Transmission Security Versus Message Security

Message security prevents an enemy from accessing the information carried in a signal by use of encryption. High-quality encryption requires that the signal be digital and adds a pseudo-random bit stream to the signal bit stream as shown in Figure 2.19. For clarity in this discussion, let us call this the encrypting signal. The summed bit stream is itself pseudo-random and makes the message nonrecoverable. In commercial applications, it is often acceptable to use an encrypting signal that repeats after as few as 64 to 256 bits. However, in secure military encryption, the encrypting signal may not repeat for years. (The shorter the encrypting bit stream, the easier it is for an enemy to crack the code.) At the receiver, the original encrypting bit stream is added to the received bit stream, which returns the signal to its original, nonencrypted form.

Figure 2.19 Message security is achieved by adding a pseudo-random bit stream to a digitized input message.

However, transmission security involves spreading the spectrum of the transmitted signal in some pseudo-random way that makes it very difficult for an enemy to detect the signal, jam the signal, or locate the transmitter. The three ways to spread the signal are frequency hopping, chirp, and direct sequence spread spectrum. They are discussed (in the context of jamming) in Chapter 5. Here we will consider these techniques from a transmission security point of view. Although there are other operational benefits, the principal benefit of transmission security is to prevent an enemy from locating the transmitter and thus being able to fire on it or use a homing weapon against it. As shown in Figure 2.20, it is most important to provide transmission security for links from high value assets to lower value assets.

A frequency hopped signal switches its full power to a different frequency every few milliseconds (for slow hoppers) or microseconds (for fast hoppers) as shown in Figure 2.21. This makes it fairly easy to detect the presence of the signal, and there are many systems that can sweep for random intercepts that allow the transmitter to be located. This is particularly true of slow hoppers. Thus, frequency hopping is the least desirable technique for protecting the transmitter location.

Chirped signals which employ a wide linear sweep move across a wide frequency range very quickly (see Figure 2.22). Like the frequency hopper, the chirped signal moves its whole signal power to one frequency at a time. However, because it tunes so quickly, a receiver cannot detect the signal unless it has a fairly wide bandwidth. The wide receiver bandwidth reduces receiver sensitivity, but the chirp signal is still fairly easy to detect. Thus, geolocation of the transmitter is fairly straightforward.

frequency range by adding a secondary digital modulation with a high rate pseudo-random bit stream as shown in Figure 2.23. Note that the bits in the high rate digitization are called chips. The frequency spectrum of a digital signal was described in Section 2.4. The null- to-null bandwidth of the input information signal is twice the bit rate while that of the spread signal is twice the chip rate. The power in the signal is distributed across this much wider spectrum. This creates a noise-like signal that literally has its energy spread across a wide frequency range in real time. Without ever receiving full power at a single frequency, it is much more difficult to determine that a signal is present. Detecting this signal requires either energy detection or very sophisticated processing to time-collapse the high rate bit stream chips to form a narrow frequency determinant. Thus, this technique is the favored approach to providing transmission security. As discussed next, the wider the signal is spread, the greater the transmission security.

Figure 2.20 It is desirable to provide a higher level of transmission security on links from higher value assets.

Figure 2.21 A frequency hopping signal moves its full transmit power to a new frequency many times during a

Figure 2.22 A chirped signal sweeps its full transmit power over a large frequency range very rapidly.

Figure 2.23 Direct sequence spread spectrum modulation spreads the signal over a wide frequency range, reducing its

power at any single frequency.

It is important to realize that transmission security techniques do not provide dependable message security. Under normal circumstances, each of the spreading techniques used will make it difficult for an enemy to recover transmitted information. However, for each technique, there are conditions under which a sophisticated enemy can read the content of the message without despreading the signal. These circumstances involve short-range receivers or the use of highly sensitive receivers and sophisticated signal processing.

2.7.1 Transmission Security Versus Transmission Bandwidth

The SNR in a receiver is inversely proportional to the system bandwidth. This means that the ability of a receiver to detect a spread spectrum signal is degraded by the amount that the signal is spread. Without transmission security, a signal can be received in a bandwidth matched to the basic information modulation. However, if a signal is spread by (for example) a factor of 1,000, the receiver bandwidth must by 1,000 times as wide to capture the full signal power as shown in Figure 2.24. This causes a reduction in receiver sensitivity of 30 dB: {10 log₁₀[bandwidth factor]}. This loss of receiver sensitivity has a fairly linear relationship to the accuracy with which the direction of arrival of a signal can be determined. We need to be a little careful with this generality, because there are processing gains associated with various emitter location approaches that depend on the specifics of signal modulations. However, the general rule remains true: the level of transmission security is a direct function of the factor by which the signal is spread.

2.7.2 Bandwidth Limitations

Now let us consider how much spreading can be applied to a signal. That depends on the bandwidth of the unspread signal. A narrowband transmitter, such as that in a command link, may be only a few kilohertz wide. For example, the command signal might have 10,000 bits per second. Depending on the modulation used, the command link bandwidth might be 10 kHz. With a spreading factor of 1,000, the command link is still only 10 MHz wide. However, a real-time digital imagery data link might be 50 MHz wide. Even if video compression can be used, it will probably still be about 2 MHz wide. If you spread this by a factor of 1,000, the resulting signal would be 2 GHz wide.

Figure 2.24 Spreading the spectrum of a signal reduces its detectability and the ability to geolocate the transmitter

proportionally to the spreading factor.

Not only is the required transmitter power proportion to the link bandwidth, but amplifiers and antennas start to lose significant efficiency when they approach 10% bandwidth. The 10% bandwidth at 5 GHz is 500 MHz. Note that microwave links (e.g., at 5 GHz) usually require directional antennas to achieve good performance. Highly mobile tactical platforms connect much more easily with links using nondirectional antennas. This makes links in the UHF frequency range (perhaps 500 to 1,000 MHz) much more

desirable. The 10% bandwidth is only 50 MHz for 500 MHz links. The point is that it is difficult to provide a high degree of transmission security to a high data rate link. The higher rate link will need to have a lower spreading ratio to fit within the practical link bandwidth.