Position location techniques and applications - Part 1: Cellular networks

For the 3G standards, complete consensus did not exist in terms of technologies, and two standards were mainly established, both based on CDMA. Although there are other proposals for 3G such as OFDM and a hybrid version of TDMA and CDMA, the cdma2000 by the Telecommunications Industry Association (TIA) and the wideband CDMA (WCDMA) by the European Telecommunications Standard Institute (ETSI) are the most used at this moment.

The wideband CDMA bandwidth is 5 MHz to accommodate data rates such as 144 Kbps and 384 Kbps, and up to 2 Mbps with favorable conditions on wireless transmission. Another important characteristic is the ability to work with different data rates on a frame-by-frame basis by using the variable spreading factor, which is the ratio of the transmission bandwidth occupied and the input bit rate. An improvement of the traditional power control of cdmaOne was also included to work with increments of up to 1 dBm and faster on both the forward and reverse channels. This standard is the evolution of GSM/GPRS, providing all services from the first day of operation. WCDMA is not backward compatible with cdmaOne. The specifications of WCDMA are developed by the 3G Partnership Project (3GPP) consortium.

cdma2000 is the 3G standard from TIA, which is backward compatible with cdmaOne, and its specifications were developed by the 3G Partnership Project 2 (3GPP2) consortium. IS-2000, known as cdma200, has also evolved. At the beginning in 1999, it had a data rate of 316.8 Kbps; version A in 2000 included voice and data. In 2002, version C was developed to have a data rate of 3.1 Mbps for the forward link, and in 2004 version D had a reverse link with 1.8 Mbps. Also in 2004, IS-856-A was developed to enhance the 3.1 Mbps of the forward link.

3G technologies provide various services such as telephony, short messaging, emergency calls, Internet access, packet and circuit switching, and more. In terms of applications, earlier technologies work in 3G (e.g., the Wireless Application Protocol (WAP), Java, Bluetooth, Wireless Markup Language (WML), and so on). Applications such as voice, messaging, Internet access, and the current location-based services (LBS) are also available.

LBS determines the location of a customer and the information is used to provide services according to location. For example, a user's location information can be employed in a map to help find the way to a destination. In this case, location should be accurate in order to position the user on the map. Location of users is also important to service providers to develop various network areas with different service charges or a dynamic charging system that changes with time according to events. For example, user PL is important in determining channel allocation strategies in base stations (BSs) where attractors are established, since at certain times the attractors might represent traffic congestion spots.

It is also important that location information of a mobile provides marketing data to many organizations and also to the user trying to locate a specific type of business. Applications such as tracking can also be implemented as an LBS; only a handset is required to track a vehicle.

Regarding user location, the network has the user's location registration made by the MSC and the HLR or VLR, which only provide information on the location area and the BS at which the user is found. This simple method is based only on the network coverage area and could be more accurate by sectoring the cells. In general, a method such as the TDOA discussed in Chapters 2 and 3 can be used, where measurements of signal parameters form hyperbolas like those shown in Figure 5.5, and after computing geometric relationships of the hyperbolas, an estimate of the position can be obtained.

The basic architecture for any network is presented in Figure 5.6, where the signal from the mobile is received through the radio access network (RAN),and depending on the service (voice or data), a location center (LC) can determine location with the help of the MSC, HLR, or VLR using algorithms described in previous chapters.

Talking specifically about GSM, we must recall that it is a system based on TDMA with eight time slots per frequency channel, occupying a transmission bandwidth of 200 KHz. The time slots are grouped into frames and then in multiframes. Each of the multiframes carries logical channels that transmit control signals or information traffic. Within one of the time slots in a frame, a 26-bit training pseudorandom sequence can be used to estimate the channel by performing the autocorrelation of the received sequence and a locally generated one. This autocorrelation can be used in localization by considering that the maximum value of such autocorrelation indicates time reference.

The base stations also transmit a synchronization sequence in the synchronization logical channel. This sequence can also be used to estimate time reference by computation of the autocorrelation. Due to the use of GMSK modulation, the autocorrelation peak is approximately 4 bits wide, which is an accuracy limitation for a time-based method. This inaccuracy is worsened by the multipath effects of the channel. If the receiver is not prepared to mitigate these effects, the peak of the autocorrelation would be shifted from its original place depending on the effects over the amplitude, phase, and time delay of the received signal.

Signal propagation time could be estimated using the time advance (TA) feature used by the base station to indicate to the mobiles the amount of time that the transmission must advance in order to receive the mobile's signal in the corresponding time slot. TDOA can be estimated using the observed time difference (OTD) that is used to help the handover mechanism where each mobile unit monitors the base stations around it. AOA, although not very accurate, could be estimated using cell sectorization. All of these techniques help to determine the circles, hyperbolas, and so on, that could be used geometrically to estimate PL as discussed in Chapters 2 and 3.

The implementation of PL in GSM could be done considering mobile-based or network-based positioning. In mobile-based positioning, the mobile handset must be equipped so that it can receive signals from at least three base stations and compute TDOA to self-determine its position. In order to do this, the base stations must be modified so that the signals from the three base stations transmitted to the mobile station are synchronized.The use of the broadcast control channel (BCCH) is appropriate since no power control or frequency hopping is used that could affect accuracy of estimation.

For the network-based PL, we have the option of three base stations monitoring the mobile signal through the traffic channel (TCH). These measurements, which are basically of TOA, are shared in a system generally called the location service center (LSC), which computes the TDOA parameters and performs the estimation with methods discussed in previous chapters. The LSC system has direct communication with the MSC and to the HLR and VLR in order to use the information to estimate the PL.

In a CDMA system, the reverse traffic channel is active only when a call is active, making this channel practically useless for PL unless a service such as emergency E911 is requested via a call. The other option is to use the reverse access channel where a mobile answers to commands sent by the base station through the forward channel. The registration update of the mobiles can be used as a signal to estimate PL. This registration update can be established periodically through a command by the base station, which will have an effect on the signaling load of the base station.

The code acquisition and tracking used in the CDMA receivers can be helpful for estimating time delay [31] through the use of sequence correlation. The estimate would have an error based on a chip duration. After that, a system such as the LSC could be used to compute the TDOA from several measurements related to the various base stations. One of the limitations for location algorithms in a CDMA network is the MAI caused by other active users, not only on the same base station but also from other cells. With the use of power control, the signal of the mobile to be located arrives at its designated base station with no strength problem because all other active users have power control as well.

On the other side, the signal from the mobile to be located will not arrive at the other base stations involved in the location algorithm with the same strength as if the mobile were in such a cell; all other users designated to such a base station will contribute to the MAI caused by the signal used for location. This is a classic problem of the near–far effect. For emergency calls, the mobile could be programmed to transmit at maximum power regardless of distance to its servicing base station. This would mitigate the near–far problem in the other base stations involved. Solutions that use multiuser detectors or interference mitigation techniques to estimate TOA or TDOA have been proposed in the literature.