This scanning RSSI receiver provides many advantages over a spectrum analyzer for evaluating the performance of WiMAX wireless networks at both 2.5 and 3.5 GHz.
WiMAX broadband wireless technology offers the promise of untethered, highspeed Internet access in supported service areas. Based on the IEEE 802.16 standard, WiMAX is specified for use in a wide range of frequencies through 66 GHz, with a variety of operational profiles differentiated by frequency band, channel bandwidth, and duplexing mode.1-3 WiMAX has been developed rapidly from inception to deployment. According to a recent (January 2008) study of the global WiMAX market, more than 400 operators are now deploying WiMAX networks.4 These WiMAX networks mainly operate in three bands, at 2.3, 2.5, and 3.5 GHz, as endorsed by the WiMAX Forum (www.wimaxforum.org). For example, the United States and Canada will have released licenses in frequency bands of 2.305 to 2.320 GHz, 2.300 to 2.400 GHz, 2.345 to 2.360 GHz, and 2.469 to 2.690 GHz. South Korea has deployed a WiBro system in the 2.3 to 2.4 GHz band, and WiMAX networks in India have used frequency bands of 3.4 to 3.8 GHz, which are also likely to be licensed for use in the European Union.5
All of these WiMAX networks require careful RF planning, optimizing, and monitoring to ensure maximum coverage, capacity, and quality of service (QoS). Propagation models are useful for network planning, and received-signal-strength-indication (RSSI) measurements can provide a simple indication of signal strength throughout the network coverage area. Several empirical models are often used in studying wireless communications systems, including the Okumura-Hata, COST231-Hata, Walfisch-Ikegami, Keenan-Motley, and Lee models. But no propagation model can account for all variations in a system. For any application, an empirical model should be calibrated to achieve the highest simulation accuracy. And RSSI measurements in the appropriate operating frequency bands offer an effective means of calibrating a wireless system model. 6-9
A spectrum analyzer is one of the more versatile of RF measurement tools, and invaluable for finding signals across broad ranges of bandwidth. But a scanning RSSI receiver can provide advantages in terms of scanning speed, measurement accuracy, and data processing and analysis, especially for evaluating the performance of WiMAX networks at different operating frequencies. To demonstrate its effectiveness for WiMAX networks, the authors designed a portable scanning RSSI receiver capable of scanning across the full WiMAX spectrum, including the 2.3-, 2.5-, and 3.5-GHz bands.
Figure 1 shows a simplified block diagram of the scanning RSSI receiver. Input signals first pass through a RF filter, which prevents out-ofband signals from mixing with the first local oscillator (LO) and creating unwanted responses within the intermediate-frequency (IF) band. A low-noise amplifier (LNA) follows the RF filter in order to boost the lowlevel signals typically encountered by a scanning RSSI receiver (compared to a spectrum analyzer). This LNA contributes most of the receiver's noise figure. Prior to the second mixer is an IF filter that provides image rejection for the system.10
Another principle difference between the architecture of a conventional spectrum analyzer and the scanning receiver designed in this work is the elimination of envelope detectors or logarithmic detectors. In the present scanning RSSI receiver design, low IF analog signals are converted to digital signals by means of a highresolution analog-to-digital converter two resolution-bandwidth (RBW) filters, with bandwidths of 50 kHz and 5 MHz, are implemented in the digital domain; by doing so with digital techniques, these bandwidths can be readily modified according to the requirements of special applications.
The scanning RSSI receiver was designed to provide a dynamic range from -110 to -30 dBm when operating with a 50-kHz resolution bandwidth (RBW). To verify the receiver scheme shown in Fig. 1, simulations were performed by means of commercial system simulation software, the Advanced Design System (ADS) suite of computer-aided-engineering (CAE) tools from Agilent Technologies (www.agilent.com). Figure 2 shows one of the simulation schematics from that program.
As noted earlier, the receiver system's noise figure is a function of the front-end LNA, and the noise figure can be lowered by placing a high-performance LNA in front of the receiver. Of course, high-level signals amplified by the LNA tend to saturate the components following the amplifier, such as mixers and other amplifiers (not to mention the LNA itself), and this can degrade the linearity of the system. There is a tradeoff between sensitivity and maximum input level. The simulation results shown in Fig. 3 indicate that when the input power is -30 dBm, the amplitude distortion measured at the input port of the ADC is less than 1 dB (-0.696), showing that an LNA can be used in the scanning RSSI receiver architecture to achieve improved low-level signal sensitivity.
In a scanning RSSI receiver (and other receivers based on superheterodyne mixing), there is a pair of image signals for each LO, with the images located on both sides of the LO frequency. Each image frequency is twice the IF away from the measured frequency. In traditional RF receivers, IF filters are used to decrease image-signal interference to an acceptable level. In performing the ADS simulation on the scanning RSSI receiver, the input power of measured signal was -110 dBm and the power of the image interference was swept from -110 to -20 dBm in 1-dB steps. The simulation results shown in Fig. 4 reveal that when the image power is -48 dBm, signals as small as -110 dBm can still be measured with error of less than 1 dB (0.748).
In making measurements on the scanning RSSI receiver (Fig. 5), input signals were generated by a model E4438C signal generator from Agilent Technologies, with experimental results from the receiver displayed on a personal computer (PC). In the test frequency bands of interest, the insertion loss of the test cable was 0.9 dB. For this report, the input power is defined as the power at the scanning RSSI receiver's RF port, including the insertion loss of the test cable.
Measurements made on the scanning RSSI receiver focused on the accuracy of the displayed average noise level (DANL), the dynamic range, the measurement speed, and other key performance parameters when using the digitally generated RBW of 50 kHz. The measurement results will be shown for two frequency bands, from 2.3 to 2.7 GHz and from 3.4 to 3.8 GHz. In the 2.3-to-2.7-GHz band, the DANL was measured with the test signal source shut off. As shown in Fig. 6, the DANL is less than -117 dBm at 2500.125 MHz.
To verify the accuracy of the RSSI measurements, the test input signal was fixed at a frequency of 2500.125 MHz and the power of the input test signal was varied. When the input test signal power was set at -30 dBm, the measured result was -29.533 dBm (Fig. 7). The test signal input power was reduced in steps and compared with the measured results, with a measured reading of -112.483 dBm resulted when the input power level was set at -113 dBm at the signal generator (Fig. 8). For a test signal power range of -113 to -30 dBm, the scanning RSSI receiver shows measurement accuracy that was better than 1 dB.
The frequency response of the RF module within the scanning RSSI receiver for the frequency band of 2.3 to 2.7 GHz tends to vary as a function of operating conditions, including temperature. But after a careful calibration, it was possible to achieve outstanding receiver amplitude flatness across the full frequency band. To verify this performance, the signal generator's input test signal was fixed at -80 dBm and the input frequency was varied from 2.3 to 2.7 GHz. The measurement results from these tests are shown in Table 1. At all frequencies, errors for power measurements to -80 dBm were less than 1 dB.
Figure 9 shows the measurement results when the receiver swept a 5-MHz frequency span in 250-kHz steps. The scanning RSSI receiver can perform continuous sweeps across a range of frequencies in 250-kHz steps or can be assigned to switch to discrete frequencies, according to the requirements of an application.
Measurements in the 3.4-to-3.8- GHz band were similar to those performed in the 2.3-to-2.7-GHz band. As shown in Fig. 10, the DANL at 3500.125 MHz is also less than -117 dBm. Figure 11 shows the measured value of DANL at 3500.125 MHz to be -29.637 dBm for an input power set at -30 dBm while Fig. 12 shows the measured value of DANL at 3500.125 MHz to be -112.406 dBm for an input power level of -113 dBm.
To verify whether 1-dB accuracy could be achieved across the full frequency band, the input power was fixed at -60 dBm and the input frequency was tuned across 3.4 to 3.8 GHz. Table 2 shows the measured values for this test, with less than 1 dB error at receiver input levels to -60 dBm. Figure 13 shows the results of measuring a continuous band of frequencies around 3.5 GHz, with the input power set at -80 dBm.
Continue to page 2
Measurement speed is one of the more important parameters to consider when evaluating the performance of a scanning RSSI receiver. According to the theorem of W.C.Y. Lee,11 at most 50 sampling points are needed to eliminate the influence of fast fading for measurements across a distance of 40 wavelengths (40?) of the operating frequency. Since RSSI measurements for the propagation model calibration are typically performed from a moving vehicle, it's necessary to calculate the fastest vehicle speed that will be supported by the scanning RSSI receiver design:
? = the wavelength of the received CW signal,
v = the vehicle speed, and
n = the measurement rate.
The middle column of Fig. 14 shows the sample count reported to the PC every 200 ms, computed from the values in the left-hand column. According to the values in Fig. 14, the average measurement rate can be found by Eq. 3.
When the measured frequency, f, is 3.8 GHz, the vehicle speed can be as great as shown in Eq. 4.
When equipped with an internal 12-channel Global Positioning System (GPS) module, the scanning RSSI receiver can measure and display RSSI information on a map, which is an essential capability for WiMAX network planning and optimization. Figure 15 shows a sampling of this capability, with the GPS information from the experiments conducted here.
In conclusion, this article provided details on a portable scanning RSSI receiver capable of measurements within frequencies of interest to WiMAX network developers, including 2.3 to 2.7 GHz and 3.4 to 3.8 GHz. In these bands, measurements were made with an RBW of 50 kHz, with DANL of less than -117 dBm and 1 dB accuracy for power measurements across the broad range of -113 to -30 dBm. The receiver showed its agility, with RSSI measurement speed estimated to be better than 500 data points per second. With a built-in GPS module, the receiver is fully capable of assisting on WiMAX network planning and optimization, in particular for calibrating WiMAX network models.
This work was supported in part by National Natural Science Foundation of China (NSFC) under Grant 60621002 and in part by the National High-Tech Project under Grant 2007AA01Z2B4.
1. WiMAX Forum, "Fixed, nomadic, portable and mobile applications for 802.16-2004 and 802.16e WiMAX networks," December 2005.
2. IEEE Std 802.16-2004, IEEE Standard for Local and metropolitan area networks: Part 16: Air Interface for Fixed Broadband Wireless Access Systems, June 2004.
3. IEEE Standard 802.16e-2005, Amendment to IEEE Standard for Local and Metropolitan Area Networks -6 Part 16: Air Interface for Fixed Broadband Wireless Access Systems- Physical and Medium Access Control 7 Layers for Combined Fixed and Mobile Operation in Licensed Bands, December 2005.
4. Carter L. Horney, "WiMAX 08 The 3G+ Broadband Alternative," Forward Concepts, Report No: 8010, Published January 2008.
5. Josh Raha and Mark Andrews, "Making Sense of WiMAX," Microwave Journal, WiMAX Supplement, pp. 6-12, November 2007.
6. Huseyin Arslan and Daljeet Singh, "Understand Requirements For WiMAX Testing," Microwaves & RF, September 2006.
7. D. Erricolo and P.L.E .Uslenghi, "Propagation Path Loss-A Comparison Between Ray- Tracing Approach and Empirical Models," IEEE Transactions on Antennas and Propagation, Vol. 50, No. 5, May 2002, pp. 766-768.
8. M.N. Lustgarten and James A. Madision, "An Empirical Propagation Model (EPM-73)," IEEE Transactions on Electromagnetic Compatibility, Vol. EMC-19, No. 3, August 1977, pp. 301-309.
9. Nuno C. Goncalves and Luis M. Correia, "A propagation Model for Urban Microcellular Systems at the UHF Band," IEEE Transactions on Vehicular Technology, Vol. 49, No. 4, July 2000.
10. Agilent Technologies, "Agilent Spectrum Analyzer Basics," publication No. 5952-0292, August 2, 2006, www.agilent.com.
11. W. C. Y Lee and Y. S. Yeh, "On the Estimation of the Second-Order Statistics of Log Normal Fading in Mobile Radio Environment," IEEE Transactions on Communications, Vol. Com-22, June 1974, pp. 869-873.