Shannon’s theorem

Definition

Shannon’s theorem, also called the Shannon capacity theorem or Shannon–Hartley theorem for band-limited noisy channels, states the maximum error-free data rate at which information can be transmitted over a communication channel of a given bandwidth in the presence of noise.

For a channel with bandwidth B hertz and signal-to-noise ratio S/N, the theoretical maximum capacity is:

$C = B \log_2(1 + S/N)$

where:

C

= channel capacity in bits per second (bps)

B

= channel bandwidth in hertz (Hz)

S/N

= signal-to-noise ratio as a linear ratio, not in dB

This theorem gives the upper limit of reliable communication. It does not say that every system can reach this limit, but it shows the best possible performance that any digital communication system can theoretically achieve under ideal coding.

Main Content

1. Channel Capacity and Bandwidth

Channel capacity

is the highest rate at which data can be transmitted with an arbitrarily low probability of error, provided suitable coding is used.

Bandwidth

is the range of frequencies available for transmission. A larger bandwidth allows more information to be sent per second, even if noise remains the same.

The theorem shows that capacity increases with bandwidth, but not in a simple linear way forever. Since capacity depends on $\log_2(1+S/N)$ , the relationship between noise, bandwidth, and data rate is fundamental in digital communication.

For example, if a channel has:

Bandwidth = 3 kHz
Signal-to-noise ratio = 15 dB

First convert SNR from dB to linear:

$S/N = 10^{15/10} \approx 31.6$

Then:

$C = 3000 \log_2(1+31.6)$

$C \approx 3000 \times 5.03 \approx 15090 \text{ bps}$

So the theoretical maximum is about 15.09 kbps.

2. Signal-to-Noise Ratio and Reliable Transmission

Signal-to-noise ratio (SNR)

measures how strong the useful signal is compared to the unwanted noise.
A higher SNR means the receiver can distinguish the transmitted signal more easily, improving the possible data rate.

Shannon’s theorem shows that if SNR increases, capacity also increases, but with diminishing returns because of the logarithmic form. This means doubling the SNR does not double the capacity.

Example:

If SNR is very low, the capacity is small.
If SNR improves from 3 dB to 6 dB, the capacity increases, but not dramatically.
To get a major increase in capacity, both bandwidth and SNR must be considered.

A useful practical interpretation is that noise places a limit on how many different signal levels can be reliably distinguished. If the channel is too noisy, increasing the bit rate will eventually cause unacceptable errors.

3. The Shannon Limit and Practical Coding

The Shannon limit is the boundary beyond which error-free communication is impossible for a given channel.
In practice, digital systems use error-correcting codes to approach this limit.

Modern communication systems such as Wi-Fi, 4G/5G, satellite links, and optical networks use advanced coding schemes to get close to Shannon’s theoretical capacity. However, no real system can exceed it on a noisy channel with fixed bandwidth and power.

A key interpretation of the theorem is:

If the transmission rate is below capacity, reliable communication is possible in principle.
If the rate is above capacity, reliable communication is impossible no matter how sophisticated the coding.

ASCII representation of the idea:

Data Rate
  ^
  |                         Impossible region
  |                              /
  |                             /
  |                            /
  |---------------------------/----  Shannon capacity
  |                          /
  |                         /
  |        Reliable region  /
  +----------------------------------> Channel conditions

This helps explain why engineers design systems close to, but not beyond, the capacity boundary.

Working / Process

1. Identify the channel bandwidth

Determine the usable frequency range of the communication channel.
Example: a voice channel may have a bandwidth of 3000 Hz.

2. Measure or estimate the signal-to-noise ratio

Find the SNR in linear form or convert it from decibels.
Conversion from dB: $S/N = 10^{(SNR_{dB}/10)}$

3. Apply Shannon’s formula

Use: $C = B \log_2(1 + S/N)$
Compute the maximum channel capacity in bits per second.

4. Compare actual data rate with capacity

If actual rate is below capacity, reliable communication is theoretically possible.
If actual rate is above capacity, errors will increase sharply.

5. Use coding and modulation accordingly

Select modulation, coding rate, and power levels that keep the system operating near the Shannon limit without exceeding it.

Example: If:

$B = 1$ MHz
SNR = 20 dB

Then: $S/N = 10^{20/10} = 100$

$C = 10^6 \log_2(101)$

$C \approx 10^6 \times 6.658 \approx 6.658 \text{ Mbps}$

So the channel can theoretically support about 6.66 Mbps.

Advantages / Applications

Gives the ultimate limit of digital communication

, helping engineers understand what is theoretically achievable.

Guides system design

by showing the trade-off between bandwidth, power, and data rate.

Used in modern communication systems

such as cellular networks, satellite communication, wireless LANs, and optical fiber systems.

Important for error-correcting code design

, because it tells designers how close a code can come to ideal performance.

Helps optimize spectrum use

, especially in crowded wireless environments where bandwidth is limited.

Summary

Shannon’s theorem gives the maximum possible data rate of a noisy communication channel.
It depends on bandwidth and signal-to-noise ratio.
It sets a theoretical limit for reliable transmission.
Important terms to remember: channel capacity, bandwidth, signal-to-noise ratio, Shannon limit, error-correcting code