The Small Cell Forum identified SON vendor interoperability to be a common concern with operators and holding back wider LTE Small Cell take-up. Network densification will result in large numbers of Small Cells being deployed ad-hoc and unplanned, making this an critical feature. Substantial progress has been made this year culminating in public operator validation and approval after extensive testing.
SON is an essential feature
In the early days of small cells, operators have been concerned that devices from new entrants could interfere and negatively impact existing networks. Perhaps that was less of a concern in remote and underserved areas where femtocells were being installed in places that had little or no coverage. But when providing capacity rather than coverage, such as in more dense urban areas, inter-cell interference was a concern. Un-coordinated cells could result in very poor service at the cell edge borders.
Unlike 3G, LTE was designed from the start with a small cell layer in mind and incorporates many features to handle that efficiently. I’ve observed that few operators were keen on using 3G for urban/outdoor and have expected LTE to be the preferred choice to increase data capacity. We are starting to see a layer of small cells become more widely introduced in major cities, especially where operators have less spectrum to play with.
Many operators are reluctant to introduce new RAN vendors into their network because there is an overhead to test and approve the equipment, develop/adapt installation and network management procedures, train staff and expand choice. This has been overcome where equipment is significantly cheaper or more suited, such as for residential and enterprise applications.
China Mobile is perhaps the most enthusiastic, having introduced ten new small cell vendors for TD-LTE. It’s not unreasonable to expect other operators to introduce and approve several new vendors in the near future, especially for in-building. They will need to be confident about interoperability both with their existing and other new vendors first.
Industry Forums already working on the issue
Both the NGMN and the Small Cell Forum have been working on this for some time. The NGMN published their “Best Practices SON” guidelines and the Small Cell Forum published “Multi-vendor X.2 Interoperability” report, both in 2014. NEC presented a summary of progress at our Cambridge Wireless SIG event that same year.
SON interoperability test documents have been produced by the NGMN and the Small Cell Forum, listing the sequence of operations and and expected results. This extended some existing SON test cases and covered a broad range of capabilities:
- - Basic self-configuration (e.g. X2 setup, Automatic Neighbour Relations, PCI selection)
- - Mobility optimisation (e.g. Mobility Robustness Optimsiation, Frequent Handover Mitigation)
- - Interference Mitigation (e.g. Inter-Cell Interference Co-ordination)
ETSI Plugfests have been held since 2010, with a growing focus on LTE SON since June 2014. The latest in June 2016 continued to focus on SON and X.2 interoperability, with over 50 test sessions active concurrently by participants. An important criteria for success of these events is that members do not release or share test results that identify vendors publicly, which encourages active and honest participation. This makes it more difficult for outsiders to assess the maturity and measure the progress made.
This year, the Small Cell Forum published a summary report listing the pass/fail rate for each test section. It's very encouraging to note that 94.5% of tests passed overall. Digging into the detail suggests that further progress needs to be made on inter-vendor interoperability, especially between macrocell and small cell vendors. Nonetheless, over 80% of those test cases passed suggesting a good level of conformance. This figure is up from 53% since the 2014 Plugfest.
BT confirms Plugfest findings match their own lab test results
As with many major network operators, UK operator BT conducts its own tests in its own lab. Unusually, they’ve published a report on their findings in conjunction with two small cell vendors and Qualcomm. This validates the private testing and confirms that both successfully interwork over the X.2 interface.
The report concludes that separately developed systems can gracefully co-exist in the same LTE network. ICIC standardisation by 3GPP, using the X.2 interface, has been a key to this success.
Areas marked for improvement relate to the differences in the provisioning systems and KPIs reported by each vendor. BT suggest this could be an area for further industry-wide standardisation – it’s certainly something that has been long overdue in the RAN industry to date.
The next steps for BT would be field trials and deployment, involving integration with the existing network management systems, core network and existing RAN. Further parameter tuning and algorithm adjustments can then be made, with performance optimisation as the primary focus.
This is encouraging and shows both progress and intent as we move towards wider take-up of LTE small cells from a more diverse range of equipment suppliers.