Software Defined Radio—the Basic Architecture
In this article the basic concepts are outlined. The RF signal from the antenna goes through a bandpass filter and amplification. The resulting signal is mixed with a locally generated RF frequency to create an Inphase and a Quadrature phase (90 degrees shifted) signal. These signals go through a lowpass filter and then into an analog-to-digital convertor. The sampling rate is typically fixed so it must be set at a rate that can sample a sufficient number of samples to recreate the signal properly across the entire range.
For reception an FPGA or host computer processor takes the signals and applies signal processing to break the signal into its component bits and symbols or to find some other set of parameters. For transmission, the processor works to create the waveforms that go back through the chain to be transmitted.
The key to software defined radio adoption in today’s world is the standard PC is now sufficiently powerful enough to perform the waveform processing. Of course DSPs and FPGAs enhance the performance of the system, but still the cost of a PC-based solution is far cheaper than that of a dedicated, proprietary system which is commonly found in military applications such as the JTRS program.
The need for more spectrum raises the need for cognitive radio techniques. Cognitive radio techniques address this issue. To implement a cognitive radio system one needs the following components:
1. Location sensors—establish its position
2. Spectral monitoring—monitor the network for open channels
3. Control software—negotiate the use of the spectrum between users