Dynamic Spectrum Sharing (DSS) is an innovative new technology that enables the parallel use of LTE and 5G in the same frequency band, therefore allowing operators not owning 5G frequency bands to immediately capitalize on the benefits of 5G NR by deploying it as part of NSA networks in existing LTE frequency bands.
But what does such technology mean for the world of network testing?
In this blog post, Part 6 in a series dedicated to initial 5G NR drive testing, we review some key points on DSS technology and what we could expected of its performance, as well as few aspects related to DSS drive testing.
Motivation
DSS is widely viewed as a key short-term enabler for deploying 5G networks quickly, as opposed to the slower, more expensive process of spectrum re-farming from one generation of wireless technology to the next.
Here are few main reasons for why this consideration viable:
First, is the fact that using DSS allows 5G NR to benefit from the low band spectrum (<1 GHz), which enables a greater coverage as compared to the mid/high-frequency bands.
Second, DSS functionality significantly reduces cost and time-to-market, which mobile operators will incur when launching 5G NR, since neither must they spend on new spectrum or engage in the tedious process of spectrum re-farming, nor do they need new or upgraded UE hardware. The latter is ensured by the fact that DSS uses the LTE multicast-broadcast single frequency network (MBSFN) transmission feature available with all LTE devices.
Last but not least, DSS enables a logistical advantage. New adopters of 5G NR will be significantly lower in number as compared to existing 4G-LTE users, and therefore a phased approach, in which the mix of NR to LTE users can be carefully monitored and managed, helps minimize this risk, without the huge investment required for dedicated 5G NR channels.
Techniques
DSS requires dynamic control of the amount of spectrum dedicated to LTE and/or 5G. 3GPP defines three main DSS techniques with the goal of combining them so that the system configuration and scheduling can avoid NR/LTE simultaneously transmitting (with the same time slot and same subcarrier), as well as inter-numerology interference between NR and LTE. Those techniques are:
- Rate matching patterns (NR, DL only) defined either with resource element (RE) granularity or symbol granularity, in order to cope with LTE’s several ‘always on’ signals with fixed time-frequency resource allocation. Thus, NR transmits only over unused LTE resources;
- MBSFN subframes (LTE, DL only) to be used in LTE to disable ‘always on’ signals from LTE in the cases for which rate-matching is not possible (e.g. NR with frequency subcarrier at 30kHz); and
- Time (time division multiplexing [TDM]) and frequency (frequency division multiplexing [FDM]) NR-LTE separation through scheduling and configuration granularity for DL/UL to avoid collision between LTE and 5G NR. This technique is only possible to use for UL since LTE UL doesn’t have any ‘always on’ signals or MBSFN subframe equivalent.
Challenges
Each of the DSS techniques, as well as an optimal combination of all three DSS, require complex control both at network- and device-side.
On the network side, the gNB scheduler has to support high flexibility, efficient spectrum management, and handling different numerologies in transmission and reception.
At device side, the reception of different numerologies and sharing RF chains between 5G NR and LTE needs to be supported.
All of these requirements can be accomplished through additional control/signaling overhead, which will likely result in degradation of the customer’s throughput experience. Therefore, a major challenge emerges from the need to keep the overhead impact at a reasonable level on LTE capacity and even more on 5G NR capacity.
Initial testing
3GPP only provides guidance for DSS implementation. Therefore, testing of DSS operability and impact both on LTE and 5G NR performance is crucial in making sure that optimal capacity and throughput is achieved at eNB/gNB and UE level, as well as verifying smooth eNB-gNB-UE interoperability; hence the need for device-based measurements to evaluate DSS performance impact on the user in drive test scenarios.
The DSS feature supports instant spectrum sharing with decisions scheduled every millisecond to share the LTE and 5G NR resources grid while maximizing resources utilization.
Quality measures of LTE and NR, as well as related QoS settings, are used when scheduling LTE and NR data dynamically. Therefore, initial DSS testing use cases need to be focused on the monitoring of the DSS signaling, which enables the analysis of the configuration set-up, and of the analysis of the RAN configuration and the performance of serving LTE and NR cells, with the DSS feature enabled by the system.
Analysis of the percentage of spectrum radio blocks (RBs) allocated per technology (LTE, NR) and NR/LTE RBs allocation by frequency within each subframe correlated with the measured overall and per technology performance and with the configuration represent the main inputs for DSS feature troubleshooting and optimization.
While performing these tests and analyses, a few things need to be kept in mind:
- In the case of low bandwidth spectrum (< 1GHx), the switch to 5G NR can happen if an LTE anchor at mid-bandwidth spectrum (1GHz – 6GHz) is co-located with the LTE low bandwidth spectrum to be shared with 5G NR;
- The switch to NR is expected to happen whenever NR is available and stronger than LTE, but only if test devices can support the use of concurrent low bandwidth frequencies; and
- In the case of low bandwidth spectrum, 5G NR has available a fairly low bandwidth (5- 15MHz) when compared to LTE 20MHz+ or even 40MHz+ when carrier aggregation (CA) is configured. Therefore, in this case, depending on the NR bandwidth value, it’s expected that LTE performs as well as NR or even better than it. Consequently, in this case, if the implemented DSS technique would allow too high a percentage of RBs allocation of NR, then these RBs are taken from the high or well performing LTE, resulting in an overall degraded rather than increased performance expected from the DSS feature.
To read more on our DSS testing solution and the broader topic of drive testing for 5G NR, download our white paper, Initial 5G NR Drive Testing with Infovista – and look out for the seventh and final part in this blog series. Now published, it focuses on another important use case: dual connectivity.
Make sure you have the right strategy and tools to test DSS feature, as well all the other 5G NR features when you come to rely on these.
