Mobile Computing

Download as pdf or txt
Download as pdf or txt
You are on page 1of 7

Mobile Communications

WS 2007/2008

Solution 1
Discussion: 31.10.2007

Solution 1.1: Modulation


Modulation is needed to press digital data onto the analogue radio transmission medium. a.) Is it possible to transmit a digital signal, e.g. coded as square wave as used inside a computer, using radio transmission without any loss? Give reasons for your answer. No, lossless transmission of an analogous signal is not possible: attenuation, dispersion, etc will be present all the time. Furthermore, each modulated signal propagates as a burst of analogue sinus waves, in extreme case as in a square wave needing an infinite number of sinus waves for exact representation, causing infinite bandwidth for communication. But, no transmission medium is able to transmit infinitely high frequencies, so all the time in a given bandwidth, we can only transfer an approximation of the digital signal. Furthermore, high frequencies will be attenuated much more than lower ones, thus after very short distance only part of the necessary frequency range for exact reconstruction of the signal is recognizable. So, digital signals cannot be transmitted without loss. [This even holds for copper cable! Even here the digital signal is transmitted as burst of waves, and attenuated while travelling over the cable.] b.) Think of a phase diagram and the points representing bit patterns for an advanced PSK scheme (see lecture slides 2.1, page 16). How can a receiver decide which bit pattern was originally sent when a received point lies somewhere in between other points in the diagram? Why is it, thus, difficult to code more and more bits per phase shift? If thinking in such points, for a received signal the corresponding point in the phase diagram is compared to the surrounding points, e.g. by measuring the distance to the other points. (Thinking in such points here only is for easier understanding, not for the real processing inside receivers.) A received point then is corrected to the allowed point which has nearest distance. In this way errors in phase (or, amplitude, considering QAM) can be detected and corrected. Having now more and more points, the risk increases to correct a received point wrongly if the average distance to the correct point becomes more that half the distance between allowed points. On this reason, advanced PSK schemes are only used if the influence by interfering signals from the environment is comparatively low. If such influences become too high, todays systems automatically can switch to a scheme with a lower number of allowed points (and thus lower data rate). Maybe, from this explanation you can also see why a n-QAM scheme is more robust than a corresponding pure n-PSK scheme: having a look below in figure c) as 16-QAM and comparing with a 16-PSK where all points would be placed on one circle, the distances between the points are higher.

Mobile Communications
WS 2007/2008 c.) Modulate the ASCII character sequence techniques: MC07 with the following modulation

* Not the codes used in reality, but better for fast signal identification in this exercise

For your solution, assume that each signal is represented by one sinus oscillation. For c.), you can estimate the correct phase/amplitude from the given position in the diagram (or, if you prefer, you can calculate it based on the angle of a virtual lines through the points). For representation of the ASCII characters, refer e.g. to wikipedia. ASCII codes: MC07 = 77 67 48 55 = 01001101 01000011 00110000 00110111 For a), encode each bit itself For b), encode 01 00 11 01 01 00 00 11 00 11 00 00 00 11 01 11 For c), encode 0100 1101 0100 0011 0011 0000 0011 0111

Solution see next page (from the exercise hour).

Mobile Communications
WS 2007/2008

Solution 1.2: Bandwidth


As told in the lecture, the term bandwidth often is used if data rate is meant. Although both terms have a relation, nevertheless they are different concepts. a.) A digital signaling system is required to operate at 9.6 kBit/s. If with one signal a 4-bit word is encoded, what is the minimum bandwidth necessary for the radio transmission? C = 9.6 kbps; number of bits N per signal = 4 Now: bandwidth B for transmitting S signals per second: B = S/2 (in Hz) Each signal carries 4 bit: C = 4 bit S

B = C/(8 bit) = 1.2 kHz

Generally spoken, there also is a formula you can use without thinking about all the relations between bandwidth, symbols, and data rate:
C = 2B log2 M (in bit/s) for M = 2N different symbols (or KN in general, for non-binary systems)

b.) When talking about signal strength or the amplification of a signal gained by an antenna, P decibel (dB) is the common measurement unit. It is defined as x = 10 log10 1 , with P1 P2 and P2 being the powers of two signals which should be compared (e.g. sending power and receiving power. Given such a scale, an antenna gain of 3 dB means that the signal strength is doubled by the antenna. Also, the influence of interfering signals is measured in dB. Here, it is talked about signalto-noise ratio (S/N ratio), which defines the power of the own signal in relation to all disturbing signals in the environment which can interfere with the own signal and by this reduce signal quality. The capacity C of a channel with bandwidth B for a given S/N ratio is lower than given in S the simple formula on the lecture slides: C B log2 1 + . [On right side, for N measurement units you have to multiply with 1 bit, stating that you are coding one bit per transferred symbol.] Given now is a channel with bandwidth of 3 MHz. We want to achieve a capacity of 20 MBit/s. What is the required signal-to-noise ratio for achieving this capacity? C = 20 MBit/s; B = 3 MHz assume: 1 Bit is encoded per signal (for measurement units), i.e. capacity C equals symbols per second. Otherwise, as above, for P bits per symbol, you have again to multiply with P on the right side of the formula in the next line:

Mobile Communications
WS 2007/2008
C = B log2 (1 + S ) N

S = 2(C / B ) 1 N S = 220 / 3 1 = 100,39 N with the dB-formular: x = 10 log10 (100,39) = 20.04 dB

When having a measurement tool for signal strength, all the time this S/N ratio is given as an important parameter, thus it is good to have some feeling about its meaning. c.) Given is a base station with a transmit power of 100 mW. The noise level is given with a power of 0.001 mW. For communication, you have the frequency band of 2.4 2.425 GHz available. What is the maximum possible data rate at distances of 10m, 50m, resp. 100m for a device communicating with the base station? Note: not identical to exercise hour because I missed two comma-positions Bandwidth B = 2.425 GHz 2.399 GHz = 0.026 GHz = 26 MHz Now use C = B log 2 (1 + S ) , and additionally consider the loss of signal strength after d
N

meters for the carrier frequency f = 2412 MHz (which is mid of the interval of 2.4 2.425 GHz), c = 3108 as speed of light, and the relation c = f: 2 2 1 1 c S = Pr = Pt = Pt 100 2 0,0001 ( in mW) = 0,01 2 ( in mW) d d 4 d 4 df

C = 26 Mhz 0.138 bit = 3,588 MBit/s After d = 50m: signal strength S = 0.000004 mW S/N = 0.000004/0.001 = 0.004 S C = B log (1 + ) C = 26 Mhz 0.00575 bit = 0.15 MBit/s N After d = 100m: signal strength S = 0.000001 mW S/N = 0,000001/0.001 = 0,001 S C = B log (1 + ) C = 26 Mhz 0.00144 bit = 0,037 MBit/s N
C = B log 2 (1 + S ) N
2 2

After d = 10m: signal strength S = 0.0001 mW

S/N = 0.0001/0.001 = 0.1

Assumed here is a PSK encoding, i.e. 1 MHz encodes 1 MBit/s! In reality, more efficient encoding is used.

Mobile Communications
WS 2007/2008

Solution 1.3: Signal Propagation


The propagation of signals in radio transmission depends on the chosen carrier frequency. a.) Why can waves with very low frequency follow the earths surface? Why are they not used for data transmission in computer networks? Below 2 MHz radio waves more or less follow earths surface. One reason is diffraction (waves are diffracted in the direction of objects which have more or less the size of their wavelength). An additional factor is electrical power which is induced in the surface by the waves. This slows down the wave front near the surface so that the wave front is brought in a curve and travels along the surface. There are several reasons why low frequencies are not used for radio data transmission: Low frequencies give us only low data rates, because the bandwidth is smaller than for higher frequencies. Regarding the large wavelength of low frequencies, for generation and receipt of the waves we need really large antennas which are not suitable for the usual size of a mobile device. The attenuation and shading of lower frequencies is less than of higher frequencies so (see the amplitude modulated radio program) we have network ranges of hundreds of kilometres, which is not useful for building up a cell network.

b.) Why are only lower frequencies (up to some hundred GHz) regulated and not higher frequencies (in the THz range)? Waves with THz-frequencies are in the range of infrared and visible light. These signals are shaded by nearly any material, thus one transmission would not that easy disturb other transmissions (e.g. if a wall is in between). In contrast, for e.g. WLAN or GSM, the radio waves go through walls, thus not everybody should be allowed to send on these frequencies. Furthermore, using waves in range of infrared/visible light for transmission is not very robust: even a simple lamp would produce lots of disturbances. And: even using the THz range below infrared is not prominent, because it is technically not simple to produce waves of that range.

Mobile Communications
WS 2007/2008 c.) Name several methods for inter-symbol interference (ISI) mitigation. How does ISI depend on the carrier frequency, symbol rate, and movement of sender/receiver? Inter-symbol interference could be mitigated by: Large guard times between symbols. This drastically reduces the achievable data rate. Lower data rate. This drastically reduces the achievable data rate. Estimation/calculation of the strongest transmission paths and adoption of the receiver to these paths, i.e. using antenna diversity for combination of different incoming symbols with different delays. This seems to be better than the above ideas. Carrier frequency: the higher the carrier frequency, the less multipath propagation we have. So, ISI becomes less problematic with higher frequencies. Symbol rate: the higher the symbol rate, the shorter is the transmission time of a single symbol, so ISI effects become more problematic. Movement: for moving users, paths need to be recalculated all time. Thus, the adoption of the receiver needs some time. The higher the movement, the more probably this cannot be done in time, giving problems with ISI.

Dependence on

Thus, in real systems: for frequencies high enough, you have on ISI problems, for other systems usually antenna diversity is used. When travelling with too high speeds, connections can break down because the adaptations cannot be done in time. c.) A vehicle travels with a speed of 30 m/s and communicates having a carrier frequency of 1 GHz (i.e. wavelength: 0.3 m). What is the maximum Doppler shift? What is the approximate fade duration? v 30 = 100Hz Maximum Doppler shift = = 0 .3 The bandwidth of the signal will be increased by approximately 200 Hz due to the motion of the vehicle (shift is possible in both directions). The time to travel a half wavelength, which produces a phase shift of 180 degree and thus leading to an inverted signal, is: 0.15 m t= = 5 m sec 30 m / s We can expect significant fades approximately every 5 msec. This is (easy to recognize) the reciprocal of 200 Hz range of increased bandwidth due to Doppler shifts.

You might also like