© 2004 Matt Welsh – Harvard University1CS263: Wireless Communicationsand Sensor NetworksMatt WelshLecture 2: RF Basics and Signal EncodingSeptember 28, 2004© 2004 Matt Welsh – Harvard University2Today's LectureBasics of wireless communicationsAnalog and digital signalsBandwidth and channel capacitySignal encoding and modulation© 2004 Matt Welsh – Harvard University3What is RF?Radio Frequency is an electromagnetic signal with a frequency between 3 kHz and 300 GhzRF signals carry analog or digital information●Analog: Information content varies continuously over time●Example: radio and TV stations●Digital: Information content consists of discrete units (e.g., 0s and 1s)●Example: Cell phones and wireless networks© 2004 Matt Welsh – Harvard University4Carrier WaveHow do we send information in a radio signal?Carrier wave●An RF signal – usually a sinusoid – that carries information●Carrier is usually a much higher frequency than the information itself!●Ex: 2.4 GHz 802.11b networks carry a lot less than 2.4 GBit/sec of data....●Rather, carry up to 11 MBit/sec of information●Why use a carrier??●Easier to generate a sinusoid signal, and it will travel further.Carrier wave frequency●The frequency of a radio transmission is the center frequency of the carrier●Actual frequency of the carrier changes over time, e.g., with FM transmission© 2004 Matt Welsh – Harvard University5Signal ModulationHow do we encode information in a carrier wave?An information signal must be modulated onto the carrier wave●That is, we must modify the carrier wave in some way...●Receiver must demodulate the carrier to get back the original signalAmplitude Modulation (AM)●Modify the amplitude of the carrier with respect to the amplitude of the signalFrequency Modulation (FM)●Modify the frequency of the carrier with respect to the amplitude of the signal© 2004 Matt Welsh – Harvard University6Signal ModulationCarrierInformationsignalAmplitudeModulation(AM)© 2004 Matt Welsh – Harvard University7Signal ModulationCarrierInformationsignalFrequencyModulation(FM)© 2004 Matt Welsh – Harvard University8Digital ModulationHow do we modulate digital signals?Amplitude shift keying (ASK)●“0” bit is the absence of the carrier (flat signal)●“1” bit is the presence of the carrier with some fixed amplitudeFrequency shift keying (FSK)●“0” bit is carrier at frequency f0 ; “1” bit is carrier at frequency f10 10 1 0 1ASKFSK© 2004 Matt Welsh – Harvard University9Other modulation techniquesLots of other modulation schemes are used in practice●Each has different properties in terms of resiliency to noise, interference,multipath effects, etc.Gaussian Frequency Shift Keying (GFSK)●Binary 1 is a positive frequency shift from base frequency●Binary 0 is negative frequency shift from base frequency●Used in BluetoothPhase shift keying (PSK)●The phase of the carrier is used to represent data●“Differential quadrature phase shift keying” (DQPSK) used by 802.11b networks●Four phase levels representing 00, 01, 10, and 11 bit sequencesQuadrature Amplitude Modulation (QAM)●Combination of AM + PSK●Use two amplitudes and 4 phase levels to represent each sequence of 3 bits© 2004 Matt Welsh – Harvard University10BandwidthA typical signal will include many frequencies●Fourier theorem: Any periodic signal is a combination of sinusoids© 2004 Matt Welsh – Harvard University11BandwidthA typical signal will include many frequencies●Fourier theorem: Any periodic signal is a combination of sinusoidsSpectrum: The range of frequencies in a signal●In this case, [f ... 5f]Bandwidth: The width of the spectrum●In this case, (5f – f) = 4f© 2004 Matt Welsh – Harvard University12BandwidthWhat happens if we increase the bandwidth?●Consider two waves, one with bandwidth 4f, the other with bandwidth 12f: ●As the bandwidth increases, the wave approximates a square wave.●A perfect square wave (or any true digital signal) has infinite bandwidth●Caution: “Bandwidth” and “data rate” often used interchangeably© 2004 Matt Welsh – Harvard University13Data Rate and BandwidthWhat is the relationship between bandwidth of a signal and its information-carrying capacity?●This wave can carry 2 bits every period, or 1 bit per T/2 sec●If T = 1 µsec, the fundamental frequency f = 1 MHz●Data rate is 2 * 106 = 2 Mbps●Caution: “Bandwidth” and “data rate” often used interchangeably“0 bit”“1 bit”Fundamental frequency = 1/T© 2004 Matt Welsh – Harvard University14Data Rate and BandwidthWhat is the relationship between bandwidth of a signal and its information-carrying capacity?●What if we double the bandwidth?●If the fundamental frequency is the same, data rate is identical!●But ... this is a “cleaner” signal.●A receiver will be much more likely to discern 0 and 1 bits from this waveform.●In general: the greater the bandwidth, the less distortion in the signal.●Caution: “Bandwidth” and “data rate” often used interchangeably“0 bit”“1 bit”Fundamental frequency = 1/T© 2004 Matt Welsh – Harvard University15Gain, loss, and decibelsRatio between two signal power levels is often measured in decibels (dB):●gain(dB) = 10 log10 (Pout / Pin)●loss(dB) = –10 log10 (Pout / Pin) = 10 log10 (Pin / Pout)●Where Pout is the output power level, and Pin is the input power levelExample●Signal with power level 10mW transmitted over a wireless channel.●Receiver gets a signal of 2mW.Loss = 10 log (10/2) = 10(0.698) = 6.98 dB© 2004 Matt Welsh – Harvard University16mW and dBmDecibels refers to relative changes in magnitude, not absolute valuesSo ... we define the dBW (decibel-Watt) as a reference●1 Watt of transmission power == 0 dBW●Example: WGBH, 89.7FM in Boston transmits at 100,000 Watts●Power in dBW = 10 * log(100,000W / 1W) = 10 * 5 = 50 dBWFor wireless networks, the dBm (decibel-milliwatt) is more useful●1 mW transmission power == 0 dBm●10 mW == 10 dBm●0.1 mW == -10 dBm●802.11b networks have a max transmit power of 100 mW == 20 dBm© 2004 Matt Welsh – Harvard University17Channel capacityHow do we get the most out of a signal of limited bandwidth?●Want to minimize noise.Nyquist Bandwidth:●Given bandwidth B, highest signal rate that can be carried is 2B log M●Where M is the number of discrete voltage levels (usually 2)●Simplest form: low-to-high transition is a “1”, high-to-low