1.4: Radio Architecture

A radio device is comprised of several key reasonably well-defined units. By frequency there are baseband, intermediate frequency (IF), and RF partitions. In a typical device, the information—either transmitted or received, bits or analog waveforms—is fully contained in the baseband unit. In the case of digital radios, the digital information originating in the baseband digital signal processor (DSP) is converted to an analog waveform typically using a digital-to-analog converter (DAC). This architecture is shown for a simple transmitter in Figure 1.3.6. When the basic information is analog, say a voice signal in analog broadcast radios, the information is already a baseband analog waveform. This analog baseband signal can have frequency components that range from DC to many megahertz. However, the baseband signal can range from DC to gigahertz in the case of some radars and point-to-point links that operate at tens of gigahertz.

The RF hardware interfaces the external EM environment with the rest of the communication device. The information that is represented at baseband is translated to a higher-frequency signal that can more easily propagate over the air and for which antennas can be more easily built with manageable sizes. Thus the information content is generally contained in a narrow band of frequencies centered at the carrier frequency. The information content generally occupies a relatively small slice of the EM spectrum. The term “generally” is used as it is not strictly necessary that communication be confined to a narrow band: that is, narrow in percentage terms relative to the RF.

The trade-offs in the choice of carrier frequency are that lower-frequency EM signals require larger antennas, typically one-quarter to one-half wavelength long, but propagate over longer distances and tend to follow the curvature of the earth. AM broadcast radio stations operate around \(1\text{ MHz}\) (where the wavelength, \(\lambda\), is \(300\text{ m}\)) using transmit antennas that are \(100\text{ m}\) high or more, but good reception is possible at hundreds of kilometers from the transmitter. At higher frequencies, antennas can be smaller, a much larger amount of information can be transmitted with a fixed fractional bandwidth, and there is less congestion. An antenna at \(2\text{ GHz}\) (where the free-space wavelength \(\lambda_{0} = 15\text{ cm}\)) is around \(4\text{ cm}\) long (and smaller with a dielectric or when folded or coiled), which is a very convenient size for a hand-held communicator.

The concept of an IF is related to the almost universal architecture of transmitters in the 20th century when baseband signals were first translated, or heterodyned, to a band around an IF before a second translation to a higher RF. Initially the IF was just above the audible range and was known as the supersonic frequency. The same progression applies in reverse in a receiver where information carried at RF is first translated to IF before finally being converted to baseband. This architecture resulted in near-optimum noise performance and relatively simple hardware, particularly at RF, where components are much more expensive than at lower frequencies.

The above discussion is a broad description of how radios work. There are many qualifications, as there are many evolving architectures and significant rethinking of the way radios can operate. Architectures and basic properties of radios are trade-offs of the capabilities of technologies, signal processing capability, cost, market dynamics, and politics.