A method of processing signals in a Global Positioning System (GPS) receiver has been invented to enable the receiver to recover some of the information that is otherwise lost when GPS signals are encrypted at the transmitters. The need for this method arises because, at the option of the military, precision GPS code (P-code) is sometimes encrypted by a secret binary code, denoted the A code. Authorized users can recover the full signal with knowledge of the A-code. However, even in the absence of knowledge of the A-code, one can track the encrypted signal by use of an estimate of the A-code. The present invention is a method of making and using such an estimate. In comparison with prior such methods, this method makes it possible to recover more of the lost information and obtain greater accuracy.
The limitation on space available for this article precludes a description of the prior methods. However, a description of pertinent generally applicable aspects of GPS signals and signal processing is presented in the next three paragraphs because it is prerequisite to a meaningful summary of the present method.
Each GPS satellite transmits two L-band signals, denoted L1 (at a carrier frequency of 1.57542 GHz) and L2 (at a carrier frequency of 1.2276 GHz). The L1 carrier is phase-modulated with two binary pseudorandom-noise codes that contain GPS information: (1) the coarse-acquisition (C/A) code, characterized by a chip rate of 1.023 MHz and (2) the precise (P) code, characterized by a chip rate of 10.23 MHz and modulated in quadrature with the C/A code. The L2 carrier is modulated with the P code only. The signals from different satellites are distinguishable from each other because each satellite transmits a unique C/A and a unique P code. Although the limitation on space also precludes a detailed description of the C/A and P codes, it can be said here that names of these codes convey an approximate idea of the roles played by these codes and of the relationship between them. The C/A and P codes of all the satellites are further modulated with a common binary code that conveys information about the satellites, their orbits, their clock offsets, and their operational statuses.
The basic principle of GPS receiver signal processing is to determine the time and the position of the GPS receiver from times of arrival of signals transmitted from several different GPS satellites. This basic principle is implemented, in practice, by use of correlations between (1) the received GPS signals and (2) model signals in the receiver constructed from model carriers modulated by model C/A, and P (and, when applicable, A) codes.
Processing is said to be done in a code mode when the receiver "knows" the code in question. Because the C/A code is not encrypted, C/A modulation is usually processed in the code mode, using the published C/A code. Processing is said to be done in an encryption mode when the receiver does not "know" the code in question. More specifically, processing is said to be done in an encryption mode when the receiver does not "know" the A code with which the P code is modulated. Hence, the present invention is characterized as a method of encryption-mode processing.
The figure is a block diagram of signal processing according to the invention. The received GPS signal is down-converted from radio frequency (RF) to baseband to obtain two pairs of quadrature components — one pair for L1, the other for L2. Unlike in some prior methods, there is no cross-processing of signals between the L1 and L2 P channels. Instead, each of the two quadrature components obtained from each RF signal, independently of the other components, is counter rotated with its respective model phase, correlated with its respective model P code, and then successively summed and dumped over pre-sum intervals substantially coincident with chips of the respective encryption code. In the encryption mode, the effect of the unknown A-code sign flips is reduced, for each quadrature component of each RF signal, by combining selected pre-sums. The resulting combined pre-sums are then summed and dumped over longer intervals and further processed to extract the amplitude, phase, and delay for each RF signal. The precision of the resulting phase and delay values is approximately four times better than that obtained from conventional cross-correlation of the L1 and L2 P signals.