You should have a look at the GNU Radio framework, you can start with SDR/DSP without any hardware. Look at

http://gnuradio.org/redmine/projects/gnuradio/wiki/Simulations

There is also various hardware that supports GNU Radio:

http://gnuradio.org/redmine/projects/gnuradio/wiki/Hardware

Depending what you want to do a cheap sound card or USB radio receiver (with custom drivers) is enough, most people use the USRP2 hardware though.

In general, a good way to start is to look at the tutorials

http://gnuradio.org/redmine/projects/gnuradio/wiki/Tutorials

http://www.linuxjournal.com/article/7505 (how to implement an FM receiver in GNU Radio)

and the existing projects (https://www.cgran.org/wiki/Projects), and start from there.

There are also youtube videos on how to do various things in GNU Radio.

Still you should expect a steep learning curve because SDR consists of many aspects in CS and EE, and there are lot of systems issues (getting GR to run, getting the hardware to run, understanding how to use GR blocks, ...)

The only thing I have achieved in this regard is to use a fast sampling digital oscilloscope like the picoscope 3206A with a free wire as an antenna. A long term FFT will show up frequency of all local AM and FM stations. The first step in all software radio is ultrafast sampling.

Thank you Sir,

Can you please elaborate more?? and what is the complexity of this job?

The only complexity was the high cost of the oscilloscope. I just attached a short length of wire to the probe to act as an antenna. The rest was done using the supplied software which includes FFT. It only identifies the frequency and relative strengths of local AM and FM radio stations. Presumably some of the signal content from each of these stations is also embedded in the oscilloscope data stream but data resolution may not be enough to pull it out for FM, but maybe for AM some progress could be made with this equipment, and some clever software.

And of course you need a PC connected to the oscilloscope. The thing about radio software is to sample rapidly and then the complexity comes in the signal analysis to discriminate channels, rather than tune to a single by hardware. This signal analysis itself is a subject of ongoing research; as simple or complex as you choose to make it depending on what your desired outcome is.

Once I have oscilloscope and wire (simple wire correct which is coming out from oscilloscope [I have not used this before]?) and do I need to do some programming for this? which kinda?

To get out any signal related to a particular station (like the audio from an AM station) out of the signal there is a lot of complex signal analysis and programming involved once you have the total high speed data stream in your computer. It is not easy! and far to complicated to go into any more detail here.

Pulkit same equipments as Andrew Jonkers has mentioned ? any thing more then that??

You should have a look at the GNU Radio framework, you can start with SDR/DSP without any hardware. Look at

http://gnuradio.org/redmine/projects/gnuradio/wiki/Simulations

There is also various hardware that supports GNU Radio:

http://gnuradio.org/redmine/projects/gnuradio/wiki/Hardware

Depending what you want to do a cheap sound card or USB radio receiver (with custom drivers) is enough, most people use the USRP2 hardware though.

In general, a good way to start is to look at the tutorials

http://gnuradio.org/redmine/projects/gnuradio/wiki/Tutorials

http://www.linuxjournal.com/article/7505 (how to implement an FM receiver in GNU Radio)

and the existing projects (https://www.cgran.org/wiki/Projects), and start from there.

There are also youtube videos on how to do various things in GNU Radio.

Still you should expect a steep learning curve because SDR consists of many aspects in CS and EE, and there are lot of systems issues (getting GR to run, getting the hardware to run, understanding how to use GR blocks, ...)

Thanks Matthias Wilhelm, I will look upon this tutorials and let you know,

And to all : hows future perspective with respect to industry demand in SDR?

Depends on the industry of course, I think in cellular they already rely on SDRs heavily because the standards are advancing fast. Currently there is much work to push the signal processing into the cloud (i.e., the antennas only record digital samples and the processing is done somewhere else), so there really is potential to save infrastructure and energy cost for them.

Military applications should also be considered, they have lots of waveforms and it's costly to implement everything. There is (was?) a project called Joint Tactical Radio System, but it seems technology is not ready for what they wanted to do ... But if you look at the standard they wanted to do (Software Communications Architecture), the problem may rather be over-engineering. So there may still be hope there.

In consumer hardware they probably do most things digitally, but you won't see programming interfaces anytime soon, since you can send anything anywhere it's hard to get it certified.

Another thing is using frequency whitespace, there may be products coming up in the next years (maybe).

This is the first of a series of very fine articles for SDR for masses:

http://www.arrl.org/files/file/Technology/tis/info/pdf/020708qex013.pdf

Thanks a lot Matthias Wilhelm and Sohail ahmeda...

Perhaps this can be helpful as an introduction to practical (and very cheap) SDR:

http://sdr.osmocom.org/trac/wiki/rtl-sdr . The hardware is an USB TV dongle with a Realtek RTL2832U chip.

Thanks Enrique Garcia, its really nice and very cheap...

But in addition to this I am trying to do completely software based experiment , with no hardware use...

Karimkhan, you will never be able to interface to the radio environment without some hardware. At the bare minimum you'd need an antenna and an amplifier. More likely, you'll want some sort of RF to IF/Baseband downconversion/upconversion as well. How else will you send/receive signals? If you want to do everything in software, then you are talking about simulation instead of implementation.

The "software-defined" in "software-defined radio" means that the hardware is minimal and most of the system configuration, algorithms, and processing that define the radio are done in software... The link above says that it gives you access to "the raw I/Q samples" which is exactly what you would need to experiment using your PC as an SDR receiver. On the down side, it is narrowband (maximum of 3.2 MS/s) and available at a coarse quantization (8-bits according to the website). Also, you'd need a transmitter if you were looking to prototype a communication system, but for use as a software-defined radio sensor, it may be promising.

No matter what you choose, if you want to experiment with SDR using actual radio signals, you will need additional hardware. If you want to experiment via simulation (a perfectly acceptable thing to do), then you won't.