Swarna Sangeetha Wipro Infotech, Mysore sangeetha.haribabu@wipro.com rsangee@hotmail.com Ph. No: 09901243859 Abstract GPS signals are well known to be used in a gamut of applications ranging from defense to commercial. Nevertheless, it has been proved after intensive research that a stand-alone GPS receiver is incapable of providing consistent navigation if the operational environment becomes non-benign. To cater to such environments the usual practice is to augment the receiver with additional sensors. These sensors either mitigate the anomalies on GPS signals or operate as an alternative source to navigation. Most recently the burgeoning mobile market has witnessed a surge in GPS receivers being integrated with the phones. This integration is not only due to the mandate by the US Government but also due to the interest in the community as a whole to identify their locations to handle catastrophic situations. This has led several mobile companies to integrate small GPS receiver chipsets with their mainstream mobile communications. However, there are several challenges that need to be addressed in such an integration environment. Firstly, the GPS signal strengths are very weak making them vulnerable to RF interference / jamming caused by GSM / GPRS signals used by phones. Secondly, the real-time integration issues and the resources requirement need to be effectively addressed with the available battery power. The focus of this paper is to address these challenges and provide mitigation solutions. A thorough literature survey is done that encapsulates these areas. The interference effects are also studied using simulation models and the results are illustrated. 1.0 Introduction: Rapid increase in navigation applications, in particular commercial applications, places a huge demand on the robustness of the navigation system. Due to its lower cost and good performance, GPS is considered as the primary source of navigation. However, GPS has its own limitations. A stand-alone GPS receiver can work well only if it can receive signals at least from 4 satellites with a reasonably good geometry. In addition, the GPS signal can also be jammed very easily by either intentional or unintentional interferences (Kaplan, 1996). The integration of GPS receivers into mobile phones also poses the same problem – if the mobile phone transmitter is not isolated well from the GPS receiver, then the receiver will get saturated by the higher power of one of the harmonics of the transmitted signal. In GPS integrated mobile phones, the GPS receiver co-exists with the mobile phones which usually transmit higher powers in comparison to the received GPS signal strength. Though a good isolation is usually guaranteed between the two systems, nevertheless any leakage from the mobile phone transmitter to the GPS receiver results in the jamming or degradation of the GPS signals. These interference or jamming signals reduce the overall C/N0 of the received GPS signals. By integrating an MEMS inertial sensor (accelerometers and gyroscopes) with the GPS receiver tracking loops in an ultra-tight integration, the complete integrated system performance (mobile phone, GPS and INS) can be improved substantially. This performance improvement is given by the indicator J/S (Jamming to Signal Strength). In order to get high positioning performance (e.g. accuracy and robustness), many multiple sensor integration schemes have been proposed, including GPS/GLONASS, GPS/INS (Wang et al., 2001). Besides these integration methods, employing antenna arrays and adaptive filters can also substantially improve the GPS based positioning performance in interference and jamming environments. An antenna array with a set of multi-channel receivers can significantly improve the position and location performance due to its anti-jamming and multipath mitigation property. As a result, the development of these spatial and temporal processing GPS receivers have become a hot topic in the research and industrial community. Recently, several techniques based on antenna-array reception have been proposed. Among them, conventional planar antenna array is the most widely used in the researches. However, the coverage of a general planar antenna array is limited in space. For example, these arrays cannot provide good reception performance if signals are received at low elevations. This array inherits the wide coverage characteristic of conventional spherical arrays and utilizes only about half the number of elements compared to a planar array to cover the upper-semi-spherical space above the earth plane. This property enhances the signal reception from both overhead and horizontal directions simultaneously. Thus, unlike common planar arrays, this novel antenna array with special geometry can receive satellite and Pseudolite signals from all directions, even from the horizon. This paper analyses the various aspects of the effects of interference signals from mobile phones on the GPS receiver performance by conducting simulation experiments using Matlab. The performance of the code and carrier tracking loops are also analysed with theoretical analyses. Finally, it is shown through simulation experiments that the receiver performance can be improved significantly if the tracking loop parameters can be optimised. 2.0 GPS Receiver Performance 2.1 Carrier Tracking Loop The conventional unaided GPS receiver uses a 2nd order carrier-tracking loop with a loop bandwidth of about 12 to 18Hz. But, to receive a dynamically varying signal the order of the loop should be increased to 3 to reduce the dynamic stress error (Ward, 1998; Irsigler et al., 2002; Jwo, 2001). However, the design of a 3rd order filter is complex, and furthermore it has stability problems. In interference and jamming systems, the tracking loop filter bandwidth can be reduced to about 3 to 5 Hz to improve the overall signal-to-noise ratio. This can be done by integrating some additional sensors such as INS which can remove most of the Doppler signals. Further reduction in bandwidth is possible if an accurate receiver clock is used, but sometimes they are too complex to be implemented. Fig. 1 shows the block diagram of the receiver tracking loop.  Figure 1. Carrier Tracking Loop Traditionally, the carrier discriminator output is passed through a low-pass filter which drives the NCO (Kaplan, 1996; Tsui, 2000). The low-pass filter attenuates the high frequency components and noise that result from the mixing process. This system architecture ensures optimal tracking as long as the biases in the measurements are not significant. However, when correlations-induced biases are present in the measurements, the tracking performance starts to degrade. 2.1.1 Discriminator Error Model Conventionally, a Costas tracking loop is used due to its insensitivity to 180º phase reversals in the I and Q data. The discriminator is the central part of the tracking loop which computes the phase error that is required to align the internally generated signals to the incoming signal. Because a two quadrant arctangent discriminator is optimal (Kaplan, 1996), it is used in most of receivers. The phase error generated by the discriminator is given by:  2.2 Code Tracking Loop Two tracking loops, the CPLL and DLL, work concurrently to demodulate the pseudo-random noise code and carrier frequency to extract the 50 Hz navigation data. User position is calculated from code phase measurements, whereas velocity is calculated from the carrier frequency measurements. Generally, the code loop is less sensitive to dynamics than the carrier loop owing to its low frequency characteristics, and also due to carrier aiding. The carrier aiding virtually removes most of the dynamics from the measurements that enter the code-tracking loop. Therefore the bandwidth of the code-tracking loop can be maintained at about 1 to 4 Hz. If the carrier loops loses lock, usually the code loop will also lose lock (Kaplan, 1996).  Figure 2. Code Tracking Loop The closed loop transfer function of the code-tracking loop shown in Figure 2 is given as: |