Loading learning content...
Signals can share a communication channel without interfering with each other by occupying different dimensions of the channel. Just as two people can occupy different rooms without collision, signals can occupy different frequencies, different time slots, different codes, different wavelengths, or statistically share capacity through packet-based transmission.
Each multiplexing technique exploits a different dimension of separation, with distinct tradeoffs in efficiency, complexity, latency, and suitability for different traffic types. Understanding these techniques and their tradeoffs is essential for network design and optimization.
This page provides a comprehensive survey of the major multiplexing families, establishing the foundation for the detailed exploration of each technique in subsequent modules.
By the end of this page, you will understand: the five major multiplexing dimensions (frequency, time, code, wavelength, statistical), the operating principles of each technique, their comparative advantages and disadvantages, and how to select appropriate techniques for specific applications.
Frequency Division Multiplexing (FDM) divides the available bandwidth into multiple non-overlapping frequency bands, assigning each channel to a dedicated band. All channels transmit simultaneously, but in different parts of the frequency spectrum.
Operating Principle
Each input signal is modulated onto a distinct carrier frequency, shifting it to its assigned position in the spectrum. A bandpass filter at the receiver extracts only the desired frequency band, recovering the original signal.
Key Components:
Guard Bands
To prevent interference between adjacent channels, FDM systems include guard bands—unused frequency gaps between channels. These bands account for:
Guard bands reduce spectral efficiency but are essential for reliable separation.
FDM Applications
Time Division Multiplexing (TDM) divides the channel into recurring time frames, with each frame divided into time slots. Each channel is assigned one or more slots per frame, transmitting in its allocated slots while remaining silent in others.
Synchronous TDM
In synchronous TDM, slots are pre-assigned to channels regardless of whether they have data to send. This provides guaranteed capacity but may waste bandwidth when channels are idle.
Key Characteristics:
The T1/DS1 System (Classic Example)
The T1 system, deployed in 1962, remains a foundational TDM example:
Statistical TDM (Asynchronous TDM)
To address synchronous TDM's inefficiency with bursty traffic, statistical TDM (also called asynchronous TDM or packet multiplexing) assigns slots dynamically based on demand:
Comparison: Synchronous vs. Statistical TDM
| Characteristic | Synchronous TDM | Statistical TDM |
|---|---|---|
| Slot assignment | Fixed, pre-allocated | Dynamic, on-demand |
| Addressing | Implicit (position) | Explicit (header required) |
| Efficiency | Low for bursty traffic | High for bursty traffic |
| Delay guarantee | Bounded (1 frame) | Variable (queue-dependent) |
| Complexity | Low | Higher (scheduling, buffering) |
| Oversubscription | Not possible | Common and beneficial |
| Best for | Constant-rate (voice) | Variable-rate (data) |
TDM enabled the digitization of the telephone network. By sampling analog voice at 8 kHz and quantizing to 8 bits, any voice call becomes a 64 kbps digital stream that can be TDM-multiplexed with others. This standardization (called PCM—Pulse Code Modulation) transformed telecommunications from analog artistry to digital engineering.
Code Division Multiplexing (CDM), most commonly implemented as Code Division Multiple Access (CDMA), allows multiple channels to share the entire frequency band and time simultaneously, separating them through unique mathematical spreading codes.
Operating Principle
Each channel is assigned a unique spreading code—a sequence of rapidly alternating values (called chips). The data signal is multiplied by this code, spreading it across a wide frequency band. At the receiver, multiplying by the same code recovers the original signal.
The key property: Spreading codes are chosen to be orthogonal or nearly orthogonal. When a receiver multiplies by code A, signals spread with code B average to zero and effectively disappear. Only the desired signal (spread with code A) is recovered.
Spread Spectrum Fundamentals
Example: IS-95 CDMA (2G Cellular)
CDMA is like a party where everyone speaks at once, but each conversation uses a different language. A listener who understands only French hears a clear French conversation while all other languages blur into indistinct background noise. The spreading codes are the 'languages' that separate signals.
Direct Sequence Spread Spectrum (DSSS)
The most common CDMA implementation is Direct Sequence Spread Spectrum (DSSS):
Advantages and Challenges
CDMA Applications
Wavelength Division Multiplexing (WDM) is FDM applied to optical fiber, using different colors (wavelengths) of light to carry independent data streams through the same fiber.
Operating Principle
Different wavelengths of light travel through optical fiber with minimal interaction. By using lasers of slightly different wavelengths, multiple channels can share a single fiber strand.
Key Components:
WDM Variants
| Technology | Channel Spacing | Channels per Fiber | Typical Use |
|---|---|---|---|
| CWDM (Coarse WDM) | 20 nm | 8-18 | Metro, access networks |
| DWDM (Dense WDM) | 0.8 nm (100 GHz) | 40-80 | Long-haul backbone |
| UDWDM (Ultra-Dense) | 0.2 nm (25 GHz) | 160-320+ | Submarine cables, max capacity |
Capacity Achievement
DWDM has enabled extraordinary capacity growth:
Example: Modern Submarine Cable
WDM Advantages and Challenges
When fiber was first deployed in the 1980s, each strand carried a single wavelength at ~100 Mbps. Today, that same fiber (with WDM upgrades at endpoints) can carry 10-20 Tbps—a 100,000x capacity increase without replacing the fiber itself. This is why 1980s fiber plant remains valuable today.
Statistical multiplexing allocates channel capacity dynamically based on actual demand rather than pre-assigning resources. Data is transmitted in discrete packets that compete for channel access, with unused capacity immediately available to active sources.
Operating Principle
Unlike the previous techniques that divide capacity in advance, statistical multiplexing:
This is the fundamental model of the modern Internet. All IP-based networks use statistical multiplexing at packet level.
Key Characteristics
Comparison with Fixed Multiplexing
| Aspect | Fixed (FDM/TDM) | Statistical (Packet) |
|---|---|---|
| Capacity allocation | Pre-assigned per source | On-demand per packet |
| Peak rate per source | Limited to allocated share | Can use full channel rate |
| Efficiency | Low for bursty traffic | High for bursty traffic |
| Delay guarantee | Bounded | Variable, potentially unbounded |
| Overload behavior | Impossible (hard limit) | Queuing then loss |
| Complexity | Low | Higher (scheduling, buffering) |
| Best traffic type | Constant rate (voice, legacy) | Variable rate (data, modern) |
Quality of Service in Statistical Systems
Pure statistical multiplexing provides best-effort service—no guarantees. To support quality-sensitive applications, networks implement QoS mechanisms:
Statistical Multiplexing Applications
The Internet was designed for data traffic, which is inherently bursty. Statistical multiplexing handles burstiness naturally—when one user isn't transmitting, others immediately get more bandwidth. This flexibility enables the Internet to support everything from occasional email to 4K streaming on the same infrastructure.
Modern communication systems rarely use a single multiplexing technique. Instead, they combine multiple techniques in hybrid or multi-dimensional schemes that exploit the advantages of each.
OFDM: Frequency + Time
Orthogonal Frequency Division Multiplexing (OFDM) divides spectrum into many narrow subcarriers, each modulated with part of the data. It combines the robustness of FDM with efficient use through orthogonal subcarrier spacing.
OFDMA: OFDM + Multiple Access
OFDM Multiple Access assigns different subcarriers to different users:
MIMO: Spatial Multiplexing
Multiple Input Multiple Output (MIMO) uses multiple antennas to create parallel spatial channels:
| Dimension | Technique | Example Application |
|---|---|---|
| Frequency | FDM/OFDM | Subcarrier assignment |
| Time | TDM | Time slot allocation |
| Code | CDMA | 2G/3G cellular, GPS |
| Wavelength | WDM/DWDM | Optical fiber networks |
| Space | MIMO | 5G Massive MIMO |
| Polarization | Dual-polarization | Satellite, advanced optical |
5G Multi-Dimensional Resource Allocation
5G NR (New Radio) exemplifies the hybrid approach:
The scheduler allocates resource blocks (time-frequency units) to users, with spatial layers multiplied through MIMO. The result is spectral efficiency of 30+ bps/Hz—100× better than first-generation cellular.
Flex Ethernet and FlexGrid
Modern network standards support flexible multiplexing:
A single communication might use all dimensions simultaneously: wavelength (WDM on fiber backbone), frequency (OFDM subcarriers on wireless), time (slot scheduling), space (MIMO layers), and statistical (IP packet switching). Each layer optimizes a different resource, combining for extraordinary overall efficiency.
Selecting the appropriate multiplexing technique depends on traffic characteristics, performance requirements, available technology, and cost constraints. This guide summarizes key decision factors.
Traffic Characteristics
| Traffic Type | Characteristics | Recommended Technique | Why |
|---|---|---|---|
| Voice (traditional) | Constant 64 kbps, delay-sensitive | Synchronous TDM | Guaranteed timing, proven |
| Video (streaming) | Constant rate, tolerates buffering | Statistical with QoS | Efficiency with prioritization |
| Video (conferencing) | Variable rate, delay-sensitive | TDM or priority QoS | Needs bounded delay |
| Web traffic | Very bursty, delay-tolerant | Statistical | Efficiency for burstiness |
| IoT sensors | Low rate, many devices | Statistical or CDMA | Many sources, low overhead |
| Fiber backbone | Aggregate, high capacity | WDM | Maximum capacity scaling |
Performance Requirements
Technology and Cost Considerations
Environment Factors
Modern systems don't choose one technique—they layer multiple techniques. Your phone call's path might include: OFDMA on the air interface, DWDM on the backhaul fiber, statistical switching through the core, and TDM through legacy telephony gateways. Each segment uses what's optimal for its specific constraints.
We've surveyed the major multiplexing techniques and their characteristics. Let's consolidate the key insights:
What's Next:
With this overview of multiplexing techniques complete, the next page explores real-world applications—how these techniques are deployed in actual systems from cellular networks to submarine cables to data centers.
You now have a comprehensive understanding of the major multiplexing techniques—FDM, TDM, CDM, WDM, and statistical multiplexing—along with their tradeoffs and selection criteria. This foundation prepares you for detailed study of each technique and their applications.