Lý thuyết điều khiển nâng cao - Chapter 7+8

 three-stage switch in terms of the input utilization p
is
93
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Circuit Switching for Telephone Networks (14)
Example: Three-stage switch for blocking probability of 0.002 and p = 0.1
Example: Three-stage switch for blocking probability of 0.002 and p = 0.7
94
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Circuit Switching for Telephone Networks (15)
Control of a multistage space switch
Circuits are established using a signaling protocol (control plane)
Establishes state in each switch so data is forwarded correctly.
95
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Circuit Switching for Telephone Networks (16)
‰ Time-Slot Interchange (TSI) Switching
ƒ Write bytes from arriving TDM stream into memory.
ƒ Read bytes in permuted order into outgoing TDM stream.
ƒ Permutation set up by control plane when connections are set up.
96
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Circuit Switching for Telephone Networks (17)
ƒ TSI can be used as the basis for a switch.
ƒ n input ports write their frames to memory.
ƒ n output ports read frames in permuted order:
–
In each time slot they pick a frame from an input port, time slot 
pair. Permutation will typically be different for each time slot.
–
Need fast memory: Max # slots = 125 μs / (2 ×
memory cycle 
time)
ƒ Different numbers of input and output slots
–
May run at different rates.
97
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Circuit Switching for Telephone Networks (18)
‰ Time-Space-Time (TST) hybrid switch
ƒ Use TSI in first & third stage; use crossbar in middle stage.
ƒ Replace n input × k output space switch by TSI switch that takes 
n-slot input frame and switches it to k-slot output frame.
98
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Circuit Switching for Telephone Networks (19)
Flow of time slots between switches
ƒ Only one space switch active in each time slot.
99
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Circuit Switching for Telephone Networks (20)
Time-share the crossbar switch
ƒ Interconnection pattern of space switch is reconfigured every time slot.
ƒ Very compact design: fewer lines because of TDM and less space
because of time-shared crossbar.
100
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Circuit Switching for Telephone Networks (21)
Example:
101
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Chapter 8: 
Spread Spectrum Systems
102
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Spread Spectrum (SS) Techniques
The spectrum of the modulated signal is spread using pseudo-noise (PN) 
signal:
⇒Unauthorized listening can be prevented.
⇒The effect of the interfering signals at the same frequency band
 decreases.
There are two main techniques: 
ƒ Frequency-hopping spread spectrum (FH-SS): carrier frequency 
is changed randomly by using PN-signal.
ƒ Direct sequence spread spectrum (DS-SS): Information signal is 
multiplied with PN-signal before modulation.
103
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Direct Sequence Spread (DSS) Spectrum (1)
‰ DSS signals
DSS transmitter:
104
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Direct Sequence Spread (DSS) Spectrum (2)
PN binary wave:
The rectangular pulses are called chips (CPS). Each chip 
has a duration of Tc
and amplitude of ±
1 so that c2(t) = 1.
Autocorrelation and Power spectrum of c(t):
105
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Direct Sequence Spread (DSS) Spectrum (3)
Consider chipped message:
. Assumed that x(t) is ergodic
 process and independent of c(t), then
For spectral spreading Wc
>> Wx
, thus 
With practical system, the bandwidth expansion factor Wc
/Wx
is from 10 
to 10000.
DSB or DPSK modulation produces a transmitted signal requiring a
 transmission bandwidth BT
>> Wx
106
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Direct Sequence Spread (DSS) Spectrum (4)
DSS receiver:
107
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Direct Sequence Spread (DSS) Spectrum (5)
At receiver, after multiplied with local PN, we obtain:
Here, we assume perfect synchronization
of local PN generator.
In the case of z(t) is white noise, for DSB system
we obtain the 
destination signal to noise ratio (same as conventional DSB):
If the message is digital and sent via BPSK system
in white noise, we can 
use the correlation receiver as:
108
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Direct Sequence Spread (DSS) Spectrum (6)
Therefore, in the case of white noise we obtain the probability of error:
where
109
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Direct Sequence Spread (DSS) Spectrum (7)
‰ DSS performance in the presence of interference:
Let z(t) stands for a single-tone interference (or CW jammer):
where the average power at frequency fc
+ fz
. Then the in-phase 
component is:
so that:
Multiplication by c(t) spreads this spectrum:
If , then the upper bound for output interference power is:
110
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Direct Sequence Spread (DSS) Spectrum (8)
Thus, the signal to interference ratio (or signal to jamming ratio) becomes:
The bandwidth expansion factor Wc
/Wx
is also called the process gain:
which is a measure of a system’s immunity of interference.
In the case of digital transmission using BPSK:
where NJ
= J/Wc
. With SR
= Eb
rb
, then
111
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Direct Sequence Spread (DSS) Spectrum (9)
Substituting rb
= Wx
, then
The jamming margin for a minimum Pe
or minimum Eb
/NJ
is defined as:
This is a measure of a system’s ability to operate in the presence of 
interference.
If the channel is corrupted by both white noise and a interference, then:
(See Example 15.1-1, pp, 677-678, [1])
112
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Direct Sequence Spread (DSS) Spectrum (10)
‰ Multiple access:
If we are sharing the channel with M-1 other spread spectrum users (as 
is the case of Code Division Multiple Access -
CDMA), each one has 
own unique spreading code
and arrival time at receiver, then the 
interference term becomes:
where Am
, cm
(t), tm
and θm
is the signal amplitude, spreading code, time 
delay and phase, respectively, of the m-th
user. Thus
Assumed that each of other users has identical signal strengths of unit 
value, then after spreading, we obtain:
113
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Direct Sequence Spread (DSS) Spectrum (11)
In the case of BPSK system (xm
(t) = ±
1), the output of the correlation 
receiver is:
Since xm
(t) = ±
1, the integration term is the cross-correlation
between the 
desired PN code and the interferer’s PN codes. Thus, minimizing the 
cross-correlation between spreading codes minimizes the interference 
between CDMA users. 
Ideally, each PN code should be chosen to be orthogonal to the other, 
therefore making z(tk
) = 0.
114
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Direct Sequence Spread (DSS) Spectrum (12)
If M users have identical signal strength, then the bit error probability for 
M users of a CDMA channel corrupted by white noise is: 
If the channel is noiseless, then
Thus, even if the channel is noiseless, the error probability is still
non-zero 
if it contains other users.
115
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Frequency Hop Spread Spectrum (1)
Practical PN sequence generation hardware limits the increasing of 
bandwidth spreading (processing gain). 
To enable larger processing gains, the PN generator can drive a frequency 
synthesizer that produces a wideband sequence of carrier frequencies that 
can hop from one frequency to another. This process is called frequency 
hopping spread spectrum (FH-SS).
Because the message is spread out over numerous of carrier frequencies, 
the interference has a reduced probability of hitting any one.
116
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Frequency Hop Spread Spectrum (2)
FH-SS transmitter:
117
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Frequency Hop Spread Spectrum (3)
FH-SS receiver:
118
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Frequency Hop Spread Spectrum (4)
‰ FH-SS signals:
At transmitter, the message is modulated using M-ary
FSK or BPSK. 
Then modulated signal is mixed with output of a frequency 
synthesizer. The BPF selects the sum term from the mixer for 
transmission on the channel.
The receiver is the reverse of this process. Due to practical limitation 
in maintaining phase coherence, most systems use non-coherent 
detection (such as envelope detection).
119
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Frequency Hop Spread Spectrum (5)
‰ FH-SS performance in the presence of interference:
Several types of interference (jamming):
White-noise interference Partial-band interference
Single-tone interference Multiple-tone interference
120
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Frequency Hop Spread Spectrum (6)
With M-ary
FSK and non-coherent detection, the bit error probability in 
the presence of white noise is [1]:
If the interference appears a white noise
over the entire bandpass
of the 
system, then:
where NJ
= J/Wc
.
For partial interference, then:
The error probability of single-tone interference
is quite small.
121
Dept. of Telecomm. Eng.
Faculty of EEE
CS2010
BG, HCMUT
Frequency Hop Spread Spectrum (7)
In the case of CDMA system with M users and M-1 potential interferers 
(multiple interference), then the error probability is:
where Y
= 2k, k equals the number of outputs from the PN generator.