5G radio access technologies focus on evolving the existing LTE architecture and new candidate technologies for the new RAT. The new evolved LTE architecture supports the control and user plane separation. Furthermore, flexible duplex schemes, such as unlicensed bands, Frequency Division Duplex (FDD) for eLTE, and Time Division Duplex for new 5G RAT will provide the extra adaptability and scalability needed to offer the latest 5G services. Moreover, the intention is to use Non-orthogonal Multiple Access (NOMA) on LTE to provide high cellular capacity and massive connectivity. Finally, the support of low-cost, long-life terminals and devices will cater for enhancing LTE towards IoT and M2M applications.
5G network techniques
For the new 5G RAT, techniques such as massive MIMO beamforming will help expand the cell radius and improve frequency usage efficiency. In addition, new radio parameters such as larger available bandwidths will offer higher capacities, while shorter TTI length improves latency figures. With a new signal waveform for the localization of interference, this new radio frame configuration also improves efficiency by reducing inter-cell interference and providing energy savings and high extendibility.
These new Multiple Access schemes provide spectrum efficiency gains in both the Downlink and the Uplink direction. In addition, non-orthogonal Multiple Access offers improved spectrum efficiency and control channel capacity limitation to provide connectivity to up to one million devices per square kilometer. Furthermore, reduced latency by grant-free transmission and simplified procedures can provide for advanced low-latency applications like e-Healthcare, autonomous driving etc.
Frame format
The new RAT design offers radio scaling since the subcarrier spacing scales with the frequency band. More specifically, higher frequency bands will have larger subcarrier spacing and shorter symbol duration.
The radio frame has better efficiency than 4G technology since, in LTE, the signals are transmitted regardless of traffic. However, this constant transmission creates a source of interference and a waste of energy. The 5G frame format sends the minimum number of radio signals to allow mobility measurements. In addition, demodulation signals are multiplexed into the data transmission for UEs. This way, the system maximizes spectral efficiency across users and base stations. Link budget and capacity are also maximized since the new format considers the target use case requirements. Finally, the new format minimizes the protocol overhead to improve scalability, reduce power consumption, and increase capacity, having an overall lower control overhead.
Multiple Access schemes
Various Multiple Access schemes can be used according to the needs of the network architecture applied. For example, in LTE Uplink, SC-FDMA can be used which offers PAPR coverage and multiplexing with OFDMA. On the other hand, in LTE Downlink, OFDMA can provide higher spectral efficiency and MIMO, and no intra-cell interference. On the downside, however, both above solutions need synchronous multiplexing, while there is Link budget loss, especially when it needs to serve simultaneous users.
Some of these problems are solved by single-carrier RSMA, which allows synchronous multiplexing, grant-less Tx with minimal signaling overhead, and Link budget gain. However, this solution is not suitable for higher spectral efficiency. OFDM-based RSMA can also be used for use cases that require grant-less Tx with minimal signaling overhead but needs synchronous multiplexing. Finally, LDS-CDMA/SCMA allows lower complexity iterative message-passing multiuser detection (when there are few users). However, on the downside, it has a higher PAPR than SC-RSMA. Furthermore, it also needs synchronous multiplexing and lacks scalability and flexibility. At last, although similar to LDS-CDMA with SIC, MUSA has a higher peak-to-average power ratio (PAPR).
Distributed vs Centralized architecture
Current Distributed RAN architecture is not suitable for context-based service delivery. A Cloud RAN is expected to separate control and user plane. This way, flexible scaling of capacity for different functions of the RAN will become available. Operators can use CRAN for many network scenarios, including a wide range of transport network solutions, base station configurations, and user applications. Thanks to its unified network architecture, Cloud RAN can be aligned with legacy deployments to reduce the overall network complexity.
There are various scenarios on the use of Cloud RAN on the transport network. For example, an attractive solution in cases of a large installed base of available fiber is to use dedicated fiber. Even when the fiber is available, the network planner must use it with caution since the cost of deploying new fiber limits its overall applicability.
Optical Transport Network (OTN) uses FEC and can increase the reach of metro optical networks. However, utilizing OTN for CPRI transport requires careful consideration as some of the highly-valuable features of OTN also add latency.
On the other hand, Passive Optical Network (PON) is a suitable solution in high-traffic areas, where small-cell deployment is the most likely candidate. However, PON is vulnerable to extra latency and power loss, since optical splitters are used for signal distribution and collection. This, in turn, reduces the cell radius and makes fault isolation difficult. Microwave technology is another solution, especially for short distances that need to have a clear line of sight.
CPRI over Ethernet (CoE) is another solution that could result in cost savings. In CoE, CPRI data on the link is not sent continuously, but as discrete Ethernet 802.3 frames. However, in order to meet latency and jitter requirements, CoE sometimes requires dedicated Ethernet links between endpoints.
Wavelength-based systems (WDM) offer another solution for CPRI transport. Coarse WDM (CWDM) supports low propagation delays and high data throughput and is an economical choice, both in equipment costs and in its use of fiber resources.