10 5G NR Physical Layer Interview Questions

10 5G NR physical layer interview questions with answers for freshers: numerology, 273 PRBs, LDPC vs polar, SSB, BWP, DMRS, HARQ and peak data rate.

By BitForBytes Editorial & Hardware Research Team, Official BitForBytes Hardware Publication · 2026-10-09 · 11 min read

⚡ Quick Summary

5G NR physical layer interviews for freshers test whether you can reason from a few core facts: subcarrier spacing is 15 × 2^μ kHz, a slot is 14 symbols, a resource block is 12 subcarriers, data uses LDPC and control uses polar codes. This article answers ten common questions with the numbers and the one-line reasoning interviewers expect.

1. Introduction & Core Motivation

A fresher interview for a modem, RAN software or wireless-test role is rarely a tour of all of 3GPP. It is a short list of physical-layer questions, each with a follow-up designed to separate a candidate who memorised a table from one who can rebuild it. "What is numerology?" is followed by "so why is the subframe still 1 ms?". "How many PRBs in 100 MHz?" is followed by "why not 277?".

Most question lists online are either architecture-level, covering network slicing and core functions, or paid question banks. The few free physical-layer lists give definitions without numbers, which is precisely what the follow-up question attacks.

Each answer below has three parts: the direct answer, the number or formula behind it, and the follow-up you should expect. The questions track the order in which a physical layer is usually taught: waveform, resource grid, coding, synchronisation, reference signals, retransmission and throughput. After reading, you will be able to answer all ten aloud in under a minute each.

2. Detailed Technical Breakdown

1. What Is Numerology in 5G NR, and Which Subcarrier Spacings Exist?

Numerology is the subcarrier spacing and cyclic-prefix configuration of an NR carrier, identified by an index μ. The spacing is

Δ f = 15 × 2^(μ) kHz

where Δ f is the subcarrier spacing and μ is the numerology index. Data channels use 15, 30 and 60 kHz in FR1 and 60 and 120 kHz in FR2; 240 kHz is used only for the SS/PBCH block. Worked example: μ = 3 gives 15 × 8 = 120 kHz.

Follow-up: "Why more than one spacing?" Higher carrier frequencies have more phase noise and Doppler, which need wider spacing; large low-band cells have long echoes, which need the longer cyclic prefix that comes with narrow spacing.

2. Why Does the Slot Get Shorter While the Subframe Stays 1 ms?

The slot gets shorter because it is defined as 14 OFDM symbols, and each symbol lasts 1/Δ f; the subframe is defined as a fixed 1 ms regardless of numerology. A subframe therefore holds 2^(μ) slots.

Follow-up: "What is the benefit of a shorter slot?" Scheduling is per slot, so a shorter slot lowers latency.

3. How Many Resource Blocks Are in 100 MHz at 30 kHz, and Why Not 277?

A 100 MHz carrier at 30 kHz has 273 resource blocks. One resource block is 12 subcarriers, so 360 kHz wide, and $100000/360$ gives 277, but 3GPP TS 38.101-1 reserves a guard band at each edge.

The occupied bandwidth is 273 × 360 = 98.28 MHz, leaving (100000 - 98280)/2 - 15 = 845 kHz of guard band per side. Spectral utilisation is 98.3%, against 90% for a 20 MHz LTE carrier.

Follow-up: "How many resource elements in one RB in one slot?" 12 × 14 = 168.

4. Why Does NR Use LDPC for Data and Polar Codes for Control?

NR uses LDPC for data because data blocks are long and need high decoder throughput and HARQ, and polar codes for control because control messages are short and must decode reliably in one attempt. LDPC decoding is parallel and handles code blocks up to 8,448 bits. Polar codes with CRC-aided list decoding perform better at payloads of tens of bits.

Follow-up: "Does PDCCH use HARQ?" No. A missed DCI is a missed grant; there is no soft combining for control.

5. What Is Inside an SS/PBCH Block, and How Many Cell IDs Exist?

An SS/PBCH block contains the primary synchronisation signal (PSS), the secondary synchronisation signal (SSS) and the physical broadcast channel (PBCH) with its DMRS. It spans 4 OFDM symbols and 240 subcarriers, which is 20 resource blocks.

 subcarriers
   239 +--------+--------+--------+--------+
       |        |        |  PBCH  |        |
       |        |  PBCH  +--------+  PBCH  |
   182 |        |        |        |        |
       +--------+        |        |        |
       |  PSS   |        |  SSS   |        |
       | (127)  |        | (127)  |        |
    56 +--------+        |        |        |
       |        |        +--------+        |
       |        |  PBCH  |  PBCH  |  PBCH  |
     0 +--------+--------+--------+--------+
        symbol 0 symbol 1 symbol 2 symbol 3

The physical cell identity is

N_ID^(cell) = 3 N_ID^((1)) + N_ID^((2))

where N_ID^((1)), from 0 to 335, is carried by the SSS and N_ID^((2)), from 0 to 2, by the PSS. That gives 336 × 3 = 1008 identities, twice LTE's 504. Worked example: N_ID^((1)) = 100 and N_ID^((2)) = 2 give a cell ID of 302.

Follow-up: "Why does the UE detect PSS first?" There are only three PSS sequences, so the search is cheap; it gives symbol timing and N_ID^((2)), after which the SSS search is narrowed.

6. What Is a Bandwidth Part, and Why Does NR Need It?

A bandwidth part (BWP) is a contiguous set of resource blocks, with one numerology, inside the carrier, on which a device transmits and receives. A device can be configured with up to four BWPs per direction, and one is active at a time.

NR needs BWPs for three reasons: a device may support less bandwidth than the 100 MHz or 400 MHz carrier; monitoring a narrow BWP when traffic is light saves power; and different BWPs can use different numerologies.

Follow-up: "PRB versus CRB?" Common resource blocks are numbered from Point A across the carrier; physical resource blocks are numbered from 0 inside a BWP.

7. When Does the Uplink Use DFT-s-OFDM Instead of CP-OFDM?

The uplink uses DFT-s-OFDM when the device is coverage-limited, because its lower peak-to-average power ratio lets the power amplifier run closer to saturation. CP-OFDM is used otherwise and always in the downlink.

Follow-up: "What does LTE use in the uplink?" Only DFT-s-OFDM (SC-FDMA). NR adds CP-OFDM so that uplink MIMO and frequency-selective scheduling work the same way as in the downlink.

8. Which Reference Signals Does NR Have, and Why Is There No CRS?

NR has four main reference signals: DMRS, PT-RS, CSI-RS and SRS. NR has no always-on cell-specific reference signal (CRS) because a permanent signal across the whole band wastes energy, causes inter-cell interference and does not fit beamformed transmission.

Follow-up: "What does 'front-loaded DMRS' mean?" The DMRS is placed at the start of the allocation so that the receiver can estimate the channel before the data arrives, which reduces decoding latency.

9. How Does HARQ in NR Differ From LTE?

HARQ in NR is asynchronous in both downlink and uplink, with timing signalled in the DCI, whereas LTE uplink HARQ is synchronous with fixed timing. NR supports up to 16 HARQ processes per cell, against 8 in LTE FDD.

NR also allows code-block-group retransmission: a large transport block is split into groups, and only the groups that failed are resent. Retransmissions use incremental redundancy from the LDPC circular buffer, selected by redundancy version.

Follow-up: "Why asynchronous?" Flexible timing is needed for dynamic TDD and for devices with different processing speeds.

10. How Do You Estimate the Peak Data Rate of a 100 MHz Carrier?

You estimate it with the 3GPP TS 38.306 formula: layers times bits per symbol times code rate times resource elements per second, less overhead.

R = v · Q_m · R_max · (12 · N_RB) / T_s^(μ) · (1 - OH)

where v is the number of layers, Q_m the bits per modulation symbol, R_max = 948/1024 the maximum code rate, N_RB the number of resource blocks, T_s^(μ) = 10⁻³/(14 × 2^(μ)) s the average symbol duration and OH the overhead (0.14 for FR1 downlink). Worked example: v = 4, 256-QAM so Q_m = 8, N_RB = 273, μ = 1. Resource elements per second are 12 × 273 × 28000 = 91.73 × 10⁶. Then 4 × 8 × 0.9258 × 91.73 × 10⁶ × 0.86 ≈ 2.34 Gbit/s.

Follow-up: "Is that what a TDD network delivers?" No. If, for illustration, 74% of symbols are downlink, the ceiling is about 1.73 Gbit/s.

Which Numbers Should You Have Ready Before the Interview?

The numbers to have ready are the handful that the ten answers above keep reusing.

3. Practical Takeaways for Students

• Exam angle: Questions 1, 2, 3 and 5 are standard short-answer and numerical material in wireless-communication papers. Write the formula, substitute, and box the result: 120 kHz, 80 slots per frame, 273 RBs, PCI 302.
• Interview angle: Answer in three beats: the fact, the number, the reason. "273. One RB is 360 kHz, which would give 277, but guard bands of 845 kHz per side are reserved." Then stop and let the interviewer choose the follow-up.
• Common mistake: Reciting LTE facts as NR facts: 504 cell IDs, 8 HARQ processes, turbo codes, always-on CRS, 90% spectral utilisation. Each has changed, and interviewers use exactly these to check whether you studied NR or reused LTE notes.
• Try it yourself: Recompute question 10 for 2 layers and 64-QAM (Q_m = 6) and check that you get about 876 Mbit/s. Revise OFDM and channel-coding basics on the BitForBytes Roadmaps.

4. Frequently Asked Questions

Which 3GPP specifications should a fresher read for physical layer interviews?

Start with four documents in the TS 38.2xx series: TS 38.211 for physical channels, modulation and numerology; TS 38.212 for multiplexing and channel coding; TS 38.213 for control procedures; and TS 38.214 for data procedures. TS 38.300 gives the overall architecture in plainer language.

Do not read them cover to cover. Use them to confirm specific facts, such as a table value, after you understand the concept from a textbook or lecture notes.

What is the difference between FR1 and FR2 in 5G?

FR1 is the lower frequency range, from 410 MHz to 7.125 GHz, and FR2 is the millimetre-wave range, from 24.25 GHz to 52.6 GHz. FR1 carriers are up to 100 MHz wide and use 15, 30 or 60 kHz spacing; FR2 carriers are up to 400 MHz wide and use 60 or 120 kHz.

FR2 offers far more bandwidth but shorter range and higher blockage loss, so it relies on beamforming with large antenna arrays.

How should I prepare for a 5G NR interview with no telecom project?

Build one small artefact you can explain line by line. A NumPy script that generates OFDM symbols for two numerologies and shows the effect of cyclic-prefix length, or a calculator that reproduces resource-block counts and peak data rates, is enough to anchor a conversation.

Then practise the ten questions above aloud. Interviewers for fresher roles weigh clear reasoning from fundamentals such as Fourier transforms, sampling, modulation and coding more heavily than specification trivia.

5. References & Further Reading

• 3GPP TS 38.211, "NR; Physical channels and modulation" (numerologies, resource grid, SS/PBCH block, reference signals).
• 3GPP TS 38.212, "NR; Multiplexing and channel coding".
• 3GPP TS 38.213, "NR; Physical layer procedures for control" (cell search, bandwidth part operation).
• 3GPP TS 38.214, "NR; Physical layer procedures for data".
• 3GPP TS 38.101-1, "NR; User Equipment (UE) radio transmission and reception; Part 1" (transmission bandwidth configuration).
• 3GPP TS 38.306, "NR; User Equipment (UE) radio access capabilities" (supported maximum data rate).
• 3GPP TS 38.321, "NR; Medium Access Control (MAC) protocol specification" (HARQ operation).
• E. Dahlman, S. Parkvall and J. Sköld, "5G NR: The Next Generation Wireless Access Technology", Academic Press.