GPS has found its way into many applications, mainly as a result of its accuracy, global availability, and cost-effectiveness. Unfortunately, however, there exist some situations in which part of the GPS signal may be obstructed to the extent that the GPS receiver may not "see" enough satellites for positioning. Examples of those situations are positioning in urban canyons and deep open-pit mining. This signal-obstruction problem, however, was successfully overcome by integrating GPS with other positioning systems. In fact, reported results showed that the performance of the integrated system is better than either system alone. Augmenting GPS is not limited to sensor integration. As shown below, GPS can be augmented with computer-based tools, such as GIS, for efficient data collection and analysis.
GPS/GIS integration
A geographic information system (GIS) is a computer-based tool capable of acquiring, storing, manipulating, analyzing, and displaying spatially referenced data [1]. Spatially referenced data is data that is identified according to its geographic location (e.g., features such as streets, light poles, and fire hydrants are linked by geography).
Spatial, or geographic, data can be obtained from a variety of sources such as existing maps, satellite imagery, and GPS. Once the information is collected, a GIS stores it as a collection of layers in the GIS database (see Figure 9.1). The GIS can then be used to analyze the information and decisions can be made efficiently. (For example, the decision to build a new road can be made by studying the effect of one feature, such as traffic volume.)
GPS is used to collect the GIS field data efficiently and accurately [2]. With GPS, the data is collected in a digital format in either real-time or postprocessed mode. A number of GPS/GIS systems that provide centimeter- to meter-level accuracy are now available on the market. Most of these systems allow the user to enter user-defined attributes for each feature. Built-in navigation functions to relocate field assets are also available. Pen computer-based systems are used by some GPS receiver manufacturers to allow the data to be edited and displayed as it is collected [2].
Many industries, including utilities management, forestry, agriculture, public safety, and fleet management, can benefit from integrated GPS/GIS systems.
GPS/LRF integration
In areas with heavy tree canopy, GPS receivers will normally lose lock to the GPS satellites. In addition, real-time differential GPS corrections may not be received as well. To overcome these problems, integrated GPS/handheld laser units, or laser range finders (LRFs), were developed [3]. The way the integrated system operates is to set up the GPS antenna in a nearby open area, which allows the GPS system to operate normally without losing lock to the GPS satellites. With the help of a digital compass, a reflectorless handheld laser, colocated with the GPS receiver, can be used to determine the distance and azimuth to the inaccessible points (see Figure 9.2). This operation is commonly known as the offset function. Software residing in the handheld computer helps in collecting both the offset data and the GPS data. At a later time, all the available information is processed using PC software to determine the coordinates ofthe inaccessible points. Collecting and processing the data may also be done in real time, while in the field, provided that the real-time DGPS corrections can be received. Once the processing is done, the user can export the output to the required GIS or CAD software. This eliminates the need to place the GPS antenna directly on the features to be mapped [2].
Figure 9.1 GPS/GIS integration.
Figure 9.2 GPS/LRF integration.
GPS/laser integration is an attractive tool, especially for the forestry industry. Tree offsets, heights, and diameters can be measured easily with the laser unit. From a single location, a stationary user in a relatively open area can offset any number of points or features. In this case, the user location will be determined precisely by averaging all the GPS data collected while taking the offset measurements. Other applications of the GPS/laser integration include mapping points under bridges, mapping points on a busy roadway, mapping highway signs, and mapping shore lines, to name a few. GPS/laser integration can be used to map point features, line features, or area features.
GPS/dead reckoning integration
Another system that has been used to supplement GPS under poor signal reception is the dead reckoning (DR) system. Dead reckoning is a low-cost system, commonly comprising an odometer sensor and a vibration gyroscope. The integrated GPS/DR system is widely used in automatic vehicle location (AVL) applications [4].
DR navigation requires that the vehicle travel-distance and direction (heading) be available on a continuous basis. The travel-distance information is obtained from the odometer sensor, while the direction information is obtained from the gyroscope. If the vehicle starts the trip from a known location, the distance and direction information can be used to determine the vehicle location at any time. In other words, assuming that the vehicle is traveling in a horizontal plane, the travel and direction information can be integrated over time to compute the vehicle location (position).
Odometer sensors are already installed in all vehicles, mainly to evaluate their age and whether a service is required. An odometer sensor counts the number of revolutions of the vehicle’s wheels, which can be converted to a travel distance through an initial calibration. This conversion is known as the odometer scale-factor determination. One way of determining the scale factor is by driving the vehicle over a known distance. Unfortunately, however, the odometer scale factor changes over time due mainly to wheel slipping and skidding, tire pressure variation, tire wear, and vehicle speed. If left uncompensated, the scale-factor error will accumulate rapidly, causing significant positional error [5].
Vibration gyroscopes, however, are low-cost sensors that measure the angular rate (heading rate) based on the so-called Coriolis acceleration. A vibration gyro outputs a voltage that is proportional to the angular velocity of the vehicle. The vehicle’s heading rate is obtained by multiplying the output voltage by a scale factor. Similar to the odometer sensors, gyroscopes suffer from error accumulation due to gyro bias and scale-factor instability. A gyro bias is a temperature-sensitive variable error that affects the gyro measurements at all times. As such, a gyro will read a nonzero value even if the angular velocity is zero. It is observable when the vehicle is stationary or when it is moving in a straight line. Gyro scale-factor error, however, affects the gyro measurements only when the vehicle is taking a turn. This error could be greatly reduced by taking equal clockwise and counterclockwise rotations [4].
It can be seen that each of the GPS and DR systems suffers from limitations. While the GPS signal may not be available in obstructed areas, the DR system drifts over time causing large positional error. This suggests that an optimal positioning solution may be developed, based on the two positioning systems. Kalman filtering technique is commonly used for system integration [5]. With the integrated system, GPS helps in controlling the drift of the DR components through frequent calibration, while DR becomes the main positioning system during the GPS outages. As such, the performance of the integrated system will be better than either system alone.
Currently, a promising new inertial navigation technology, microelec-tro mechanical system (MEMS) technology, is under development. MEMS technology will be used to provide the heading and the traveled distance of the vehicle, replacing the traditional DR system. MEMS-based gyroscopes and accelerometers are expected to overcome the size and the cost of the current technology [6].
GPS/INS integration
There exist a number of applications that require high-accuracy positioning in obstructed areas and/or under high dynamic conditions. Examples of these applications are deep open-pit mining and airborne mapping.As discussed earlier, a major problem with GPS is its limitation when used in obstructed areas. In addition, a GPS receiver has limited dynamic capabilities. As mentioned in Section 2.7, GPS signal obstruction and high receiver dynamics can cause temporary signal losses, or cycle slips. To overcome these limitations, GPS can be integrated with a relatively environment-independent system, the inertial navigation system (INS).
An INS is a system that, once initialized (by acquiring the initial position, velocity, and orientation information), becomes an autonomous navigation system providing 3-D position, velocity, and attitude information [7]. An inertial sensor, also known as the inertial measurement unit (IMU), is a device consisting of accelerometers, gyroscopes, other electronics components, and a computer. When mounted on a moving object, the accelerometers measure the object’s acceleration plus the gravitational force, while the gyroscopes provide information on the orientation of the inertial platform. These sets of information are accumulated by the sensor’s computer to produce the velocity and position information. In addition to being a relatively environment-independent system, an inertial system provides accuracy as high as that of GPS for the short period of time following the initialization [7]. Moreover, inertial systems provide very high update rates compared with GPS. A major drawback of the inertial system, however, is that it suffers from drift if left unaided for a long period of time. In particular, the performance of the gyroscopes limits the overall performance of the inertial system.
Integrating GPS and INS overcomes the limitations of both systems [7]. In fact, GPS and INS complement each other. While GPS provides the initialization and the calibration to the inertial system, the latter bridges the GPS gaps when the satellite signal is blocked or temporarily lost. GPS/INS integration is commonly done in either of two modes, namely, loose coupling or tight coupling mechanisms. Loosely coupled integration is carried out in the solution domain, while tightly coupled integration is carried out in the raw measurements domain. In addition, tightly coupled integration requires extensive computations as compared with loosely coupled integration. It results, however, in a nearly optimal integration solution. Similar to the GPS/DR, the Kalman filtering technique is commonly used for GPS/INS integration [5].
GPS/pseudolite integration
One of the fastest growing applications of GPS is open-pit mining. The use of GPS in open-pit mining can remarkably reduce the cost of various mining operations. The availability of real-time GPS positioning at centimeter-level accuracy has attracted the attention of the mining industry. This is mainly because accurate real-time positioning is a key component that leads to automating the heavy and expensive mining machines. As such, smart mining systems can be developed that not only increase mining safety but also reduce costly labor [8].
Unfortunately, similar to the earlier cases, the satellite signal will be partially blocked as the pit deepens (see Figure 9.3). As such, in deep open-pit mining, GPS alone cannot be used reliably for mining positioning. One promising system that can augment GPS to ensure high-accuracy positioning at all times is the pseudolite (short for pseudosatellite) system. A pseudolite is a ground-based electronic device that transmits a GPS-like signal (code, carrier frequency, and data message), which can be acquired by a GPS receiver. Unlike GPS, which uses atomic clocks onboard the satellites, pseudolites typically use low-cost crystal clocks to generate the signal [9].
The addition of pseudolite signals improves both system availability and geometry. The number and locations of the pseudolites can be optimized to ensure the best performance of the system.
Figure 9.3 GPS/pseudolite integration.
The vertical dilution of precision, in particular, can be improved dramatically, which leads to improved accuracy for the height component. Another advantage of using the pseudolites is that, being ground-based transmitters, their signals are not affected by the ionosphere. Pseudolites, however, suffer from a number of drawbacks that must be overcome to ensure high-accuracy positioning. The first is known as the near-far problem, which results from the variation in the received pseudolite signal power as the receiver-pseudolite distance changes. The closer the receiver to the pseudolite transmitter, the higher the signal power, and vice versa. This problem does not exist with GPS-only positioning, as the received GPS signal power remains almost constant, because the satellite-receiver distance does not change significantly. Consequently, in GPS/pseudolite integration, if the pseudo-lite signal is much stronger than the other pseudolite and GPS signals, it may overwhelm the other signals and jam the receiver. This is what is known as the near problem. However, if the pseudolite signal is much weaker, the receiver may not be able to track it, which is known as the far problem. Transmitting the pseudolite signal in short pulses with a low duty cycle may, however, minimize the effect of the near-far problem [9].
The use of inaccurate clocks to generate the pseudolite signal causes synchronization error in the sampling time. This error will cause a range error, even if double differences are formed. A possible solution to this problem is through the use of a content-free data message of a master pseu-dolite. Another problem that requires the pseudolite user’s attention is the multipath error. Pseudolite multipath error occurs as a result of reflected signals from objects surrounding the antennas of both the receiver and the transmitter. Some researchers have suggested the use of patterned antennas as a feasible way of reducing the multipath effect. Unlike GPS-only positioning where ephemeris errors do not affect the position solution significantly, errors in the pseudolite coordinates will be propagated into the solution, causing large positioning errors. This is caused by the relatively short receiver-pseudolite separation [10]. Careful calibration of the pseu-dolite antenna location solves this problem.
It should be pointed out that the application of the integrated GPS/ pseudolite system in not limited to deep open-pit mining. Such an integrated system has been successfully used in precise aircraft landing, deformation monitoring, and other applications. Being similar in principle to GPS, pseudolite-only positioning has the potential of being the system of the future for indoor applications, such as underground mining (see Figure 9.3). A challenging problem to overcome, however, is the pseudolite location problem.
GPS/cellular integration
Cellular communication technology is becoming widely accepted throughout the world. Both the number of subscribers and the cellular coverage areas are increasing continuously. In addition, more advanced digital cellular coverage is on the rise, allowing voice and data to be mixed seamlessly. This makes the cellular system very attractive to a number of markets, including emergency 911, AVL, and RTK GPS.
A major limitation with the current cellular system, however, is its ability to precisely determine where a call was originated [11]. Although this limitation is not critical for applications like RTK GPS, it is of utmost importance for other applications such as emergency 911 and AVL. In the United States, for example, about one-third of all emergency 911 calls come from cellular phones. Of these, nearly one-fourth cannot describe their location precisely, which makes it very difficult for an operator to effectively send out assistance. As such, the U.S. Federal Communications Commission (FCC) has made it mandatory that, as of October 2001, wireless emergency 911 callers must be located with an accuracy of 125m (67% probability level) or better [11].
To meet the FCC location requirement, wireless network operators can either use the network-based location or the handset-based location. Most network-based caller location systems employ either the time-difference of arrival (TDOA) approach or the angle of arrival (AOA) approach to determine the caller’s location. The former measures the differences in the arrival times of an emergency 911 signal at the cell sites or base stations. The caller’s location can be determined if the signal is received at a minimum of three base stations. Obviously, time synchronization is essential with this technique, which can be ensured by equipping each cell site with a GPS timing receiver. The second technique, the AOA, uses phased-array antennas to compute the angles at which the signal arrives at the base stations. A minimum of two sites is required to compute the caller’s location with this method. As both the TDOA and AOA methods have advantages and drawbacks, some network operators combine the two methods [11].
Handset-based location technology integrates GPS with cellular communication through the installation of a GPS chipset in the handset of the wireless phone. With selective availability being turned off permanently, this technology would locate the wireless emergency 911 callers with an accuracy that exceeds the FCC requirement by a factor of ten. Unlike network-based technology, handset-based location technology is very simple to implement and does not require the installation of additional equipment at the base stations (e.g., GPS timing receivers). One of the drawbacks of the handset-based location technology, however, is that only new cellular phones can be equipped with GPS. In addition, the GPS signal is very weak to be received inside buildings. This limitation, however, could be efficiently overcome in the near future with the development of integrated GPS/MEMS technology, described in Section 9.3.
In the near future, the development of a new generation of cellular technology, the 3G wideband digital networks, will be completed. The 3G cellular technology supports voice, high-speed data, and multimedia applications. In addition, this technology uses common global standards, which not only reduces the operational cost but also makes the system useable worldwide. Moreover, with this new technology, devices can be turned on all the time for data transmission, as subscribers pay for the packets of data they receive/transmit.
The advances in the wireless communication and caller’s location technologies discussed earlier will greatly impact a number of industries. The vehicle navigation market, for example, is expected to greatly benefit from the advances in wireless communication, location, and Internet technologies (see Section 10.11 for details about vehicle navigation). Currently, vehicles use complex systems that integrate location technology with in-car computer navigation systems containing electronic digital road maps and other related information. Clearly, the in-car system will not be aware of any real-world changes in the navigation system’s database (e.g., a change in the traffic direction). With the availability of wireless Internet service, however, an up-to-date database residing at a central location could be accessed by drivers, eliminating the need for a complex in-car computer navigation system. Furthermore, with the availability of a precise location system, drivers could customize the information they need according to their locations, such as turn-by-turn navigation, traffic information, and local weather conditions. This method is simple, cost-effective, and flexible, and has the potential of being the way of the future.
