Contents

Digital Communication
Other Subjects
Section Progress23%

3 of 13 articles

RZ Line Coding

Return to zero, wider bandwidth, self-clocking properties.

Darshan N
Updated: 19 March 2026
11 min read

Return-to-Zero (RZ) line coding is a baseband signaling scheme in which the signal level returns to zero within every bit period, regardless of what the bit value is. This mandatory zero crossing within each bit interval is what fundamentally distinguishes Return-to-Zero (RZ) from NRZ coding. The guaranteed transitions in RZ carry clocking information embedded within the signal itself, which is a significant practical advantage in synchronous communication systems.

RZ Line Coding VariantsData Bits:101101Unipolar RZ:+V0VPolar RZ:+V0V-VKey Properties:Bandwidth = 2 x R_bSelf-clocking: YESDC component: YES (unipolar)Zero crossing each bit periodGuaranteed transition per bitPulse width = T_b / 2Higher BW than NRZUsed in optical systemsBetter clock recoveryPulse returns to 0V at half-bit period in every bit slot
Figure 1: RZ encoding for bit sequence 1 0 1 1 0 1 showing mandatory zero crossings and half-bit-period pulse widths

Core Concept of RZ Encoding

The fundamental rule of RZ encoding is that the signal must return to zero during every bit period. In Unipolar RZ, bit 1 is represented by a positive pulse that occupies the first half of the bit period, followed by zero for the second half. Bit 0 is represented by zero for the entire bit period. In Polar RZ, bit 1 uses a positive pulse in the first half and bit 0 uses a negative pulse in the first half, both returning to zero in the second half.

The mandatory return to zero within each bit period means that a transition always occurs at the midpoint of every bit slot that carries a 1 in unipolar RZ, or at every bit slot in polar RZ. This guaranteed transition pattern allows the receiver to extract clock timing directly from the received data waveform without requiring a separate clock signal. This property is called self-clocking and is the primary advantage of RZ over NRZ.

However, RZ encoding uses shorter pulses than NRZ for the same bit rate. A pulse that is only half the bit period wide occupies a wider frequency band. The spectral content spreads out more because narrower time-domain pulses correspond to wider frequency-domain representations. This means RZ requires twice the bandwidth of NRZ for the same data rate, which is its main disadvantage in bandwidth-limited systems.

Mathematical Expression and Bandwidth

The PSD of an RZ signal is also sinc-squared shaped, but the argument is based on the pulse width rather than the full bit period. Since the pulse width in RZ is T_b/2 (half the bit period), the first null of the spectrum occurs at 2/T_b, which equals 2R_b. Therefore:

B_RZ = 2 R_b is the null-to-null bandwidth of RZ. This is exactly twice the bandwidth of NRZ for the same bit rate. The spectral efficiency of RZ is 0.5 bps/Hz, compared to 1 bps/Hz for NRZ. This trade-off is the cost of achieving self-clocking. The Fourier transform of a rectangular pulse of width tau gives a sinc spectrum with first null at 1/tau. For RZ, tau = T_b/2, so first null = 1/(T_b/2) = 2/T_b = 2R_b.

In polar RZ, the DC component is theoretically zero when data has equal probability of 1s and 0s, because positive and negative pulses average out. In unipolar RZ, however, the DC component is +V/4 on average (positive pulse of height V for half the duration, half the time). This DC component limits use in AC-coupled transformer-based systems.

Practical Understanding of RZ

The self-clocking property of RZ makes it well-suited for optical fiber communication systems. In fiber optic links, the transmitter emits light pulses where a pulse represents bit 1 and no light represents bit 0. Optical RZ (OOK-RZ) ensures that there is always a falling edge at the midpoint of a bit-1 period, from which the clock can be extracted using a phase-locked loop at the receiver.

Another practical consideration is pulse shaping in RZ systems. Rather than using hard rectangular pulses, real systems use raised cosine or Gaussian-shaped pulses to reduce inter-symbol interference and limit spectral spread. The fundamental idea of returning to zero is retained, but the exact waveform shape is optimized for the channel.

One important variant is Alternate Mark Inversion RZ (AMI-RZ), used in early T1 telephone carrier systems. Here, successive 1s alternate between positive and negative half-duration pulses, while 0s produce no signal. This eliminates the DC component and adds built-in single-bit error detection capability, since two consecutive same-polarity pulses indicate an error.

Numerical Example

The bandwidth of RZ being twice that of NRZ is a frequently tested relationship. Given bit rate, one can compute required channel bandwidth directly. The relationship also appears in multi-level signaling questions where baud rate differs from bit rate, but for binary RZ the formula is straightforward.

Example
Given:
Bit rate R_b = 4 Mbps
Scheme: Unipolar RZ
Voltage: +3V pulse, 0V otherwise

Why this formula applies:
RZ pulse width = T_b/2, so first null frequency = 2R_b

Formula:
B_RZ = 2 x R_b

Substitution:
B_RZ = 2 x 4 x 10^6

Calculation:
B_RZ = 8 x 10^6 Hz

Final Answer:
Minimum null-to-null bandwidth required = 8 MHz
(Twice the bandwidth of NRZ at same bit rate of 4 Mbps)
Exam Tip: RZ bandwidth is always 2 x R_b and NRZ bandwidth is R_b. If a GATE question asks which scheme needs more bandwidth for same bit rate, the answer is always RZ. Also, RZ is self-clocking but NRZ is not. These two facts are among the most commonly tested comparison points.

RZ Mechanism Summary

  • Signal returns to zero voltage in every bit period. This return happens at the midpoint of the bit period for active pulses.
  • Unipolar RZ: Bit 1 = +V for first T_b/2, then 0. Bit 0 = 0 for entire T_b. Self-clocking only for 1s.
  • Polar RZ: Bit 1 = +V first half, then 0. Bit 0 = -V first half, then 0. Every bit produces a transition. Fully self-clocking.
  • Bandwidth = 2 R_b. Spectral efficiency = 0.5 bps/Hz. Requires twice the channel bandwidth of NRZ for same data rate.
  • Self-clocking advantage: Clock recovery via PLL is straightforward because transitions are guaranteed in every bit period (polar RZ) or every 1 bit (unipolar RZ).
  • Used in optical fiber OOK-RZ modulation, AMI-RZ in T1 systems, and disk storage read channels.

Quick Revision

  • RZ returns signal to zero within every bit period. Pulse width is T_b/2 for active levels.
  • Bandwidth formula: B_RZ = 2 R_b. Always twice the bandwidth of NRZ for same bit rate.
  • RZ is self-clocking. Transitions guaranteed in every bit period (polar) or every 1-bit (unipolar) allow clock recovery.
  • Unipolar RZ has DC component (+V/4 on average). Polar RZ has zero DC for balanced data.
  • Spectral efficiency of RZ = 0.5 bps/Hz versus 1 bps/Hz for NRZ.
  • Exam trap: Do not say RZ has same bandwidth as NRZ. RZ always needs double the bandwidth. This is a very common GATE distractor.
  • AMI-RZ variant eliminates DC and adds error detection by alternating pulse polarity for successive 1s.

RZ Line Coding Quiz

Test your knowledge of RZ coding bandwidth requirements and self-clocking properties.

Question 1 of 3

Q1.Unipolar RZ coding uses pulses of duration Tb/2. Compared to unipolar NRZ at the same bit rate Rb, the first-null bandwidth of unipolar RZ is: