A Primer on Zadoff Chu Sequences
Listen
Radio episode about this paper
Transcript
Introduction to the show: ident: Computer Science. The day in computer science: software, systems and theory on arXiv.
Tom: We're looking at the paper: "A Primer on Zadoff Chu Sequences".
Paper summary: Jane: We've covered what ZC sequences are, but let's look closer at what this primer actually claims about them in terms of their overall importance. The paper is laying out the groundwork for understanding why these sequences have become so important in modern wireless communications.
Tom: The thesis of this paper is that ZC sequences are a really important manifestation of spread spectrum, and they have to some extent displaced PN and Walsh sequences which were the mainstays of 3G cellular systems like WCDMA and cdma2000, as well as the older 2G-era IS-ninety-five <ref:2211.05702#pg1,they have to some extent displaced PN and Walsh sequences which were>.
Lu: So it’s positioning them as successors or at least important alternatives to those older codes that were dominant in previous generations.
Meng: That makes sense because we're constantly looking for better ways to handle interference and spreading, and these sequences offer a new mathematical structure for that.
Jane: The paper claims that ZC sequences are complex sequences with unit amplitude and particular phase shifts, which is a key difference from the real, binary values of PN or Walsh codes.
Tom: And the paper immediately sets up the core idea: because of these properties, they are used for many key functions in current cellular systems, and they are expected to be prevalent in future ones too.
Jane: Specifically, it points out that in LTE and 5G NR, they are widely used for a number of important initial access and overhead channel functions that engineers often overlook when focusing only on data transmission <ref:2211.05702#pg1,in LTE and 5G NR, they are widely used for a number>.
Lu: It highlights their use in initial access both in the downlink synchronization sequences and the uplink random access premables as well.
Meng: That means they're involved right at the start of establishing a connection or getting your signal received, which is critical for reliability.
Tom: Plus, they are also used for transmitting uplink control information, which is another area where they shine.
Jane: So what does this mean for the world? It means that these specific complex sequences provide a very robust foundation for the signaling and control aspects of modern wireless networks.
Lu: They offer a structured way to handle those crucial early stages of communication that are often overlooked in typical data transmission analysis.
Conclusion: Tom: So, looking at the paper, "A Primer on Zadoff Chu Sequences," the authors are primarily doing an overview and primer on this topic. It's not a deep dive into a specific new algorithm or a massive experimental result.
Jane: That’s right; it't more about establishing the foundational knowledge about what these sequences are and why they matter in the long run for cellular infrastructure development.
Lu: The paper lays out the definition of the ZC sequence, and then it goes on to clearly lay out those key properties, like constant amplitude, zero cyclic autocorrelation, and fixed cyclic cross correlation.
Meng: It’s giving engineers a clear reference point for understanding the structure without getting bogged down in complex math.
Tom: The implication is that these properties are so strong that they are likely to keep them relevant across different generations of cellular technology, meaning we should expect to see them in future systems as well.
Jane: So the final thought is that ZC sequences aren't just a temporary trend; they seem like a stable and important component for the long-term evolution of cellular standards.
Jeffrey G. Andrews
University of Texas at Austin
cs.IT, math.IT
Submitted: 2022-11-10
Updated: 2026-11-17
Comments: Tutorial article, not submitted for publication
License: http://creativecommons.org/licenses/by/4.0/
The gist: Zadoff-Chu (ZC) sequences are an important manifestation of spread spectrum in modern cellular systems…
Key concepts
- Zadoff-Chu Sequence Definition
- A ZC sequence is defined by the formula sq[n] = exp(-jπq n(n + 1) / Nzc), where Nzc must be an odd number, often prime. This formula generates a complex sequence with unit amplitude and specific phase shifts based on the index q and length Nzc.
- Constant Amplitude Property
- ZC sequences maintain a constant amplitude of 1 across all samples. This is beneficial because it simplifies implementation, as the peak-to-average power ratio is 1, which helps limit the Peak-to-Average Power Ratio (PAPR) of the resulting continuous signal.
- Zero Cyclic Autocorrelation
- The cyclic autocorrelation of a ZC sequence is optimal, meaning it is zero for all non-zero shifts. This property allows for excellent correlation characteristics, which are crucial for efficient signal processing in communication systems.
- Application in Cellular Standards
- ZC sequences are used in LTE and 5G NR for functions like initial downlink synchronization and random access. Examples include using Nzc = 63 sequences with specific q values for the Primary Synchronization Sequence (PSS) in LTE.
Terminology
Summary
Zadoff-Chu (ZC) sequences are an important manifestation of spread spectrum in modern cellular systems, including LTE and 5G NR They have to some extent displaced PN and Walsh sequences which were the mainstays of 3G cellular (WCDMA and cdma2000) and the 2G-era IS-95 ZC sequences are complex sequences with unit amplitude and particular phase shifts, as opposed to Walsh and PN codes which are real and binary valued, most commonly ±1 Because of these properties, they are used for many key functions in current cellular systems, and are likely to be prevalent in future cellular systems as well
Definition and Mathematical Formulation
A Zadoff-Chu sequence has two key parameters, (i) the root index q = 1, 2..., Nzc − 1 and (ii) the length of the sequence, Nzc, which must be an odd number, and is often a prime number Given these two parameters, the q th ZC sequence sq[n] is defined as sq[n] = exp −jπq n(n + 1) / Nzc (1) where n = 0, 1, 2. p. 0 , Nzc − 1 and (ii) the length of the sequence, Nzc, which must be an odd number, and is often
a prime number. Given these two parameters, the q
th ZC sequence sq[n] is defined as
sq[n] = exp
−jπq
n(n + 1)
Nzc
(1)
J.Where the paper says it
Key Sequence Properties
ZC sequences have several highly desirable properties. These properties include:
** Property 1: Constant Amplitude:**
ZC sequences have a constant amplitude of 1 and only the phase changes from sample to sample. This is desirable for several implementation reasons, for example the peak to average power ratio of the sequence is also 1, and also limits the PAPR of the eventual continuous time signal as well. p. 2 Obviously, from (1) since they are unit amplitude complex
numbers, all values of sq[n] have a constant amplitude of 1 and only the phase changes from sample
to sample. This is desirable for several implementation reasons, for example the peak to average power
ratio of the sequence is also 1, and also limits the PAPR of the eventual continuous time signal as well.Where the paper says it
** Property 2: Zero Cyclic Autocorrelation:**
The cyclic autocorrelation of a ZC sequence is optimal, in that it is zero for all nonzero shifts of the sequence. Namely, the unnormalized cyclic autocorrelation function of sq[n] is Nzcδ[τ], where τ ∈ Z is the cyclic shift. p. 2 The cyclic autocorrelation of a ZC sequence is optimal,
in that it is zero for all nonzero shifts of the sequence. Namely, the unnormalized cyclic autocorrelation
function of sq[n] is Nzcδ[τ ], where τ ∈ Z is the cyclic shift.Where the paper says it
** Property 3: Fixed Cyclic Cross-Correlation:**
For any two distinct ZC sequences of the same length – i.e. the same value of Nzc but one having root index q = q1 while the other with root index q2 ≠ q1 – the normalized cyclic cross correlation is exactly 1/√Nzc. p. 2 For any two distinct ZC sequences of the same length
– i. the same value of Nzc but one having root index q = q1 while the other with root index q2 ̸= q1
– the normalized cyclic cross correlation is exactly 1/
√
Nzc. the same value of Nzc but one having root index q = q1 while the other with root index q2 ̸= q1
– the normalized cyclic cross correlation is exactly 1/
√
Nzc. This assumes that Nzc is prime, or more
generally that |q1 − q2| is relatively prime to Nzc, i. This assumes that Nzc is prime, or more
generally that |q1 − q2| is relatively prime to Nzc, i. the only positive integer evenly dividing Nzc
and |q1 − q2| is 1.Where the paper says it
Where the paper says it
Where the paper says it
** Property 4: The DFT or IDFT of a ZC Sequence is a ZC Sequence:**
Since the DFT of a sequence x[n] is a sum of complex exponentials with rotating phase shifts weighted by the sequence x[n], if the sequence x[n] is a ZC sequence, which itself is a rotating sequence of phase shifts, the result is also a ZC sequence.
Handling Non-Prime Lengths
Most commonly, this means the sequences are of length Nzc wherein Nzc is a prime number. An arbitrary length ZC sequence can be created from a prime length one using either cyclic extension or truncation, however the above important Properties 2 and 3 no longer precisely hold, although the loss may be tolerable. p. 3 For example in many practical cases the desired spreading factor
will not be a prime number. An arbitrary length ZC sequence can be created from a prime length one
using either cyclic extension or truncation, however the above important Properties 2 and 3 no longer
precisely hold, although the loss may be tolerable. An arbitrary length ZC sequence can be created from a prime length one
using either cyclic extension or truncation, however the above important Properties 2 and 3 no longer
precisely hold, although the loss may be tolerable. Confusingly, such truncated or extended ZC sequences are often still called ZC sequences by practicing engineers (and their wireless standards), even though they do not satisfy the above theoretical
properties and are not actually ZC or CAZAC sequences.Where the paper says it
Where the paper says it
Application in Cellular Standards
ZC sequences are utilized in LTE and 5G NR for many key functions. Examples of their use include:
** Initial downlink synchronization:**
The Primary Synchronization Sequence (PSS) in LTE is based on a Nzc = 63 sequence, and values q = 29, 34, 25 are used (a different value of q for each sector, usually).
** Random access:**
The Physical Random Access Channel (PRACH) uses Nzc = 839 for the long PRACH preamble and Nzc = 139 for the short PRACH preamble, in both LTE and 5G NR.
** Uplink control information:**
ZC sequences (length 12, extended from Nzc = 11 sequences as in the above example) are used for all but format 3.
Comparison to Other Codes
PN sequences are probably the most commonly used sequences for spread spectrum, and they have a higher autocorrelation (in discrete time), in addition. Walsh codes are orthogonal codes, meaning they have ideal (zero) cross-correlation, providing Nsf such orthogonal codes for spreading factor Nsf, where Nsf is necessarily a power of two. p. 6 Walsh codes are another commonly used spreading code. They are orthogonal codes, meaning they
have ideal (zero) cross-correlation, providing Nsf such orthogonal codes for spreading factor Nsf, where
Nsf is necessarily a power of two. ZC sequences consist
of particular phase shifts of a unit amplitude complex exponential. In this sense, ZC sequences are more
closely related to the otherwise obscure Complementary Code Keying (CCK) modulation used in the
first commercially successful WiFi standard, IEEE 802.Where the paper says it
Where the paper says it
Specific Sequence Examples
Example 1 shows a Length 5 ZC sequence with q = 1. Example 3 illustrates that the normalized shifted cross correlation of s1[n] and s4[n] can readily be computed to indeed be 1/√5 = 0. p. 4 Meanwhile, the sequence for q = 4 can be found to be
s4[0] = exp(0) = 1
s4[1] = exp(j2π/5)
s4[2] = exp(−j4π/5)
s4[3] = exp(j2π/5)
s4[4] = exp(−j4π) = 1
Needless to say, this sequence has the same cyclic autocorrelation property as for q = 1. Although not
very obviously so, the normalized shifted cross correlation of s1[n] and s4[n] can readily be computed
to indeed be 1/
√
5 = 0. Although not
very obviously so, the normalized shifted cross correlation of s1[n] and s4[n] can readily be computed
to indeed be 1/
√
5 = 0. 4472 for all possible such shifts. Their normalized autocorrelations are plotted in Fig. 1,
where it can be seen that the autocorrelation for z4[n] is much worse than for z1[n].Where the paper says it
Where the paper says it
Where the paper says it
Sequence Generation Methods
To create a cyclically extended ZC sequence of arbitrary length Nsf, using (1), compute zq[n] = sq[n mod Nzc] for n = 0, 1..., Nsf − 1 and where Nzc is the largest prime number less than or equal to Nsf. p. 3 To create a cyclically extended ZC sequence of arbitrary length Nsf, using (1), compute
zq[n] = sq[n mod Nzc]
for n = 0, 1, . , Nsf − 1 and where Nzc is the largest prime number less than or equal to Nsf. That
is, we simply wrap the necessary number of values of the sequence sq[n] around to the back of the
sequences to “top it off” to the desired length. To create a truncated ZC sequence of arbitrary length is also simple. Just create a ZC sequence of
length Nzc ≥ Nsf and cut off the last Nzc − Nsf values of the sequence. In practice, cyclically extended
sequences tend to be preferred to truncated sequences for reasons that are not entirely clear to me.Where the paper says it
Where the paper says it
Where the paper says it
Origin and Context
The origin of Zadoff-Chu sequences dates to the late 1950s. The ZC sequences themselves were publicly proposed by Chu in 1972 [4], for any arbitrary length. p. 8 My understanding of their origins is as
follows. The ZC sequences themselves were publicly proposed by Chu in 1972 [4], for any arbitrary
length. The ZC sequences themselves were publicly proposed by Chu in 1972 [4], for any arbitrary
length. However, ZC sequences as known today (i. However, ZC sequences as known today (i. for odd length sequences) were previously proposed
much earlier in a secretive patent application [5] filed in 1957 (and issued in 1963), as noted in [6]
but likely unbeknownst to Chu. for odd length sequences) were previously proposed
much earlier in a secretive patent application [5] filed in 1957 (and issued in 1963), as noted in [6]
but likely unbeknownst to Chu. Thus, the dual recognition of both Zadoff and Chu is appropriate, with
Zadoff listed first as in [7], where the name “Zadoff-Chu” as a descriptor for these sequences appears
for the first time (to the best of my knowledge).Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
References
[1] C. Li and W. Huang, “A constructive representation for the Fourier dual of the Zadoff-Chu sequences,” IEEE Transactions on Information Theory, vol. p. 8 REFERENCES
[1] C. Huang, “A constructive representation for the Fourier dual of the Zadoff-Chu sequences,” IEEE Transactions on
Information Theory, vol. REFERENCES
[1] C. Huang, “A constructive representation for the Fourier dual of the Zadoff-Chu sequences,” IEEE Transactions on
Information Theory, vol.Where the paper says it
Where the paper says it
[2] S. Sesia, I. Toufik, and M. p. 8 4221–4224, 2007. Toufik, and M. Toufik, and M. Baker, LTE: The UMTS Long Term Evolution, 2nd ed.Where the paper says it
Where the paper says it
[3] E. Dahlman, S. p. 8 Wiley, 2011. Dahlman, S. Dahlman, S. Parkvall, and J. Parkvall, and J. Skold, 5G NR: the next generation wireless access technology.Where the paper says it
Where the paper says it
Where the paper says it
[4] J. Chu, “Polyphase codes with good periodic correlation properties,” IEEE Trans. p. 8 Elsevier, 2018. Chu, “Polyphase codes with good periodic correlation properties,” IEEE Trans. REFERENCES
[1] C. Huang, “A constructive representation for the Fourier dual of the Zadoff-Chu sequences,” IEEE Transactions on
Information Theory, vol.Where the paper says it
Where the paper says it
[5] S. Zadoff, “Phase coded communication system,” US Patent 3099796, July 1963
[6] R. L. Frank, “Comments on polyphase codes with good correlation properties,” IEEE Trans. p. 8 Zadoff, “Phase coded communication system,” US Patent 3099796, July 1963. Frank, “Comments on polyphase codes with good correlation properties,” IEEE Trans.Where the paper says it
[7] B. M. Popovic, “Generalized chirp-like polyphase sequences with optimum correlation properties,” IEEE Trans. p. 8 Frank, “Comments on polyphase codes with good correlation properties,” IEEE Trans. Popovic, “Generalized chirp-like polyphase sequences with optimum correlation properties,” IEEE Trans.Where the paper says it
Summary
ZC sequences are complex sequences with unit amplitude and arbitrary phase, as opposed to Walsh and PN codes which are real and binary valued (for our purposes, at ±1). They have a number of remarkable and desirable properties that we define in the next section. p. 0 ZC sequences are
complex sequences with unit amplitude and particular phase shifts, as opposed to Walsh and PN codes
which are real and binary valued, most commonly ±1. ZC sequences have a number of remarkable and desirable properties that we define in the next section. ZC sequences have a number of remarkable and desirable properties that we define in the next section. Because of these properties, they are used for many key functions in current cellular systems, and are
likely to be prevalent in future cellular systems as well.Where the paper says it
Where the paper says it
How the training signal is built
The sequence sq[n] is defined by the formula sq[n] = exp −jπq n(n + 1) / Nzc (1). The sequence has a length Nzc, which must be an odd number, and is often a prime number. p. 0 A Zadoff-Chu sequence has two key parameters, (i) the root index q =
1, 2, . , Nzc − 1 and (ii) the length of the sequence, Nzc, which must be an odd number, and is often
a prime number.Where the paper says it
Where the gains come from
The desirable properties of ZC sequences stem from their structure. These properties are used for many key functions in current cellular systems, and are likely to be prevalent in future cellular systems as well. p. 0 ZC sequences have a number of remarkable and desirable properties that we define in the next section. Because of these properties, they are used for many key functions in current cellular systems, and are
likely to be prevalent in future cellular systems as well. The exact mapping for the
IDFT or DFT of a ZC sequence depends on the length Nzc of the sequence and can be found in [1]. The implementation benefit of this property is that a ZC sequence can be generated directly in
the frequency domain without actually taking the DFT of the sequence. The exact mapping for the
IDFT or DFT of a ZC sequence depends on the length Nzc of the sequence and can be found in [1]. The implementation benefit of this property is that a ZC sequence can be generated directly in
the frequency domain without actually taking the DFT of the sequence. The implementation benefit of this property is that a ZC sequence can be generated directly in
the frequency domain without actually taking the DFT of the sequence. This is particularly useful
for OFDMA or SC-FDMA waveforms that utilize the frequency domain for signaling. The implementation benefit of this property is that a ZC sequence can be generated directly in
the frequency domain without actually taking the DFT of the sequence. This is particularly useful
for OFDMA or SC-FDMA waveforms that utilize the frequency domain for signaling. The unnormalized cyclic cross correlation is√
Nzc. This is in fact the optimal cross
correlation for any two sequences with the optimal autocorrelation defined above. The unnormalized cyclic cross correlation is√
Nzc. This is in fact the optimal cross
correlation for any two sequences with the optimal autocorrelation defined above.Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
Where the paper says it
ZC sequences are utilized in LTE and 5G NR for many key functions. These include initial downlink synchronization, random access, and uplink control information. p. 0 For example, ZC sequences are used for initial access in
both the downlink (synchronization sequences) and uplink (random access premables). They are also
used for transmitting uplink control information, and as pilot symbols for both uplink channel sounding
and fine-grained channel estimation. Synchronization is not only how the UE acquires the BS timing, but also how it receives the key
system information necessary for subsequent communication. Specifically, the Primary Synchronization
Sequence (PSS) in LTE is based on a Nzc = 63 sequence, and values q = 29, 34, 25 are used (a
different value of q for each sector, usually). 8
In 5G, the PSS is no longer a ZC sequence, but rather is based on PN sequences of length 127. Random access.Where the paper says it
Where the paper says it
Where the paper says it
The sequence sq[n] = exp −jπq n(n + 1) / Nzc (1) has a length Nzc, which must be an odd number, and is often a prime number.
The desirable properties of ZC sequences stem from their structure. The implementation benefit of the DFT or IDFT of a ZC Sequence is that a ZC sequence can be generated directly in the frequency domain without actually taking the DFT of the sequence. This is particularly useful for OFDMA or SC-FDMA waveforms that utilize the frequency domain for signaling. The optimal cross correlation for any two sequences with the optimal autocorrelation defined above is √Nzc.
Green — the paper says it, and the page is printed. Red — no passage carries it, so treat that sentence as unverified. The same check runs on your own text: check a summary against its source, $5.
Improvements for AI systems
-
Improved AI systems can be utilized for initial access in cellular systems by leveraging ZC sequences for both downlink synchronization and uplink random access premables. This allows AI to synchronize with base stations more reliably across different network conditions.
-
AI-driven channel estimation can be enhanced by using ZC sequences as pilot symbols for fine-grained channel estimation, as mentioned in the paper. This enables more precise channel state information acquisition for better data transmission quality.
-
Frequency domain processing in AI systems, such as those used in OFDMA or SC-FDMA waveforms, can benefit from the DFT or IDFT property of ZC sequences being a ZC sequence itself. This allows for direct generation of sequences in the frequency domain without explicit FFT computation, improving implementation efficiency for signaling.
-
Systems can be designed to utilize cyclically extended ZC sequences to create spreading codes of arbitrary length, such as Nsf = 12 in LTE and 5G NR uplinks. This flexibility allows AI systems to adapt spreading factors to specific resource block sizes or channel requirements without relying solely on prime length sequences.
-
AI systems can implement robust random access mechanisms by utilizing ZC sequences for the Physical Random Access Channel (PRACH) preamble, specifically using Nzc = 839 and Nzc = 139 in both LTE and 5G NR. This ensures strong shifted cross correlation properties for handling multiple simultaneous users at different timing offsets.
-
AI systems can optimize uplink control information transmission by employing cyclically extended length 12 ZC sequences for formats used in LTE and 5G NR, such as formats 0 and 1. This provides a spreading code to allow multiple UEs to send control information concurrently while maintaining good correlation properties.
Abstract
Zadoff-Chu (ZC) sequences are an important manifestation of spread spectrum in modern cellular systems, including LTE and 5G NR. They have to some extent displaced PN and Walsh sequences which were the mainstays of 3G cellular (WCDMA and cdma2000) and the 2G-era IS-95. ZC sequences are complex sequences with unit amplitude and particular phase shifts, as opposed to Walsh and PN codes which are real and binary valued, most commonly plus or minus1. ZC sequences have a number of remarkable and desirable properties that we define in the next section. Because of these properties, they are used for many key functions in current cellular systems, and are likely to be prevalent in future cellular systems as well. In LTE and 5G NR, they are widely used for a number of important initial access and overhead channel functions that are often overlooked by engineers who focus on data transmission. For example, ZC sequences are used for initial access in both the downlink (synchronization sequences) and uplink (random access premables). They are also used for transmitting uplink control information, and as pilot symbols for both uplink channel sounding and fine-grained channel estimation. It is not an exaggeration to say that most types of signals other than the data transmissions in modern cellular standards utilize ZC sequences. In this primer, we define ZC sequences and introduce their key properties, and provide some examples. We also discuss modified ZC sequences that are commonly used in practice, but are not, strictly speaking, ZC sequences. We also overview their uses in LTE and 5G.
Related papers
- The Gaussian Multiuser Channel Capacity Region
- The mathematical landscape of partial information decomposition: A comprehensive review of properties and measures
- Analytically Characterized Optimal Power Control for Signal-Level-Integrated Sensing, Computing and Communication in Federated Learning
- Information Spectrum Methods for Multiple-Access Channels--Part I: epsilon-Capacity Regions in the Mixed Gaussian Case
- Seed Recycling: Explicit Codes for Mixed Synchronization Errors Matching the Leading Term of the Gilbert--Varshamov Bound
- Second-Order Image Characterization with Distortion and Secrecy