How can you obtain 5G NTN connectivity for your current GEO system?

• A channel bandwidth of 200kHz is required for both the service and feeder link – although if the satellite has regenerative capabilities the spectrum requirement for the feeder link can be diminished.

 

• Note: that potentially multiple legacy system channels can be used to form the 200kHz channel if they are contiguous and filter notches do not have a detrimental effect on the performance.

 

• Check TR36.763 Sec 6.2.2 (case 1, 4 & 7) for indicative link budget results for GEO satellites. Indicative performance of an NTN NB-IoT system was included in the slides.  The satellite should be able to deliver an appropriate spot EIRP.

 

• Note: that power budget can be saved on legacy GEO satelittes by utilizing quasi-fixed earth cell topology. Here the cells are turned ON and OFF in scheduled manner, allowing UEs to turn of AS functions between overage perods AND allowing the GEO satellite to save power

You may verify the requirements listed in the above answer are fulfilled or reach out to us to have a discussion, a preliminary assessment or startup a feasibility study specific to your constellation.

Legacy GEO satellites are in general built for transparent architectures, which is the architecture supported by the 3GPP Rel17 cellular specifications.

A PHY throughput of ~250 kbps is the limit for DL and around 22 kbps is the limit for UL. This is the per transport block throughput and not application level end-to-end throughput. Any E2E measure is significantly lower due to the large propagation involved in the GEO scenarios – 800 ms to 13 s depending on the procedural messages exchanged.

Any feature that limits the number of messages exchanged in the AS and NAS to transmit data will optimize the latency and E2E throughput. Furthermore, due to the large satellite coverage it is advantageous to have any RA and paging optimizations. Some features are listed here:

• R14: Non-anchor RACH & Paging
• R15: Early Data Transmission
• R15: Wake Up Signal
• R15: SR HARQ disabling
• R16: Group WUS
• R17: 16QAM

The maritime use-case is especially interesting for cellular NTN: NTN coverage may be used to support TN infrastructure, but it is far more relevant where there is no TN infrastructure available ie. on the sea for cellular connectivity for logistic tracking or broadband access for sailors. Maritime will likely be an attractive use case, but according to both NSR and GSMA, use cases like agriculture, logistics and energy would also be interesting.

The smallest spectral bandwidth that is required for NB-IoT operation is 2x200 kHz to allocate 1 DL anchor carrier and 1 UL carrier. NB-IoT is designed to fit within the LTE numerology and single carriers are thus only equivalent to the 200kHz, or a single LTE-PRB.

This depends on the link-budget ie. the satellite payload configuration (antenna, front-end noise figure) and the UE antenna gain. However, it can be expected to be in the high-end of cellular device capabilities i.e. 20 dBm to 23 dBm.

Phased arrays would be useful in the service link for creating the earth-fixed beams in the earth-fixed cell scenario. I would expect pahsed arrays to be used in a NGSO scenario – like the fast moving LEO scenario where electrically steering beams fast to maintain a fixture on the ground would be advantageous. In a GEO scenario, I would expect a tendency towards large aperture sizes, which could also mean mechanically steered antennas. In any case, a large antenna array will also be advantageous when it comes to the UL link-budget.

Check TR36.763 Sec 6.2.2 (case 1, 4 & 7) for an indication. For a UE NF of 7 dB:

• Spot EIRP   -->    DL CNR
• 81.6 dBm   -->    -3.3 dB
• 76.1 dBm   -->    -8.5 dB
• 84.4 dBm   -->    -2.2 dB

Yes, an increased link budget allows for transmission of more information per second since less redundancy in terms of coding and modulation selection is required to safeguard the demodulation of the transferred modulation from noise in the receiver.

It depends on the context of 5G support. If we talk about high-speed broadband, then we will probably have to wait some years. Using GEO satellites for 5G backhaul – being NR backhaul – will likely require new GEO satellites to be launched for the service. It is a scalable endeavor where just a few (3) GEO satellites could provide global coverage for a NTN broadband service and more could be added as the traffic load increases with adoption. For new GEO satellites, it would be wise to plan ahead for regenerative use-cases and potentially reconfigurability of service upon the end of the GEO satellite lifetime, e.g. by utilizing SDRs and FPGAs onboard. This ensure long-term rentability of the satellites after launch, even when their first targeted service case becomes deprecated. Also, it enables a switch of service IF the market for the preliminary service does not mature as expected after launch of the satellite. In the case of legacy GEO satellites, they may be useful for providing narrow-band NB-IoT cellular access to IoT devices on a global scale. Here, the same scalability applies, but in addition the satellites are already in-orbit and may be running deprecated services at little financial gain to their owners. According to GSMA Intelligence, there is an untapped D2D addressable revenue potential of $3obn, or 3% of existing telco revenue base – just on D2D. By narrowband use cases we mean IoT, messaging, Push-to-talk etc.

The 5G core is located after the eNB stack, so that would be in the ground-segment in the transparent architecture and in the regenerative architecture the 5G core can be placed either partially onboard the satellite along with the eNB or in the ground segment. Placing the eNB in the satellite alone reduces the propagation delay for AS messaging exchanges by half and rids you of the overhead of transmitting a copy of the RAN onto the feeder-link. (since the RAN is designed for radio access of cellular devices, and the feeder link involves massive satellite dishes in the ground segment, the copying of RAN onto feeder link is a very inefficient use of the feeder link). Placing some core network features onboard the satellite along with the eNB – such as MME and S-GW could allow for example the attachment and authorization of some UEs without the involvement of the ground-segment core network… this would reduce the overall delay for NAS procedures.

Beamforming is used in order to create separate beams on the ground to deploy geo-fixed cells and to achieve a high link-budget within beams. On the UE side little beamforming is expected – the expectation is near-isotropic antennas in the case of IoT devices for NB-IoT