• Jordi J. Giménez

FeMBMS, the origin for LTE-based 5G Terrestrial Broadcast

Updated: Jun 30, 2019

FeMBMS stands for Further enhanced Multimedia Broadcast Multicast Service, which as the name indicates is an evolution of eMBMS. FeMBMS was standardized in LTE Release 14 and is considered as the first approach to Terrestrial Broadcast in 3GPP, at least the first solutions that meets several of the requirements defined for broadcasting.

What is the origin of FeMBMS?

As all standardization activity, FeMBMS is the result of contributions from many companies involved in 3GPP (Third Generation Partnership Project).

The possibility to operate eMBMS in a similar fashion as traditional broadcast systems (e.g. like DVB-T/T2, ATSC3.0 or DAB+) came as a result of a study item in Release 14 to develop functionalities to meet requirements from broadcasters. Big players such as Qualcomm, Huawei, Nokia or Ericsson played an important role in standardizing this evolution of eMBMS. The European Broadcasting Union (EBU) together with the research branches of public service broadcasters in UK (with the BBC) and Germany (with the IRT) started engaging the activity as the possibitily of deliver TV and radio services to smartphones and tablets in the "conventional" way broadcasters are used to do that (with a broadcast downlink only and receive only mode, free-to-air, with a high level of control, etc) is very attactive.

Work in 3GPP towards Release 16 will end up in a new evolution of FeMBMS which will be finally known as LTE-based 5G Terrestrial Broadcast.

MBMS was initially standardized in 3GPP for the first phase of UMTS standards in release 6, this means 2004. Later on, eMBMS was standardized as part of LTE release 9. Concepts like introducing SFNs (Single Frequency Networks) or creating a service layer to ingest audiovisual content come from that time. In fact, the initial LTE version of MBMS, comes with the so-called Multicast-Broadcast Single-Frequency Network (MBSFN) bearer supporting broadcast only services based on OFDM, like other well-known mobile broadcast standards (at that time DVB-H, -SH or -NGH).

Qualcomm's patent US 9,325,552 B2 already brings a wide number of features to convert eMBMS in a Terrestrial broadcast system similar to standards like DVB, ATSC or DAB.

However if we would need to set the company that more closely addressed the Features at the physical layer for supporting Terrestrial Broadcast operation, this is Qualcomm. In fact, the Patent US 9,325,552 B2 by G.K. Walker et alt. is an excellent approach to what a Terrestrial Broadcast system based on 3GPP technology would need in terms of addressing a variety of use cases even using different network topologies and infrastructure.

Work in 3GPP towards Release 16 will end up in a new evolution of FeMBMS which will be finally known as LTE-based 5G Terrestrial Broadcast. Some of the solutions introduced in this patent will indeed be useful for such system although it is unclear to what degree these will become a reality.

Some interesting extracts of the patent are shown here:

"Specific broadcast use cases in various terrestrial broadcast systems, such as the Advanced Television Systems Committee (ATSC), may be served by a derivative form of LTE Broadcast/eMBMS. Depending on the specific conditions, these use cases may benefit from a cyclic prefix that is longer than those nominally available in the existing LTE specification, such as regular and extended cyclic prefixes.

  • The first use case considered provides a low power, low tower height, mobile, single frequency network (SFN) net work. This use case defines a typical LTE Broadcast application, which may support tablets and smartphones, whether indoors, outdoors, or vehicular. Low power typically refers to a network that has transmitter sites in the range of 2 kW effective isotropic radiated power (EIRP) per 5 MHz. Low tower typically refers to a radiation height in the range of 30 m, and mobile typically refers to a network type that Supports all classes of service for which the receiving antenna is not stationary. The applicable range of Doppler Velocity is generally 3 km/hr to 200 km/hr for ATSC. The currently defined 16.66 and 33.33 us cyclic prefixes should be sufficient in both mixed and dedicated carrier modes. This selection of cyclic prefixes may be duplicated in any standalone mode. This deployment style may also be suitable for indoor reception by nominally fixed receivers. The high Doppler rate is generally not required for this use case, however the low speed Doppler may be beneficial. The appropriate channel model for such reception is a multipath Rayleigh fading model. Indoor recep tion is likely dominated by close-in reflections in a temporal sense. Given that this use case is dominated by a Rayleigh fading dominated channel models, there may be significant efficiency gain possible due to the use of MIMO. The potential benefits of MIMO here depend on the deployment style of the network. This deployment style is Single Frequency Network (SFN). Typically, the frequency reuse in such deployments is 100% and the bits per second (bps)/Hz of the deployment is less than or equal to two bps/Hz, although the selection of modulation coding scheme (MCS) may ultimately be determined by the site density and the total number of sites within a multicast broadcast multimedia service SFN (MBSFN) area. There are interference regions about the transition from one MBSFN to another that must be considered in the net work design.

  • Another use case considered provides a medium power, high tower, mobile, SFN. Medium power typically refers to transmitter with a maximum radiated power of 50 kW effective radiated power (ERP), and high tower typically refers to a transmit radiation height above 200 m. This deployment style is a potential ATSC/LTE Broadcast application that may Support tablets and Smartphones, whether indoors, outdoors or in a vehicle. The applicable range of Doppler velocity may generally range from 3 km/hr to 200 km/hr. This deployment style is also potentially suitable for indoor reception by nomi nally fixed receivers. As with the low power, low tower height, mobile SFN, a high Doppler rate is generally not required for this use case, however the low speed Doppler may be benefi cial. The appropriate channel model for Such reception may also be a multipath Rayleigh fading model. As currently defined, the LTE Broadcast physical layer has limitations for application to this use case, such that the existing cyclic prefixes may not belong enough to adequately guard against the long differential delays. The presence of high transmit towers in the network may lead to Such long differential delays, in which cyclic prefixes of more than a 90 us may be beneficial.

  • Another use case considered provides a high power, high tower, multi-frequency network (MFN), with rooftop reception. Roof top reception generally refers to the receiving antenna being stationary and at a receiving height in the range of between 9 and 10 m. This type of deployment style is also a potential ATSC/LTE Broadcast application that supports roof top reception for consumption with nominally fixed 5 receivers. The currently defined channel model is an additive white Gaussian noise (AWGN)-based model, although it is known that a Ricean model may also be useful for defining the appropriate channel in this use case. The duration of the channel for this style of deployment is typically less than 30 10 us for the 99" percentile reception locations. However, there are known cases of paths up to 100 us. Because of these longer paths, there may be a need for cyclic prefix durations greater than 100 us, in order to support this use case. This dimension of cyclic prefix may also support medium power SFN within 15 the high power footprint, e.g., in geographically shadowed areas.

The ATSC target efficiency is currently defined at 4.2 bps/ HZ at 15 dB carrier-to-noise (C/N) for AWGN channels. In order to achieve this level of capacity, the pilot overhead 20 should be decreased, as compared to the mobile profile. This combination of highbps per HZefficiency with relatively long cyclic prefix will also result in a larger fast Fourier transform (FFT) requirement. It should be noted that the lack of time diversity in this use 25 case may be treated by the use of Cyclic Delay Diversity (CDD) or other related methods, e.g., Space-Frequency Block Code (SFBC), although this is most effective for Ray leigh channels, which may be more likely to occur for indoor reception, which may also need to use more pilot energy. 30 Thus, there may be a conflict between a maximum rooftop reception efficiency and an indoor reception. Another use case considered provides a low power, lower tower, SFN, rooftop reception. This use case may not be nominally required within the context of ATSC, however, it 35 may fall within the range of cyclic prefixes that might otherwise Support other use cases. This use case is generally based on city/suburban coverage for indoor and vehicular handheld reception with a moderately dense deployment and rooftop reception in a far more sparse rural deployment. Simulations have demonstrated that cyclic prefix durations up to 200 us 55 may be beneficial for the rural reception use case. As an SFN deployment style, wherein spectral reuse may approach 100%, the exact spectral reuse depends on the use of directional receive antennas with a sufficient front to back ratio in the rural border areas of the respective multicast- 60 broadcast single frequency networks (MBSFNs). Interior areas of the SFN may utilize omni-directional antennas. An MBSFN transition within suburban areas may use similar considerations of interference as in the previous low power, low tower use case."