Ontological Layer: Network Economics & Sharing
Multiplexing
Many Voices Traveling as One
THE CHANNEL IS ONE. THE VOICES ARE MANY.
Dividing the Resource
A single communication link is expensive. Dedicating an entire cable or frequency band to a single user who may frequently be silent is an immense waste of infrastructure. Realm 10 solves this by slicing the physical medium into manageable domains. We can divide the spectrum into distinct bands (FDM), slice time into rapidly repeating slots (TDM), or dynamically assign capacity only when a user speaks (Statistical Multiplexing). The MUX merges; the Channel carries; the DEMUX separates.
The 16 Multiplexing Nodes
Ontological Mapping: FDM, TDM, and Statistical Allocation
The Multiplexing Problem
- Many independent users need to communicate
- Giving each user a dedicated wire is impossible
- Results in wasted bandwidth and infrastructure
- Solution: Share one high-capacity medium
MUX & DEMUX
- Multiplexer (MUX): Combines multiple streams
- Composite Signal: Travels over shared channel
- Demultiplexer (DEMUX): Separates streams at receiver
- End-to-End transparency for the user
Frequency Division (FDM)
- Different signals given different frequency bands
- All users transmit simultaneously
- Requires modulation to shift baseband to passband
- Frequency is the shared resource
Guard Bands
- Adjacent FDM channels cannot touch exactly
- Unused spectral gaps prevent interference
- Allows for imperfect practical filters
- Trade-off: Consumes valuable total spectrum
Time Division (TDM)
- Different users occupy channel at different times
- Users share the exact same frequency band
- Sequential, interleaved transmission
- Time is the shared resource
TDM Frames
- Time Slot: Defined period assigned to a source
- Frame: Repeating structured collection of slots
- Sync Word required to mark the start of the frame
- Receiver uses Sync to identify which slot belongs to whom
Synchronous TDM
- Each source gets a predetermined, fixed slot
- Highly predictable, simple structure
- Fatal Flaw: If a user has no data, the slot is sent empty
- Results in wasted (idle) channel capacity
Statistical TDM
- Slots allocated dynamically based on demand
- Inactive users do not consume channel time
- Requires address headers to identify payload
- Vastly more efficient for "bursty" computer traffic
Wavelength (WDM)
- Optical equivalent of Frequency Division (FDM)
- Multiple light colors travel down one fiber
- CWDM (Coarse) vs DWDM (Dense) spacing
- Allows terabits of capacity on existing infrastructure
Code & Space (CDM/SDM)
- CDM: Same time & freq, separated by spreading codes
- SDM: Physically separated paths (cables/antenna beams)
- Polarization (PDM): Orthogonal EM wave states
- Hybrid Multiplexing: Combining FDM + TDM + SDM
FDM vs TDM Matrix
| FDM | TDM | |
|---|---|---|
| RESOURCE | Freq | Time |
| TX | Simultaneous | Sequential |
| SEPARATION | Filters | Slots |
| OVERHEAD | Guard Bands | Sync/Frames |
- FDM: "Who gets which frequency?"
- TDM: "Who gets which time?"
- TDM requires strict network clock synchronization
- FDM requires precise analog filtering
Packet Queuing
- Statistical MUX requires buffering incoming bursts
- Scheduler decides: Who transmits? When? How much?
- Round Robin, Priority, or Weighted policies
- If Queue fills up → Delay increases → Packets drop
Overhead & Utilization
- MUX adds headers, sync words, and control bits
- Usable payload is always < total transmitted bits
- Utilization = Used Capacity / Available Capacity
- Very high utilization causes network congestion
Multiple Access
- Multiplexing: Combining streams into a shared pipe
- Multiple Access: Many independent users sharing a system
- FDMA (Frequency), TDMA (Time), CDMA (Code)
- Different applications of the same core resource division
Hierarchy
- Multiple low-speed sources → MUX → Medium-speed
- Multiple Medium-speed → MUX → High-speed backbone
- TDM hierarchical levels (e.g., T1, E1 carrier systems)
- Multiplexing does not create infinite capacity
16 — The Network Laboratory
We have successfully combined many voices onto a single physical road. But as we stuff more and more data into this shared channel, we must ask: Are we transmitting efficiently? Do we actually need to send every single bit of this data, or is the source repeating itself? You are now prepared for Realm 11: Source Coding, where we learn to mathematically compress information and eliminate redundancy.