An Influence of Intentional Jamming Signals on Dsp Based Demodulators

Resistance of d~~erent types of digital radio signals to certain interferences and jamming signals is essential @om milita~ point of view. An experimental method was chosen to take into consideration injluerwe of all factors and elements to achieve more accurate results. This paper presents results of experiments concerning influence of intentional jamming on digital FSK and PSK demodulators. Several types of noise and imitating signals are considered, including autoregressive models of the jammed signal. The demodulators are sof~are DSP demodulators working in real time. INTRODUCTION Digital modulations are widely used in military wireless communications. Narrowband signals demodulators can be realized in classic analog way, using specialized digital integrated circuits or digital signal processors (DSP). The last option is a flexible and cost effective solution. It is usefid in case that the user needs several types of demodulators, using the same hardware. The resistance of such demodulators to different types of j amming signals is a very important property. Theoretical analysis is usually performed under simplifying assumptions and used mathematical models neglects influence of receiver’s filters, mixers, AGC, etc. Whereas the jamming signals have influence both on radio receiver HF amplifiers, AGC unit, IF unit, filters, demodulators and so on. All these units interacts, and analytical analysis is complicated and imprecise. v I IF ..................... .............. Furthermore fixed precision of calculations in DSP, roundoff errors propagation, non ideal AD conversion are another sources of inaccuracy. To analyze overall jamming signals influence on DSP demodulators a laboratory experiment was performed. DSP BASED DEMODULATORS In, the experiment two types of modulation were considered: frequency shift keying (FSK) and phase shill keying (PSK). To analyze the BER, and influence of certain demodulator type, three types of demodulators were elaborated: ● discriminator based FSK demodulator, ● correlation based FSK demodulator, . PLL DPSK demodulator. All three demodulators were realized using two TMS 320C31 DSPS. Block diagrams of used hardware is shown on fig 1. The RF unit is Rohde & Schwarz EK 890 HI? radio with 1.44 MHz intermediate ffequency. Analog to digital conversion is performed with accuracy 12 bits and sampling frequency 5 MSps. Digital signal is then fed to HARRIS HSP 50016 downconverter, containing tunable quadrature digital local generator, complex mixer and two sets of switched decimating filters. Signal on the output of the downconverter is a complex signal with zero center frequency and sampling frequency 1.7 times higher than selected 3 dB bandwidth. It provides proper selectivity and rejects aliased components [5]. .......~. . ........................... ............... Y=H=lt-U-h=li=l,! ;Cl ::{ local ~ generator Digital \ ; Digital downconverter ; ! demodulator :~ ..................... . ... ..... ...... ............................................. Fig. 1. Block diagram of the DSP based demodulator c1 Pc 0-7803-4902-4/98/$10.00 (c) 1998 IEEE The signal is then converted to TI serial standard and fed do serial input of the fwst DSP. In case of FSK signal, the first DSP works as frequency discriminator or performs calculation of correlation fictions. Instantaneous frequency is calculated on the basis of signal argument. Complex correlation functions are calculated using digital mixers and local quadrature generators. In case of PSK signal, the fwst DSP performs carrier recovery. Average frequency is calculated first, and it is subtracted from signal frequency. It enables extension of the frequency offset range. Then phase locked loop (PLL) is used for fine carrier synchronization. After serial transfer to the second DSP, signal is fed to software Schmitt relay and bit synchronization procedure. Synchronization and decision unit are realized using three shifted counters comparison method. The early and delayed counters are used for data rate generator phase correction. Output fi-omthe middle counter is used by decision unit to determine bit’s value. Output bit stream is transmitted to PC processor trough ISA interface. LABORATORY TEST STAND An efficiency of intentional jamming signals depends on their level, structure, and parameters. The resistance of radio link to the jamming signals depends on used type of modulation, its parameters and type of demodulator. To analyze all these interactions a laboratory test stand was established (fig 2). In this stand pseudonoise data signal and IQ modulation is performed by transmitting DSP 1, equipped with two D/A converters, Low frequency output IQ signal, with zero center tiequency modulate a radio signal generator. The source of .... . . .. . ... . ... ........ ..... ..... . ..... ........ DSP 1 i~ ~ lpseupdatat-FM.......................... ........ jamming signal signal generator + generator jamming signal is the HP 715 workstation. Its stereo AUDIO interface don’t support zero frequency (constant level), so it generate complex IQ signal with 15 kHz offset from the center frequency. This signal also modulate a signal generator, with center frequency shifted by -15 kHz in comparison with the useful signal generator, to eliminate the workstation offset. Both signals are added in wideband signal combiner and fed into radio receiver input. Intermediate frequency analog signal is the digitized and downconverted. Digital IQ signal from the radio is transmitted to DSPS performing a real time demodulation. To perform bit error rate (BER) measurement, additional, delayed output of data form DSP 1 is provided. The second DSP software is also modified, and BER meter is added. It compares the transmitted and received data. To do it correctly, data time delay was adjusted according to selected demodulator type and modulation parameters. On the basis of received bits counter calculates current BER and indicates error bits in received bits stream. All these data are then transmitted to PC memory for tier analysis. Analog scope enables observation and measurement of signal parameters in different points of demodulation process. JAMMING SIGNALS The efficiency of j armning signals depends on structure of the used signal and their structure and parameters. An influence of jamming signal level, center frequency error, frequency shifl error, data rate error and bandwidth versus usefid modulation scheme were analyzed. ............. .. ...... .. ... ....................... DSP 2, DSP3