For those not familiar with the concept of frequency sweeping to enhance spatial resolution, the following is a good example of the CHIRP technique:-
STARFISH CHIRP SIGNAL PROCESSING
CHIRP (Compressed High Intensity Radar Pulse) techniques have been used for a number of years above the water in many commercial and military RADAR systems. The techniques used to create an electromagnetic CHIRP pulse have now been modified and adapted for acoustic imaging sonar systems. StarFish uses CHIRP techniques at the core of its acoustic engine.
To understand the benefits of using CHIRP acoustic techniques, we need to analyse the limitations using conventional single frequency (monotonic) techniques.
An acoustic pulse consists of an on / off switch modulating the amplitude of a single carrier frequency. The figure below shows how this relationship exists between the transmitted signal and the output produced by the receiver circuitry in the sonar. It can be seen that the receiver does not decode each cycle of the transmitted pulse, but instead produces the 'envelope' of its overall amplitude...
The ability of monotonic acoustic systems to resolve targets is better if the pulse duration is short; this, however, has its drawbacks. Ideally, we need long transmit pulses to get enough acoustic energy into the water for good identification of the furthermost targets, but due to the velocity of sound (VOS) through water (typically around 1500 metres / second), each pulse will occupy an equivalent 'distance' related to its pulse duration - this is referred to as 'range resolution', and can be given by the following equation...
Range Resolution = (Pulse Duration x Velocity of Sound) / 2
EXAMPLE...
In conventional monotonic side scan sonar systems the pulse duration is typically 100 micro seconds, and combining this with the typical VOS of 1500 metres / second, a range resolution of 75mm is obtained.
The 'range resolution' effectively determines the ability of the sonar to identify separate targets; so, using the example above, if two targets are less than 75mm apart then they cannot be distinguished from each other. The net effect is that the system will display a single large 'combined' target, rather than multiple smaller targets, and any fine sonar detail is lost.
CHIRP SIGNAL PROCESSING OVERCOMES THESE LIMITATIONS!
Instead of using a pulse of a single carrier frequency, the frequency within the burst is changed (swept) through the duration of the transmission, from one frequency to another. For example, at the start of the transmission the sonar may operate at 100KHz, and at the end, it may have reached 150KHz - the difference between the starting and ending frequency is known as the 'Bandwidth' of the transmission, and typically the centre frequency of the transmission is used to identify the sonar (in this case it would be a 125KHz sonar).
By constantly changing its frequency over time, this 'chirped' transmission can be thought of as having a unique acoustic signature, and so if two pulses now overlap (as the targets are closer than the range resolution), we can use the known 'frequency versus time' information to tell them apart.
With modern high-speed digital-signal-processing (DSP) techniques, the StarFish sonar receiver contains a 'pattern-matching' circuit that looks for its transmitted 'CHIRP' being echoed back from targets, and its receiver now produces a sharp 'spike' when a good match is found (whereas the monotonic sonar produces an output the same duration as its transmit pulse)...
This means that the critical factor in determining range resolution is no longer the pulse duration, but the bandwidth of the CHIRP, so the range resolution can be found by...
Range Resolution = (Velocity of Sound) / (Bandwidth x 2)
EXAMPLE...
The bandwidth of the StarFish CHIRP system is typically 40kHz, and using the same VOS of 1500 metres / second, our new range resolution is 18.75mm... a theoretical improvement by a factor of 4 over the monotonic example above!
The figure above shows that on a chirped sonar, when two acoustic echoes overlap, the CHIRP pulses do not merge into a single acoustic return (as their frequency is different from each other at the overlapping points), and the sonar is able to resolve and display the two targets independently.
Therefore, we now can have longer transmissions (and see targets further away) without a loss in resolution; and additionally, CHIRP signal processing techniques offer improvements in background noise rejection (as the side scan sonar is only looking for a swept frequency echoes, removing random noise or out-of-band noise).
----------------------------------
The above data in its original form is at:-
CHIRP Signal Processing - StarFish Seabed Imaging System - Side Scan Sonar Acoustics
mm43