20-06-2013, 04:50 PM
Minimum Shift Keying Modulation/Demodulation Trainer ST2116
Minimum Shift Keying.pdf (Size: 3.24 MB / Downloads: 26)
Theory
The ever increasing demand for digital transmission channels, in the radio frequency
(RF) band presents a potentially serious problem of spectral congestion and is likely
to cause severe adjacent and co-channel interference problems. This has, in recent
years, led to the investigation of a wide variety of techniques for solving the problem
of spectral congestion. Some solutions to this problem includes:
1. New allocations at high frequencies
2. Better management of existing allocations
3. The use of frequency-reuse techniques such as the use of narrow-beam antennas
and dual polarizing systems
4. The use of efficient source encoding techniques
5. The use of spectrally efficient modulation techniques.
From the design point of view we are concerned here with the fifth option i.e. the use
of spectrally efficient modulation techniques. One such modulation scheme, which is
spectrally efficient, is known as Minimum Shift Keying (MSK). We shall study the
MSK signal format and its relation to other schemes such as quadrature phase shift
keying (QPSK), offset QPSK (OQPSK), and frequency shift keying (FSK). The main
attributes of MSK, such as constant envelope, spectral efficiency, error rate
performance of binary PSK, and self-synchronizing capability will all be explained on
the basis of the modulation format.
Spectral Efficiency And Minimum Shift Keying :
In any communication system, the two primary communication resources are the
transmitted power and channel bandwidth. A general system-design objective would
be to use these two resources as efficiently as possible. In many communication
channels, one of the resources may be more precious than the other and hence most
channels can be classified primarily as power-limited or band-limited. (The voice
grade telephone circuit with approximately 3 KHz bandwidth is a typical band-limited
channel, whereas space communication links are typically power limited). In powerlimited
channels, coding schemes would be generally used to save power at the
expense of bandwidth, whereas in band-limited channels “spectrally efficient
modulation” techniques would be used to save bandwidth. The primary objective of
spectrally efficient modulation is to maximize the bandwidth efficiency, measured in
bits/s/Hz. The primary objective of spectrally efficient modulation is to maximize the
bandwidth efficiency, defined as the ratio of data rate to channel bandwidth (in units
of bits/s/Hz). A secondary objective of such modulation schemes may be to achieve
this bandwidth efficiency at a prescribed average bit error rate with minimum
expenditure of signal power. Some channels may have other restrictions and
limitations, which may force other constraints on the modulation techniques. For
example, communication systems using certain types of nonlinear channels call for an
additional feature, namely, a constant envelope, which makes the modulation
impervious to such impairments. This is needed because a memory less nonlinearity
produces extraneous sidebands when passing a signal with amplitude fluctuations.
Such sidebands introduce out-of-band interference with other communication
systems.
FSK and PSK :
The constraint of a constant envelope feature for the modulation scheme narrows the
search to two major signaling techniques, namely, FSK and PSK. Consider binary
communication-transmitting a pulse every T seconds (at the signaling rate of 1/T
baud) to denote one of two equally likely information symbols, +1 or -1. FSK denotes
the two states by transmitting a sinusoidal carrier at one of the two possible
frequencies, whereas binary PSK (BPSK) uses the two opposite phase of the carrier
(i.e., 0 and 180"). Figure 1 shows typical signals in the two types of modulation. Note
that BPSK is also equivalent to amplitude modulating the carrier with the information
bit stream, i.e., multiplication with +1 or -1. The two schemes can be compared on the
basis of their bit error rate (BER) performance (i.e., the average number of errors in
transmitting a long bit stream) through an ideal channel. The ideal channel is taken to
be a linear all-pass channel, corrupted only by additive white Gaussian noise with a
constant (one-sided) power spectral density of No W/Hz. The required ratio, Eo /No, of
signal energy per bit (Eo) and noise level No to achieve a given BER (such as 1 error
in l05 bits) is the quantity of interest.
QPSK and OQPSK :
The optimum Eo/No performance achievable with BPSK led to a search for
mechanisms to improve the bandwidth efficiency of PSK schemes without any loss of
performance. It was found that since cos2πfct and sin2πfct (where f, is the carrier
frequency) are coherently orthogonal signals, two binary bit streams modulating the
two carrier signals in quadrature can be demodulated separately.
Extensions and Generalizations :
MSK or continuous phase FSK (CPFSK) may be generalized to include other values
of Δf, the frequency separation, and a longer bit memory before the decision has to be
made. For larger observation intervals such as 3T or 5T, a maximum improvement of
0.8 dB has been reported for Δf = 0.715/T. However, the complexity of the circuits
involved does not seem to favor these schemes over the simple yet efficient MSK
modulation.
Similarly, while retaining the advantage of good bit error rate performance, the
spectral properties of MSK can be improved by shaping the data pulses further. Note
that in MSK, the symbol pulse shape p(t) is cos[πtg(t)/2T] where g(t) = 1, 0 ≤ t ≤ T.
Other choices of g(t) are possible with the spectral falloff rate depending on the endpoint
behavior of the shape chosen. For example, a function such as g(t) =
sin(2πt/T)/(2πt/T) (known as sinusoidal frequency shift keying) results in a much
smoother p(t) and produces an asymptotic spectral falloff that is twice as fast as in
MSK. Unfortunately, all these generalizations tend to produce a broader (main lobe)
spectrum than MSK thus worsening the performance at low bandwidth/bit rate values.
MSK has been extended to multiple level pulses, known as multiple amplitude MSK
(MAMSK). Other recent works indicate that application of an efficient baseband
coding scheme such as correlative coding to MSK may be the answer to further
spectral economy and good performance.