Coded orthogonal frequency division multiplexing. Orthogonal Frequency Division Multiplexing (OFDM) is a modulated multi-carrier transmission technique, which splits the available bandwidth into many narrow sub-band channels (typically 2000-8000). Each carrier is modulated by a low rate data stream. The modulation scheme can vary from a simple QPSK to a more complex 64-QAM (or other) depending on the required binary rate and the expected transmission robustness.

For those familiar with Frequency Division Multiple Access (FDMA), OFDM is similar. However, OFDM uses the spectrum much more efficiently by providing a closer packing of the sub-band channels. To achieve this, all the carriers are made orthogonal to one another. By providing for orthogonality of carriers, each carrier has a whole number of cycles over a given symbol period. By doing this, the occupied bandwidth of each carrier has a null at the center frequency of each of the other carriers in the system. This results in minimal interference between the carriers, allowing then to be spaced as close together as is possible. Each individual carrier of the OFDM signal has a narrow bandwidth (for example 1kHz), and the resulting symbol rate is low. This results in the signal having high immunity to multi-path delay spread, as the delay spread must be very long to cause significant inter-symbol interference (> 500 milliseconds).

Coded Orthogonal Frequency Division Multiplexing (COFDM) has the same principle as OFDM except that Forward Error Correction (FEC) is applied to the signal prior to transmission. This overcomes errors in the transmission as a result of lost carriers from multiple propagation effects including frequency selective fading and channel noise.

COFDM can transmit many streams of data simultaneously, each one occupying only a small portion of the total available bandwidth. This approach can have many advantages with proper implementation: 1. Because the bandwidth occupied by each sequence of symbols is relatively small, its duration in time is bigger. As a result, the immunity against multi-path echoes can be higher. 2. Frequency selective fades are spread over many carriers. Instead of completely destroying a number of adjacent symbols, many symbols are instead distorted only slightly. 3 By dividing available bandwidth in multiple narrow sub-bands, the frequency response over each of the individual sub-band channels is essentially flat even with steep multi-path induced fade. This can mean easier equalization requirements.

See also: DVB and the Engineering & Transmission chapter.