Glossary term

Inter-Carrier Interference

Engineering definition of inter-carrier interference covering OFDM orthogonality loss, normalized frequency offset, phase noise, cyclic-prefix leakage and validation evidence.

Definition

phenomenon

Inter-carrier interference is OFDM impairment caused when energy from one subcarrier leaks into other subcarriers because orthogonality is not preserved.

Inter-carrier interference, often abbreviated ICI, is the effect of OFDM subcarrier orthogonality loss. It can be caused by residual carrier frequency offset, Doppler spread, phase noise, sampling-clock error, time variation during the symbol, or energy leaking beyond the cyclic prefix. ICI raises the apparent interference floor, degrades EVM, increases BER or packet errors and can force a lower modulation-and-coding mode.

Inter-carrier interference is interference between OFDM subcarriers caused by loss of orthogonality. In an ideal OFDM receiver, each subcarrier can be demodulated without energy from its neighbors. ICI appears when that assumption is broken.

The result is not just a lower received power. ICI behaves like impairment inside the demodulator: EVM rises, constellation points blur, channel estimates become less stable, BER increases and high-order modulation can fail even when average SNR looks acceptable.

OFDM Orthogonality

In an ideal useful symbol interval, subcarriers are spaced by:

\displaystyle \Delta f_{sc}=\frac{1}{T_u}

where T_u is useful OFDM symbol time. Orthogonality depends on frequency alignment, timing, phase stability and the channel being compatible with the receiver’s cyclic-prefix and equalization assumptions.

If those assumptions are violated, the receiver’s FFT bin for one subcarrier receives leakage from other bins.

Normalized Offset Screen

Residual carrier frequency offset is often normalized by subcarrier spacing:

\displaystyle \epsilon=\frac{\Delta f_{err}}{\Delta f_{sc}}

where:

  • Delta f_err is residual frequency offset after tracking;
  • Delta f_sc is subcarrier spacing;
  • epsilon is normalized offset.

A simple guarded margin is:

M_\epsilon=\epsilon_{limit}-|\epsilon|-U_\epsilon

where U_epsilon is measurement and operating allowance. Positive margin suggests the residual offset is inside the chosen release screen.

Causes

ICI is an effect, not one root cause. Common causes include:

  • residual carrier frequency offset;
  • Doppler shift or Doppler spread;
  • phase noise across the OFDM symbol;
  • sampling-clock mismatch;
  • channel variation during the useful symbol;
  • delayed energy outside the cyclic prefix;
  • poor synchronization or tracking-loop instability.

Because the causes differ, the fix may be oscillator control, carrier recovery, pilot density, subcarrier spacing, cyclic-prefix choice, antenna/path change or a more robust mode.

Worked Example

An OFDM mode uses subcarrier spacing:

\Delta f_{sc}=15\ \text{kHz}

Residual frequency offset after tracking is:

\Delta f_{err}=210\ \text{Hz}

The normalized offset is:

\displaystyle \epsilon=\frac{210}{15000}=0.014

The release screen allows:

\epsilon_{limit}=0.020

and measurement/temperature allowance is:

U_\epsilon=0.004

The guarded margin is:

M_\epsilon=0.020-0.014-0.004=0.002

The mode barely passes this frequency-offset screen.

If mobility increases residual offset to:

\Delta f_{err}=300\ \text{Hz}

then:

\displaystyle \epsilon=\frac{300}{15000}=0.020

and:

M_\epsilon=0.020-0.020-0.004=-0.004

The same waveform now fails the guarded ICI screen even before considering phase noise, cyclic-prefix leakage or pilot-estimation error.

Engineering Interpretation

ICI can look like noise, but it is often structured. It may affect edge subcarriers, high-Doppler conditions, long packets, high-order QAM, poor oscillator states or specific mobility profiles more strongly than average SNR suggests.

For adaptive modulation, the receiver should not select a high spectral-efficiency mode only from RSSI or SNR. It should also check residual frequency offset, phase noise, EVM trend, pilot quality, cyclic-prefix margin and packet-error behavior.

The diagnostic signature often appears as a mode-dependent impairment. Robust QPSK may still decode, while 64-QAM or 256-QAM loses margin because smaller constellation spacing leaves less tolerance for subcarrier leakage. That is why ICI evidence should be tied to the selected modulation order.

Distinction From Co-Channel Interference

Co-channel interference comes from another transmitter using the same channel resource. Inter-carrier interference is internal to the OFDM demodulation process: energy leaks between subcarriers because the receiver no longer sees an orthogonal basis.

Both reduce usable SINR, but they call for different evidence and mitigation. Channel planning helps co-channel interference; synchronization, numerology, tracking and waveform changes often control ICI.

Common Mistakes

Do not treat every OFDM impairment as thermal noise. If EVM rises with residual CFO, Doppler or phase-noise conditions, the failure may be ICI-limited. Do not increase transmit power as the first response to ICI; it may leave the orthogonality problem unchanged.

Do not evaluate ICI using one static frequency-offset measurement only. Temperature, oscillator aging, mobility, loop bandwidth and burst acquisition can change the residual offset during operation.

Validation Evidence

A defensible ICI assessment should include:

  • subcarrier spacing and useful symbol time;
  • residual CFO and normalized offset;
  • phase-noise and Doppler conditions;
  • cyclic-prefix margin and delay-spread evidence;
  • pilot density and tracking-loop behavior;
  • EVM by subcarrier or resource region when available;
  • BER, packet-error and fallback behavior;
  • measurement uncertainty and temperature/mobility envelope.

With that evidence, inter-carrier interference becomes a diagnosable OFDM impairment rather than a vague label for poor receiver performance.

REF

See also