Dr. Riki Morikawa, George Mason University
 

Spread Spectrum

Back to Index

What is Spread Spectrum?

There are three types of spread spectrum techniques:  (1) direct sequence spread spectrum (DSSS), (2) frequency hopping (FHSS) and (3) time hopping (THSS).  In each technique, a psuedorandom code (PN code) that is known and shared between the transmitter and receiver, is used to spread the frequency bandwidth of the signal itself.  I will go over DSSS, but the same basic ideas apply to FHSS and THSS.

To begin, we need to recognize that there is a relationship between frequency bandwidth in Hz, and data rate capacity in bps.  Increasing frequency bandwidth enables us to increase data throughput and vice versa.  We also know that by increasing transmit power, we can increase data throughput.  The Shannon-Hartley[1] formula shows the relationship between capacity (bps), frequency bandwidth and signal-to-noise ratio (SNR):

C (bps) = BW_frequency x log₂(1+SNR)  (1)

Let’s first consider the case of a non-spread narrowband signal.  If an RF interferor (e.g., RFI from other transmitters, electrical equipment, etc.) overalps our narrowband signal, then our SNR can be dramatically reduced, possibly to the point that no signal is received at our receiver.  With a reduced SNR, capacity is also reduced or eliminated.

During WWII, it was recognized that a jammer could easily stop narrowband communications, therefore a solution was needed that would enable communications to continue despite attempts at jamming the signal by an adversary.   Thie is where spread spetrum techniques got their start.

The basic idea is to spread the signal across a much wider frequency bandwidth (i.e., one greater than the bandwidth of a jammer or RF interferor).  That way, enough of the signal is allowed to pass through.   So how do we spread the frequency bandwidth of the signal?

Consider digital communications, and the capacity formula shown in Shannon-Hartley.  From this equation, we see that if we increase C, then BW also increases provided we keep SNR at the same value[2].  To increase the capacity, we introduce the PN code, which is used to interleave bits into our original data rate (i.e., information rate).  This has the effect of increasing the bits/second, and thus the frequency bandwidth of the signal.  Not all of the bits generated are our signal, therefore, we call this “chip rate” vice information data rate.  The process of adding bits via PN code is called “spreading”.

So now we transmit more bits at a certain”chip rate”, resulting in a much wider frequency bandwidth, which can be very close, if not below, the noise floor (i.e., remember that we’re using the same amount of power, but within certain limits, this is not a problem).   Now if a jammer or RFI exists, it only effects a small portion of our signal, not the entire signal as in the narrowband case.

Upon reaching the receiver, a despreading process using the shared PN code occurs.  The depreading process recovers the original signal, and essentially minimizes the RFI.

Another major advantage of DSSS, is that a receiver will only receive data according to the PN code assignment.  That means that other spread signals can coexist in the same frequency bandwidth, as long as their PN code are different enough (i.e., they must meet an orthoganlity requirement).  Stacking several calls together increases the efficiency of the used bandwidth.  This is a major reason why cellular providers have gone towards spread access technologies such as CDMA, CDMA2000 and WCDMA.