Some Theoretical Observations on Spread-Spectrum Communications

In general, spread-spectrum communications refers to a class of modulation methods by which the information-bearing signal is transmitted via a modulated signal having much greater bandwidth. Two common methods are used to accomplish the spreading. In one method, direct-sequence modulation, the information signal is multiplied by a rapidly varying waveform. This waveform, which the receiver is required to know, may be thought of as having a pseudo-random character. For practical reasons, it has finite duration and is repeated in time. We refer to any particular suitable waveform, or a collection of 2013 such, as a code. The other method for spectrum spreading is frequency hopping. In this case, the available transmission band is divided into a large number of disjoint frequency intervals, and the information is conveyed by hopping from one such frequency to another. The information may be transmitted by phase-shift-keying each frequency or by using a particular set of frequencies for a particular symbol, etc. Again the code for frequency hopping must be known to the receiver. The initial motivation for introducing such a scheme appears to be in its military use as an anti-jamming device. The jammer, not knowing the transmitter's code for spectrum spreading, must thus blanket all codes. Most of the jammer's power is wasted in codes that are orthogonal to the one in actual use. An additional application of more commercial interest was introduced by Costas.1 His idea was to use spread spectrum as a way to make a large bandwidth, W, available as a communication resource to many potential users without preassigning frequency divided channels (FDM) (and thus overlimiting the number of potential users) and without having a dynamic assignment of FDM (thus incurring the need and cost of external control).