Geometric Representation of Bandpass Signals
M -ary Amplitude Shift Keying (Pulse Amplitude Modulation)
Transmitted Signals:
For m {0, 1, . . . , M 1}
Am cos (2fc t) , for 0 t T
sm (t) = Am = A(M 1 2m)
0 , elsewhere
It is useful to express sm (t) in terms of a unit energy rectangular pulse,
1
, for 0 t T
hT (t) = T
0 , elsewhere
so
sm (t) = Am T hT (t) cos (2fc t)
Basis Signals:
Only one basis signal is required to represent {sm (t) | m = 0, 1, . . . , M 1}:
0 (t) = hT (t) 2 cos (2fc t)
Note: 0 (t) is normalized:
2 2 2
0 (t) = 0 (t) dt = hT (t) 2 cos (2fc t) dt
T T
1 2 2 t sin (4fc t)
= (2) cos (2fc t) dt = +
0 T T 2 8fc 0
2 T sin (4fc T ) sin (4fc T )
= + =1+
T 2 8fc 4fc T
= 1
n
since fc T 1 (e.g., fc = 1Ghz, T = 0.1ms, fc T = 105 ). Equality holds if fc T = 4 for some integer n.
Signals:
For m {0, 1, . . . , M 1}
T
sm (t) = Am 0 (t)
2
= sm,0 0 (t)
with
T
sm,0 = Am
2
Signal Space Diagram (M =4):
s3 s2 s1 s0
-
0 (t)
3A T2 A T2 0 A T2 3A T2
Equivalent lowpass signal:
sm (t) = Am T hT (t) cos (2fc t)
= Am T hT (t)Re ej2fc t
T j2f t
= Re Am hT (t) 2e c
2
= Re sl,m (t) 2ej2fc t
where
T
sl,m (t) = Am hT (t)
2
SYSC 3503 1 Winter 2010/11
is the equivalent lowpass signal for sm (t). Also, dene
T
sm = Am
2
so
sl,m (t) = sm hT (t)
M -ary Phase Shift Keying
Transmitted Signals:
For m {0, 1, . . . , M 1}
2
sm (t) = A T hT (t) cos (2fc t + m ) m = m
M
Basis Signals:
Note: sm (t) can also be written as
sm (t) = A T hT (t) cos m cos (2fc t) A T hT (t) sin m sin (2fc t)
Note: cos(2fc t) and sin(2fc t) are orthogonal over [0, T ]:
T
cos(2fc t), sin(2fc t) = cos(2fc t) sin(2fc t) dt
0
T
sin2 (2fc t)
=
4fc 0
sin2 (2fc T )
=
4fc
= 0
since fc is typically very large.
Therefore, only two basis signals are required to represent {sm (t) | m = 0, 1, . . . , M 1}:
0 (t) = hT (t) 2 cos (2fc t)
1 (t) = hT (t) 2 sin (2fc t)
Signals:
For m {0, 1, . . . , M 1}
T T
sm (t) = A cos m 0 (t) + A sin m 1 (t)
2 2
= sm,0 0 (t) + sm,1 1 (t)
with
T T
sm,0 = A cos m sm,1 = A sin m
2 2
SYSC 3503 2 Winter 2010/11
Signal Space Diagram:
M =4 M =8
1 (t) 1 (t)
s1 6 s2 6
s3 s1
s s0 s s0
2 - 4 -
0 (t) 0 (t)
A T2 q A T2 q
s5 s7
s3 s6
? ?
Quarternary Phase Shift Keying (QPSK)
Equivalent lowpass signal:
sm (t) = A T hT (t) cos (2fc t + m )
= A T hT (t)Re ej(2fc t+m )
T jm j2f t
= Re A e hT (t) 2e c
2
= Re sl,m (t) 2ej2fc t
where
T jm
sl,m (t) = A e hT (t)
2
is the equivalent lowpass signal for sm (t). Also, dene
T jm
sm = A e
2
so
sl,m (t) = sm hT (t)
Quadrature Amplitude Modulation (QAM)
Because cos(2fc t) and sin(2fc t) are orthogonal, it is possible to double the capacity of PAM.
Transmitted Signals:
For m {0, 1, . . . , M 1}
sm (t) = Ac,m T hT (t) cos (2fc t) As,m T hT (t) sin (2fc t)
where Ac,m is the amplitude of the cos carrier (the in-phase component), and As,m is the amplitude of the sin
carrier (the quadrature phase component).
Basis Signals:
0 (t) = hT (t) 2 cos (2fc t)
1 (t) = hT (t) 2 sin (2fc t)
SYSC 3503 3 Winter 2010/11
Signals:
For m {0, 1, . . . , M 1}
T T
sm (t) = Ac,m 0 (t) + As,m 1 (t)
2 2
= sm,0 0 (t) + sm,1 1 (t)
with
T T
sm,0 = Ac,m sm,1 = As,m
2 2
Signal Space Diagram:
M =4
1 (t)
s1 6 s0
A T2
- 0 (t)
A T2 A T2
s2 A T2 s3
?
M = 16
1 (t)
s12 s13 6 s14 s15
3A T2
s8 s9 s10 s11
A T2
- 0 (t)
3A T2 A T2 A T2 3A T2
A T2
s4 s5 s6 s7
3A T2
s0 s1 s2 s3
?
SYSC 3503 4 Winter 2010/11
M =8
1 (t)
s4 s5 6 s6 s7
T
A 2
- 0 (t)
3A T2 A T2 A T2 3A T2
A T2
s0 s1 s2 s3
?
Equivalent lowpass signal:
sm (t) = Ac,m T hT (t) cos (2fc t) As,m T hT (t) sin (2fc t)
= Re [Ac,m + jAs,m ] T hT (t) [cos(2fc t) + j sin(2fc t)]
T j2f t
= Re [Ac,m + jAs,m ] hT (t) 2e c
2
= Re sl,m (t) 2ej2fc t
where
T
sl,m (t) = [Ac,m + jAs,m ] hT (t)
2
is the equivalent lowpass signal for sm (t). Also, dene
T
sm = [Ac,m + jAs,m ]
2
so
sl,m (t) = sm hT (t)
M -ary Frequency Shift Keying (FSK)
Transmitted Signals:
For m {0, 1, . . . , M 1}
sm (t) = A T hT (t) cos (2fm t) fm = fc + mfc
Basis Signals:
One basis signal is required
for each transmitted signal:
m (t) = hT (t) 2 cos (2fm t)
For proper operation, the transmitted signals should all be orthogonal:
k (t), l (t) = k (t)l (t) dt
= hT (t) 2 cos(2fk t)hT (t) 2 cos(2fl t) dt
2
= hT (t) 2 cos(2fk t) cos(2fl t) dt
T
2
= cos(2fk t) cos(2fl t) dt
T 0
T
2 sin(2[fk fl ]t) sin(2[fk + fl ]t)
= +
T 4[fk fl ] 4[fk + fl ] 0
2 sin(2[fk fl ]T ) sin(2[fk + fl ]T )
= +
T 4[fk fl ] 4[fk + fl ]
sin(2[fk fl ]T ) sin(2[fk + fl ]T )
= +
2[fk fl ]T 2[fk + fl ]T
SYSC 3503 5 Winter 2010/11
sin(2fc [k l]T ) sin(2[2fc + fc (k + l)]T )
= +
2fc [k l]T 2[2fc + fc (k + l)]T
sin(2fc [k l]T )
=
2fc [k l]T
To ensure orthogonality, fc T must be an integer multiple of 1/2.
Signals:
For m {0, 1, . . . , M 1}
T
sm (t) = A m (t)
2
= sm,m m (t)
with
T A T2 , if m = n
sm,n = A mn =
2 0 , otherwise
Signal Space Diagram (M = 2):
1 (t)
6
A T2 s1
s0
- 0 (t)
A T2
Equivalent lowpass signal:
sm (t) = A T hT (t) cos (2fm t)
= A T hT (t)Re ej2(fc +fc m)t
T j2mfc t j2f t
= Re A e hT (t) 2e c
2
= Re sl,m (t) 2ej2fc t
where
T j2mfc t
sl,m (t) = A e hT (t)
2
is the equivalent lowpass signal for sm (t).
SYSC 3503 6 Winter 2010/11
