Standalone LTE in unlicensed spectrum: design principles
A standalone LTE system, operating entirely in unlicensed spectrum, would broaden the reach of LTE performance benefits in unlicensed bands. This in turn would open the LTE ecosystem to non-traditional network providers and enable new RAN-sharing capabilities for telecom operators. In this blog post, we provide an overview of the challenges, design principles, and new opportunities of standalone LTE in unlicensed spectrum.
Data traffic in mobile networks continues to outpace the spectrum resources available for mobile wireless systems. Here in the Ericsson Research blog, we have previously described how licensed assisted access (LAA) enables mobile device users to be served by a combination of licensed and unlicensed spectrum bands at the same time. You find those posts here – a short summary can be found at the end of this post.
However, LAA is fundamentally tied to the requirement of licensed carriers, which act as anchors for mobility and control signaling. An alternative would be a standalone LTE system, operating entirely in unlicensed spectrum. Recognizing the potential benefits of such a system, the MuLTEfire Alliance has initiated a technical study of the necessary design requirements.
Radio access network design challenges
A standalone system must satisfy the following requirements to be viable:
- It must adhere to any applicable regulatory requirements for unlicensed band operation, such as limits on transmit power spectral density and channel bandwidth occupancy.
- It must achieve fair coexistence with other nodes of either the same or a different technology, e.g., Wi-Fi.
- It must deliver robust performance for critical functionalities such as initial access and mobility.
- It must deliver downlink and uplink throughputs comparable to or better than other unlicensed technologies.
The first two aspects can be realized by adopting the physical channel designs and listen-before-talk (LBT) channel access mechanisms of 3GPP Rel-13 LAA and Rel-14 eLAA. In particular, LAA channel access mechanisms on both DL and UL have been designed to ensure that LAA is a better neighbor to Wi-Fi than another Wi-Fi network. This has been verified through extensive simulations and lab tests, which we described in this post.
Unlike licensed carriers, the timing of DL system information (SI) transmission and UL transmission opportunities for initial access or hybrid ARQ (HARQ) feedback is generally uncertain due to the need for LBT. SI transmission can be made more robust by expanding the time-frequency resources assigned for the Physical Broadcast Channel (which carries the Master Information Block (MIB)), and including MIB/SIB transmissions in subframes carrying discovery reference signals (DRSs). UL transmission opportunities for control information can be increased by allowing DL HARQ feedback and random access preambles in both regular UL subframes and in the end of subframes with partial transmissions, as shown in Figure 1.
Supporting mobility between standalone networks is also challenging due to uncertainties in the timing of reference signals for inter-cell measurement and the possibility that networks are unsynchronized. For synchronized and loosely synchronized standalone networks, the DRS-based radio resource management (RRM) measurement principles defined for LAA can be used. When networks are not synchronized, the UE will have to spend more time searching for neighbor cell DRSs, which impacts power consumption and mobility performance. In this case, it is beneficial for standalone eNodeBs to configure separate DRS measurement instances for the serving cell and adjacent frequencies, and to allow the inter-frequency measurement configurations to be more dynamic in order to increase the likelihood of encountering neighbor cell DRSs. Mobility from a standalone LTE in unlicensed spectrum to a normal LTE network on licensed spectrum is expected to be seamless, while mobility in the opposite direction is also feasible.
Neutral host capabilities
From a network architecture perspective, standalone should be designed with the flexibility of using either a traditional PLMN evolved packet core (EPC), or directly using an IP network for connectivity. The latter case gives rise to a so-called neutral host network (NHN) mode, where multiple operators can share a single NHN ID across standalone cells, without having to deploy separate radio access networks. UEs are consequently given increased flexibility in how they connect to the standalone network: either with a PLMN subscription or with a subscription to a service provider affiliated with the NHN. An example of an NHN is shown in Figure 2.
Standalone opens the LTE eco system to new access providers, such as building owners, systems integrators and enterprises
LAA has established a solid foundation for bringing LTE to unlicensed bands. Standalone will open the door to further innovation and novel use cases as we tackle the associated design challenges and increase the choice of technologies that may be used in unlicensed spectrum.
Ericsson Research, Silicon Valley.