Q. What are the similarities and differences of an L1 vs. L1/L2 receiver? How do they differ from the C/A Code? - J.T. Camarillo, Calif.

A. John C. Bohlke, Sokkia Corp: The difference between an L1 and an L1/L2 receiver lies in the number of carrier frequencies they can track. An L1 receiver tracks only one of two carrier frequencies transmitted by the GPS satellites whereas an L1/L2 receiver tracks both carrier frequencies. Tracking two frequencies enables an L1/L2 receiver to achieve the same amount of accuracy as an L1 receiver but with less observation time and over longer baselines. Additionally, L1/L2 receivers have the capability of On-the-Fly (OTF) ambiguity resolution. A kinematic survey receiver with OTF ambiguity resolution allows the user to reinitialize the receiver without returning to a known point after losing lock.
      Most L1 receivers are capable of achieving an accuracy of 1cm + 1 part per million (ppm), whereas most L1/L2 receivers are capable of a 5mm + 1ppm accuracy. However, L1/L2 receivers typically cost much more than L1 receivers.
      C/A code receivers are generally less expensive than L1 receivers but can only achieve an accuracy of several meters. As opposed to various static and kinematic surveying methods available with L1 and L1/L2 receivers, C/A code receivers are only capable of post-processing data that was collected using the differential GPS method.

Carl Carter, Allen Osborne Associates: All receivers monitor GPS satellites by listening for the L1 signal modulated by the C/A code. The L1 signal is called the carrier frequency, and is transmitted at 1575.42 MHz. The C/A code is a digital bit sequence that is used to modulate the L1 signal at a bit rate of 1.023 MHz. In addition, the satellites broadcast another carrier called L2 at 1227.60 MHz. Both the L1 and L2 carrier signals are modulated with a bit pattern referred to as the P code, which has a bit rate of 10.23 MHz. This rate is exactly 10 times that of the C/A code, which is found only on the L1 carrier.
      Receivers referred to as C/A code receivers monitor only the L1 signal, and search for the C/A code pattern by generating a copy of the pattern and attempting to correlate it with the received signals. Since the signal transmitted by the GPS satellites takes some time to travel from the satellite to the receiver, the pattern received from any satellite will be delayed in time from when it was transmitted by an amount related to the distance between the satellite and receiver. The amount of time it is delayed is determined in the C/A code receiver by having the internally generated pattern shifted in time until a match is found, and then determining how that time shift relates to the receiver's internal clock. This time shift is called the pseudorange, since it is related to the actual range, but also contains several other elements including a factor due to the receiver's clock not exactly matching GPS time. The ability of a receiver to measure this time pre-cisely is related to the period of the pattern it is matching, and usually can be done down to some tenths of a percent, or from one to a few parts in 1000. Since the C/A code has a period of about 1 µsec, the C/A code receiver is generally able to resolve the time to some number of nanoseconds.
      Receivers which refer to both L1 and L2 monitor both carrier frequencies. Since signals traveling through the ionosphere tend to be altered as they travel, and the amount of alteration tends to vary with the frequency of the signal, a receiver that monitors both L1 and L2 is able to compute the amount of that alteration and correct for it. In the process, the receiver must somehow obtain a copy of the P code (currently the P code is encrypted, so what is seen is properly called the Y code). This copy of the code provides a means to refine the distance measurement by the factor of 10 available from the faster P code bit rate.
      One additional case should be noted. Receivers which are generally referred to as L1 receivers could be constructed to use the P code, but this is not a common case. P codes are generally used only by dual-frequency, or L1/L2 receivers.

Wendy Corcoran, NovAtel Communications: GPS satellites transmit data on two frequencies: L1 (1575.42 MHz) and L2 (1227.60 MHz). Modulated on this carrier frequencies is code information. On the L1 carrier, the C/A code and P code are modulated while on the L2 carrier only P code is modulated.
      In the majority of L1 GPS receivers, the data collected from the satellites is the L1 carrier and the C/A code. On the L1/L2 receivers, the information collected is the L1 carrier, L2 carrier, C/A code and in some cases the P code on one or both of the carrier frequencies. The only similarity between an L1 vs. L1/L2 receiver is both receivers track the L1 and C/A code.
      With L1/L2 receivers the RF section is doubled to track both frequencies and the number of channels are doubled to track the L1 signal of each satellite and the L2 signal of the same satellites. Also, the amount of memory has to be increased because the data collected has increased. This results in additional hardware requirements which in turn increase the cost of these receivers.
      In functionality, the L1/L2 receiver can compensate for ionospheric interference by comparing the L1 signal to the L2 as it travels through the ionosphere. This error increases in significance the longer the distance between the base and the mobile GPS receiver. For the L1 GPS receivers, the recommended distance is less than 25 km for surveying accuracy but, if L1 and L2 are measured, the distance can be beyond 25 km. Another functional advantage to an L1/L2 receiver is its ability to compute centimeter accuracies virtually instantaneously in real-time. Although certain L1 receivers are used for centimeter accuracies as well, the distance is restricted and there is a longer occupation time for real-time. For L1/L2 receivers, the two frequencies can be combined to form a longer wavelength frequency. This technique is called 'widelaning' and it can be used to rapidly determine the unknown number of cycles between the satellite and the GPS receiver. This unknown number of cycles or "ambiguity" is key to achieving centimeter results. With L1/L2 receivers these ambiguities can be computed within seconds.
     In order to obtain accuracies below 30 cm, the carrier frequencies have to be tracked and recorded. The C/A code GPS receivers are not able to compute accuracies like the carrier-based GPS receivers. The C/A code resolution coupled with Selective Availability, restricts the receiver's ability to compute accurate positions (50-100m). If the C/A code receiver is used differentially, the typical accuracy is 3-5m - although there is a group of C/A code receivers that can obtain accuracies of 0.5-1.5m. These C/A code receivers are referred to by various names, therefore you should check the accuracies quoted in the specific product literature.

Arthur Lange, Trimble Navigation: The GPS satellites broadcast both C/A and P(Y) modulation on the L1 frequency. Only the P(Y) code is broadcast on the L2 frequency. All GPS receivers must use the L1 C/A code. P(Y) code receivers use C/A code in order to synchronize with the lower power P(Y) code. Only military receivers can directly decode the Y code. The L2 frequency is used by survey grade receivers to measure the offset between the P(Y) codes on the L1 and L2 frequencies, and thus the differential delay caused by the ionosphere. Once the differential delay is computed, then a better model for the total delay of the ionosphere can be computed. Survey grade GPS receivers must use the better model for the ionosphere correction to obtain the best long-baseline survey accuracy.
      A GPS receiver that receives L2 must have additional circuitry above and beyond the L1 requirements. This additional L2 circuitry includes an antenna, amplifiers, mixer, and digital channel circuitry which essentially duplicates the required L1 circuitry.

Dr. Frank van Diggelen, Ashtech Inc.: L1 and L2 are the two frequencies at which each GPS satellite transmits signals. Both L1 and L1/L2 receivers track the C/A code (which happens to be on L1). L1/L2 receivers also track the P code which is transmitted on both frequencies, and is usually encrypted. Proprietary techniques are used to extract the P code from the encrypted signal.
      In non-differential mode, L1 and L1/L2 receivers will per-form similarly, providing horizontal positional accuracies of 100m (95 percent). In differential mode, L1/L2 receivers outperform L1 only receivers because they can remove the ionospheric errors, and because they can resolve ambiguities faster to achieve cen-timeter accuracy. The biggest difference in performance is the time required to resolve ambiguities and achieve centimeter accuracy. The comparable performance of professional systems is summarized in the following table.

Q. Why is there such a large variability in the data storage requirements of GPS systems? Some GPS systems require several hundred bytes to store a single GPS position. Other systems can store a single position in only a dozen or so bytes. If it's just a simple position, how can there be that much difference? - A.W. Miwaukee, Wis.

A. Bohlke.: When collecting just a simple GPS position, receivers may record the position in a binary or an ASCII format. A binary format uses less storage space than the ASCII format. In other cases, the data storage requirements of GPS rely on a number of variables. The length of observation and the data collection rate significantly affect the amount of memory necessary to store GPS data. The required amount of storage space may depend on the relationship between the number of channels and the number of available satellites. A greater number of channels and/or satellites increases the need for memory. Storage requirements also rely on whether the receiver employs various data compression techniques and whether the user enters attribute data associated with the position. Lastly, the receiver may use memory to store the observables (used for post-processing) in addition to storing the GPS position.

Carter: The large difference in storage is because in some receivers, it isn't just a simple position that is stored. In receivers designed simply for navigation, all that is stored is the location and time error that are the results of position computations. But many receivers are designed to collect raw data for later (post) processing, a method that can result in much more precise position determinations. These receivers generally store the observations made from the satellites rather than just the position computations. Observations consist of such elements as the measured range to a satellite computed from the C/A code, range computed from the P code, range computed from counting whole and fractional wavelengths of the carrier wave on both the L1 and L2 frequency carriers, and error estimates associated with each of these. Each of these values must be recorded for each satellite being tracked, and receivers may track up to 12 satellites simultaneously. In addition, in order to support processing at a later time, the ephemeris data (orbit definition) broadcast by each satellite must be recorded as well.

Corcoran: Data storage requirements change based on what the receiver measures and records. Raw data is going to consume a larger memory than position only. If we take the case of position only receivers, the variation may be due to what they record with the position. In some receivers, the latitude, longitude and height are recorded whereas in others, factors that affected the position may also be recorded such as DOP, number of satellites, specific SV numbers, residuals and position RMS. This information is important to evaluate the quality of the position you're getting. If position only is output, there are no indicators of how accurate the position is or what may have caused inaccuracies.

Lange: Some models of GPS/GIS data collectors use a data storage and post-processing differential correction scheme that only requires about 20 bytes per 3-D position. These data storage requirements are the same whether the position is real-time DGPS corrected or will be post-processed DGPS corrected. The low data storage requirement GPS/GIS data loggers are able to do this recording in so few bytes because of the way the differential correction post-processing software is organized to perform a differential correction on a computed GPS position.
      In other GPS/GIS data storage schemes, when the data will be post-process differentially corrected, the data file does not contain an uncorrected computed position, but rather the full pseudo-ranges for each of the satellites in view. Then in the post-processing differential correction software, the rover pseudo-ranges are first corrected and then a position is calculated. To store the uncorrected pseudo ranges may require over 200 bytes per 'position' compared to less than 20 bytes for the more efficient data storage scheme.
      Storing the data more efficiently may result in a ten-fold reduction in field storage requirement over the less efficient method. A side benefit of the more efficient method is that the same format is used in data collectors for both real-time and post-processed differential GPS, resulting in rover files that can have mixed real-time and post-processed positions with the post-processing software only correcting the rover positions that were not real-time corrected in the field.

van Diggelen: When it comes to storing a GPS position, the biggest factor is whether the system is set up for post-processing or real-time operation.
      If a system is designed for real-time differential operation, it requires a radio link to receive differential data from a base station. This data is processed in the GPS receiver at the remote site, the result of the processing is a single position which can be stored using a few bytes. However, for systems designed for post-processing, differential processing is done "back at the office" with data stored in both the base station receiver and the rover receiver. The GPS receiver must store all the GPS measurement data, as well as overhead information, such as satellite ephemerides. For the highest possible accuracy the measurement data includes six observables for each satellite: C/A code and carrier, P(L1) code and carrier, and P(L2) code and carrier measurements. The best receivers have 12 channels, so up to 72 measurements may be stored at each epoch, requiring several hundred bytes. Why bother with all this data? The answer is: you get what you pay for, and for the price of many observables you get high position accuracy.

About the Participants:
John C. Bohlke is GPS support manager at Sokkia Corp. in Overland Park, Kan. He may be reached at 913-492-4900 or 800-4-SOKKIA in the U.S. Carl Carter is a software systems engineer at Allen Osborne Associates in Westlake Village, Calif. He may be reached at 805-495-8420. Wendy Corcoran serves as manager, survey products at NovAtel Communications Ltd. in Calgary, Alberta, Canada. She may be reached at 403-295-4900. Arthur Lange is GIS product manager at Trimble Navigation in Sunnyvale, Calif. He may be reached at 408-481-2994. Dr. Frank van Diggelen is a marketing manager at Ashtech Inc. in Sunnyvale, Calif. He may be reached at 408-524-1508.

Back