

These are the posts and pages that got the most views in 2010.

1

D-Link DWA-125 on Linux Ubuntu 10.04 October 2010

2

Ad Hoc Networking an Xbox 360 and a Windows PC October 2010

3

USRP2 FM Broadcast Receiver Demo September 2010

4

Configuration of the D-Link DWA-125 in Managed Mode October 2010

 

5

KoreK Chopchop Attack on WEP Metworks September 2010


Herzberg, Foucault and Polar Orbits

November 28, 2010

This is the building where I work. It is called the Herzberg building, after the name of the Canadian scientist and Nobel laureate (1971) Gerhard Herzberg:

Let’s get inside the building, through the main entrance. On the wall, next to the elevator there is a commemorative plaque with a short biography of Herzberg:

You may read it while waiting for the elevator. If you wish to stay in shape, it is best to walk and take the stairway. While you are climbing the stairs you may look at an interesting demo, of a physical phenomenon, embedded in the staircase. The apparatus was designed by the French scientist Léon Foucault. It is called the Foucault pendulum. A huge pendulum is suspended from the ceiling of the five–floor building. It is oscillating above a large orthographic projection of the earth centered on Ottawa. This short video was taken around 9:45 AM:

Here is another video, taken from the same observation point around 4:30 PM:

Has the direction of the pendulum oscillation changed? No! It is the map of the earth that has rotated. The direction of the oscillation is fixed. It is like if the pendulum was suspended from a fixed location, far in space, and was oscillating above the rotating earth. This idea is illustrated by Melodie:

By the same principle, a satellite orbiting around the earth will oscillate following a fixed circle around a rotating earth. Because of the rotation of the earth, the satellite will cover from one cycle to another different parts of the globe as shown in the following video:

This type of orbit, called polar, is used for telecommunications, observation, spy or wheatear satellites because the birds have the possibility to scan the entire earth after a couple of rotations.


Follow

Get every new post delivered to your Inbox.