3GPP Channel Model Types for 4G and 5G

Understanding 3GPP Channel Models in LTE and NR


Introduction

  • 3GPP defines various channel models for LTE and NR to simulate real-world propagation conditions and evaluate system performance. These models are specified in:

    • TS 36.101 – LTE User Equipment Radio Transmission and Reception

    • TS 38.141 – NR Base Station Conformance Testing

    • TS 38.901 – Channel Model for 5G

     

  • The common channel models include:

    • AWGN (Additive White Gaussian Noise)

    • EPA (Extended Pedestrian A)

    • EVA (Extended Vehicular A)

    • ETU (Extended Typical Urban)

    • MBSFN (Multimedia Broadcast Single Frequency Network)

    • TDL Models (TDL-A, TDL-B, TDL-C, TDL-D, TDL-E)

  • These models help evaluate device performance under different multipath fading and mobility scenarios.

  • They can be configured to mimic diverse propagation scenarios allowing testing under different propagation conditions, simulating real-world scenarios.

  • For LTE channel models, they are defined in TS 36.101. They include models like EPA, EVA, ETU and MBSFN are commonly used. 
  • For NR, they are defined in TS 38.141, TS 38.901. They are the TDL models (A, B, C, D, E) and are used for advanced multipath and MIMO testing. 
  • AWGN serves as a universal baseline. 
  • TDL models simulate how signals bounce in the real world. 3GPP [TR 38.901] defines different TDL models differentiated with how many taps each model has and the delay and power with each tap (echo). These taps affect how signal fades and spreads out over time. By different configuring different number of taps, we can simulate all kinds of scenarios from NLOS to LOS scenarios.
  • TDL-A, B and C have more taps to represent scenarios where the signal reflects all sorts of surfaces like in urban areas with plenty buildings. They are used for NLOS.
  • TDL -D and E have fewer taps with more extended multipath delays such as hills or rural areas. Imagine standing on a hill directly facing a tower. They are LOS models, and follow Rician fading distribution [TR 38.901]. 


Overview of 3GPP Channel Models

  1. Additive White Gaussian Noise (AWGN)

- A basic channel model with a constant spectral density for static performance measurement. No fading or multi-paths exist for this propagation model (TS 25.141). 
        - It represents an ideal ad noise-limited environment without fading or multi-paths like the lab. 
- It is ideal for testing baseline performance of devices in a clean, controlled environment. 
- It can be used to evaluate throughput performance without the complexity of fading.
- It can be experimentally used for High-speed scenario if the correct Doppler shift is defined e.g., TS 36.101 section B3 and  TS 36.141 Annex B.

2. Extended Pedestrian A (EPA) 


- This is a low delay spread fading model simulating pedestrian mobility in urban and suburban areas. 
- It has a doppler shift of 5Hz. 
- Defined in TS 36.101, TS 36.116, TS 36.104
- Best used to simulate slow-moving devices e.g., pedestrians in environments with minimal multipath such as open urban streets or parks. 
- Ideal for handover and data reliability at low speeds.
- It has 7 multipath components. Each tap can be modeled to have a Rayleigh or Rician distribution with a mean value described in the table below.


3. Extended Vehicular A (EVA)

- This is a medium spread model with a relatively higher Doppler shift e.g., 70 Hz, representing vehicular mobility in urban settings. 
- Defined in TS 36.101, TS 36.116, TS 36.104.
- It is best to simulate vehicle-mounted devices moving at moderate speeds (30 - 50 KM/H) in urban areas with some multipath reflections (e.g. from buildings). 
- It has 9 multipath components as shown in the table below. 


4. Extended Typical Urban (ETU)

- High delay spread model with significant doppler effect (300 Hz), mimicking complex urban environments with fast moving devices. 
- Defined in TS 36.101, TS 36.116, TS 36.104.
- It is ideal for simulating high-speed urban scenarios (e.g., vehicles with 120+ km/h) with extensive multipath from dense buildings or infrastructure. 
- It has 9 multipath components.

All the taps in the tables above have a classical Doppler spectrum. 
In addition to a multipath delay profile, each multipath fading propagation condition has a maximum Doppler frequency as shown below:


5. Multimedia Broadcast Single Frequency Network (MBSFN)

- This is a channel model designed for broadcast services over LTE with longer delay spreads and specific multipath profiles. 
- Best for testing broadcast or multicast services e.g., eMBMS in LTE or 5G scenario. 
- Simulate scenarios like stadiums, urban squares, or rural areas where multiple cells transmit identical signals assessing synchronization and coverage for video streaming or emergency alerts. 

6. Tapped Delay Line A, 30 ns delay spread (TDLA30)

- Defined in TS 38.901 with a low/short delay spread of 30ns, representing line-of-sight (LOS) or near LOS conditions with minimal multipath. 
- Suitable for simulating indoor or small-cell environments e.g., offices, homes or outdoor rural areas with clear LOS. 
- Use it for short -range indoor Wi-Fi like conditions. 

7. Tapped Delay Line B, 100ns delay spread (TDLB100)

- Defined in TS 38.901, a TDL model with moderate/nominal delay spread, 100ns, typical for usage in suburban areas or urban areas with non-line-of-sight (NLOS) conditions. 
- Best for suburban residentials areas or campus with moderate multipath from scattered buildings or trees. 
- Suitable for pedestrians or low-speed vehicular traffic. 

8. Tapped Delay Line C, 300ns delay spread (TDLC300)

- It has high/long delay spread, 300ns, desired for urban NLOS environments with significant multipath. 
- Best for simulating dense urban settings, e.g., City centers with reflections from tall buildings
- Suitable for NLOS conditions at low to moderate speeds ( city speeds, up to 50KM/H). 
- Good for capacity testing under rich-multipath conditions. 

9. Tapped Delay Line A

- 

10. Tapped Delay Line B

- Suitable for NLOS conditions with moderate delay spread 
- Urban or industrial areas with NLOS propagation. 

11. Tapped Delay Line C

- Suited for NLOS conditions  with larger delay spread, representing multipaths, TDLC300 is a variant. 
- Ideal for dense urban areas and metropolitan areas with significant NLOS multipath. 
- Suitable for checking robustness and interference management, and high capacity urban networks. 

12. Tapped Delay Line D

- Model from TS 38.901, suitable for LOS conditions. 

13. Tapped Delay Line E

- A TDL with extended delay characteristics, often used for scenarios requiring longer multipath delays such as rural or hilly terrain.
- Suitable for wide area coverage scenario focusing on long-range connectivity and delay tolerance. 

  • By selecting the appropriate channel types, we can simulate and replicate specific environments - urban, suburban, rural, indoor, and vehicular - while ensuring realistic performance evaluation.
  • TS 36.104 also provides for the High speed Train scenario for open space and tunnel for multi-antennas in non-fading propagation channel one tap. For high speed train condition for a leaky cable is a fading propagation channel with one tap. The Rician factor, K is the ratio between the dominant signal power and the variant of the other weaker signals. 


Where D_s/2 is the initial distance of the train from the BS, D_min is the BS-Railway track distance both in meters, v is the velocity of the train in m/s and f_d is the maximum doppler frequency. 

  • TS 36.141, provides for LTE receiver moving propagation conditions. The ETU200, with propagation conditions derived from Band 1, with 200 Hz Doppler frequency is used. Also, an AWGN-only profile using a single path with a Doppler frequency of 622.653 Hz is used. AWGN does not account for the Doppler shift. 



  • TS 38.141 in section G.2.1 provides the steps for the simplified delay profiles from the TR 38.901. Of interest, are the delay profiles for FR1, section G.2.1.1, that gives the number of channel taps with maximum delay spread for TDLA30, TDLB100 and TDLC300. Further, in section G.2.2 of TS 38.141, we have the combination of channel model parameters, which is channel model name and a maximum Doppler frequency. E.g., TDLA<DS>-<Doppler> where <DS> is the desired delay spread, and <Doppler> is the maximum Doppler frequency (Hz).
  • The tables below show the propagation conditions that are used for the performance measurements in multipath fading environments for low, medium and high Doppler frequencies for FR1 and FR2 respectively.

Table for Channel model parameters for FR1

Combination Name

Model

Maximum Doppler Frequency

TDLA30-5

TDLA30

5 Hz

TDLA30-10

TDLA30

10 Hz

TDLB100-400

TDLB100

400 Hz

TDLC300-100

TDLC300

100 Hz

TDLC300-600

TDLC300

600 Hz

TDLC300-1200

TDLC300

1200 Hz

Table for Channel model parameters for FR2

Combination Name

Model

Maximum Doppler Frequency

TDLA30-35

TDLA30

35 Hz

TDLA30-75

TDLA30

75 Hz

TDLA30-300

TDLA30

300 Hz

TDLC60-300

TDLC60

300 Hz

TDLD30-75

TDLD30

75 Hz


  • TR 38.901, provides a sample way of scaling parameters for TDL models


Conclusion

By selecting the appropriate 3GPP channel models, we can replicate specific environments—urban, suburban, rural, indoor, and vehicular—ensuring realistic performance evaluation of LTE and NR systems. These models are crucial for network optimization, device testing, and system validation in various real-world conditions.

References: 

3. 3GPP TR 38.900 version 14.2.0 (2016-12), 3rd Generation Partnership Project; Technical
Specification Group Radio Access Network; Study on channel model for frequency spectrum
above 6 GHz (Release 14) (TR 38 900)


Location: 65760 Eschborn, Germany

Comments

Post a Comment