LIKE MANY organisations, the military looks for ways to do more with less. Smaller and smarter munitions are becoming candidates for in-flight guidance, such as the plan to use GPS guidance in army artillery shells and in navy deck gun projectiles.

The GPS receiver would have to meet certain requirements for in-flight guidance including gun-hardening, fast signal acquisition and jammer tolerance, selective availability anti-spoofing module (SAASM), interfaces and operating modes to support initialisation of the GPS receiver, small size, low power consumption and low cost. Many of these requirements need research into how they would operate in specific circumstances but some criteria already established in other applications would apply. For example, a projectile-firing sequence illustrates features needed in a projectile GPS receiver.

The Army Excalibur 155mm projectile is guided in-flight. The rear fins de-spin the projectile after firing while the canards in front provide aerodynamic guidance control. The nosecone contains the guidance section that includes the GPS receiver/IMU. Initialisation sequences are well understood and the process would require minimal adaptation to serve the requirements suggested.

Acquisition On receiving a start acquisition command, a projectile GPS receiver initiates a direct Y-code search for satellite signals. When it has detected and acquired signals from at least four satellites, the receiver outputs pseudo-range and delta-range measurements, aligns with the IMU and begins to guide the projectile. As the projectile leaves the gun barrel, it deploys tail fins that de-spin it through pitch control. Subsequently it deploys, near the front, canards controlled by a drive guidance unit that make use of the navigation data derived by the GPS/IMU. GPS tracking continues until the munition dispenses or until jamming levels preclude GPS tracking. During test firings, the projectile transmits telemetry back to the ground monitor stations for performance evaluation.

Gun hardening A ball grid array mounting of digital devices with high input/output pin count, results in a dense, low-cost and very strong component mount. Proper layout and packaging will allow many receiver components to survive gun-fire shock.

Fast acquisition A projectile receives its kinetic energy from the gun charge. The projectile GPS receiver must track at least four GPS satellites for projectile guidance to commence. The earlier this occurs, the longer the interval available for guidance manoeuvres. Jamming levels typically will increase as the projectile approaches the target. Acquisition requires a much higher signal-to-noise ratio (SNR) than tracking, so it is important that acquisition occurs as early in the flight as possible.

Signal search The GPS signal search is a process in time and frequency. The time search adjusts the local code generator to match the arrival code-state of the satellite signal and the frequency search adjusts the frequency of the carrier numerically via a controlled oscillator to be at or near the carrier frequency of the satellite signal. This is necessary to provide a carrier-phase reference to phase-demodulate the Y-code from the satellite L1 carrier successfully. This is correlated with a local replica code that pre-positions the carrier loop for pull-in.

Minimising search time The acquisition search time is minimised by reducing the uncertainties to be searched and also by maximising the uncertainties searched by the receiver. The contribution of the receiver design to this is to minimise the frequency uncertainty of the receiver's reference oscillator.

Frequency uncertainty Major contributors to frequency uncertainty at acquisition are the projectile's velocity uncertainty and frequency uncertainty of the receiver's reference oscillator. Projectile velocity uncertainty is minimised by using information received during initialisation. A frequency calibration process for the receiver oscillator is part of the initialisation process to minimise its frequency uncertainty because the projectile may have been stored for years with no servicing.

Calibration is carried out during initialisation using the time intervals between PS time marks that also are used to transfer GPS time to the projectile. It is essential to consider oscillator drift during the hold period between initialisation and start of acquisition, and the oscillator frequency shift induced by the gunfire setback shock. The two main contributors to time uncertainty at acquisition are uncertainty of the projectile's position and of the receiver clock bias with respect to GPS time. Projectile-position uncertainty is kept to a minimum using methods as for projectile velocity. The major part of time uncertainty at gunfire is receiver-clock bias uncertainty.

Acquisition correlator The second part in minimising acquisition time is maximising the span of parallel time and frequency search embedded in the projectile's GPS receiver. We traded off our desire for large parallel time and frequency search spans with recurring cost and power-dissipation considerations.

Tightly coupled IMU IMU measurements are used to rate-aid the GPS carrier and code-tracking loops. This removes most dynamic stress from the loops, allowing them to operate in a narrower bandwidth than otherwise would be the case. This reduces the jammer and thermal noise power in the loop bandwidth, improving measurement accuracy and lowering the tracking drop-lock threshold.

The two main factors limiting the benefit of IMU aiding are the IMU measurement accuracy and the latency of the IMU data when it arrives at the tracking loop. The projectile's IMU is not powered until after gunfire. It receives no transfer alignment during initialisation because there is insufficient power available during the time-hold period to sustain the IMU and because its dynamic range is insufficient to track the 15kg acceleration of gunfire.

The IMU must depend on the GPS track to align it in flight. The IMU aiding becomes effective only after GPS has acquired and tracked over an interval sufficient to align the IMU. This makes early GPS acquisition very important. The latency is minimised by converting the raw inertial sensor data, changes in velocity and changes in direction to the satellite line-of-site range, range rate and range acceleration. The latency is thus held to less than 20 milliseconds. This is adequate because the projectile flight does not contain any sudden high-amplitude dynamic events that would demand a shorter latency interval.

The projectiles are tactical weapons that will be used on or near a battlefield. The projectile GPS receiver must be secure for this environment, so a selective availability/anti-spoof module (SAASM) is mandatory. All the digital GPS signal processing functions are embedded in the SAASM and inputs to the SAASM are the sampled L1 and L2 satellite signals, and outputs are the satellite measurement set.

The specific SAASM components are the GPS ASIC containing 12 tracking channels each capable of tracking L1 or L2, C/A or P/Y code, the acquisition correlator ASIC, a key data processor (KDP) chip set and flash memory for signal processing programme storage. The projectile nose- cone contains the initialisation coupler, GPS antennas, the IMU, a guidance computer, batteries and power regulators as well as the GPS receiver.

The projectile GPS receiver has been designed to address the needs of projectile fire sequence and guidance. Of course many of the features and the performance of this receiver are desirable in any GPS receiver intended for military applications. The basic chip-set, the SAASM and software that provide these capabilities can be used in a wide variety of applications.