Ontological Layer: The I/Q Dimension
Digital Modulation II
PSK & QAM: Multiplying Information in Phase Space
The Grid of Symbols
Simple amplitude and frequency shifting are inefficient. To truly maximize spectral efficiency, digital information must be folded into the orthogonal dimensions of In-Phase (I) and Quadrature (Q) carriers. By combining Phase Shift Keying (PSK) and Quadrature Amplitude Modulation (QAM), we map binary data onto dense, two-dimensional constellation grids. Realm 08 defines how increasing the number of points multiplies our data rate, but critically narrows the Euclidean distance between them—making our symbols highly vulnerable to the crushing effects of Additive White Gaussian Noise.
The 16 Constellation Nodes
Ontological Mapping: From Phase Shifts to 256-QAM
Phase as Information
- Amplitude and frequency kept constant
- Digital symbols mapped to discrete phase states
- c(t) = A_c cos(2πf_ct + φ)
- Forms the basis of the PSK family
BPSK Constellation
- Binary Phase Shift Keying (2 states)
- State 0 → 180° phase (-A)
- State 1 → 0° phase (+A)
- Points lie on opposite sides of the origin
BPSK Waveform
- Continuous carrier undergoes sharp phase flip
- s_1(t) = A_c cos(2πf_ct)
- s_0(t) = -A_c cos(2πf_ct)
- Requires coherent detection at receiver
Decision Boundary
- The geometric line separating valid symbols
- Receiver decides based on which side the sample lands
- Noise can physically push a signal across the line
- Result: Symbol Error
QPSK Concept
- Quadrature Phase Shift Keying
- Four equally spaced phase states
- log₂(4) = 2 bits per symbol
- Double data rate of BPSK for same bandwidth
I/Q Space Representation
- s(t) = I(t)cos(ω_ct) - Q(t)sin(ω_ct)
- I(t) = In-Phase component
- Q(t) = Quadrature component (90° offset)
- Orthogonal carriers do not interfere
Gray Coding
- Assigns binary labels to constellation points
- Adjacent points differ by only ONE bit
- If noise pushes symbol to neighbor, only 1 bit is lost
- Minimizes bit error rate for a given symbol error rate
M-PSK Scaling
- Generalization to M phase states on a circle
- k = log₂(M) bits per symbol
- As M increases, points get closer together
- Higher spectral efficiency vs higher noise sensitivity
Differential PSK (DPSK)
- Information encoded in phase CHANGE between symbols
- Receiver compares current symbol to previous
- Eliminates need for absolute carrier phase reference
- Cost: Slight performance/SNR penalty
QAM Principle
- Quadrature Amplitude Modulation
- Combines Multiple I amplitudes + Multiple Q amplitudes
- Creates a 2D grid instead of a circle
- Vastly increases possible symbol states
16-QAM to 256-QAM
- 16-QAM: 4x4 grid (4 bits/symbol)
- 64-QAM: 8x8 grid (6 bits/symbol)
- 256-QAM: 16x16 grid (8 bits/symbol)
- R_s = R_b / log₂(M) (Symbol Rate vs Bit Rate)
Euclidean Distance (d_min)
- d = √[(I₁−I₂)² + (Q₁−Q₂)²]
- d_min: Smallest distance between any two points
- Major measure of constellation robustness
- Fixed power + more points = smaller d_min
AWGN Noise Clouds
- Additive White Gaussian Noise creates spread
- High SNR = tight clouds, easy decisions
- Low SNR = wide clouds, overlapping decision regions
- E_s/N_0 metric dictates error probability
Phase Offset & Impairments
- Phase offset rotates the entire constellation
- Requires Carrier Phase Recovery loop
- Freq Offset = Continuous spinning
- Amplitude distortion = points pushed inward/outward
Symbol vs Bit Error
- Symbol Error: Receiver picks wrong constellation point
- Bit Error: Specific 1s and 0s that were flipped
- Soft Decisions vs Hard Decisions
- Why E_b/N_0 is used for fair comparisons
16 — The Constellation Laboratory
By mastering Phase Shift Keying and Quadrature Amplitude Modulation, we have learned how to pack massive amounts of information into a single analog symbol. But the receiver is currently blind—it receives only a noisy, distorted wave. It must somehow answer: "Where is the symbol?" and "Which symbol was sent?". You are now prepared for Realm 09: Digital Receivers & Detection.