Case Study - KPI Optimization with RF Shaping & Layer Management

Introduction: Spectral efficiency is key to supporting a large subscriber base on limited available spectrum. To maximize the use of spectrum, a mix of footprint optimization or RF shaping is required in conjunction with parameter tuning and feature optimization. Cluster/Area optimization of a LTE and 3G network with macro and oDAS layers requires a complete cell footprint view in order to effectively optimize and maintain performance in a mature wireless network. The entire area and complimentary cell interactions are required rather than individual worst offender optimization typically undertaken today. To improve poor performing sites, oDAS nodes and areas, surrounding cell footprints need to be optimized to provide the best capacity and network quality.

In order to efficiently optimize, a complete and mature process needs to be executed consisting of a mix of detailed drive data, user based recordings and KPIs. It is essential to analyze and optimize oDAS and macro layer by defining proper RF footprints and HO boundaries between the layers. The OptPCS-iCE Optimization process recommended by Telecom Technology Services, Inc (TTS Inc) in conjunction with IMNOS involves this combination of detailed drive data, switch recordings, and cell performance KPIs. This is not a simple look exercise, but instead a baseline and improvement process to ensure a quality VoLTE/ LTE/ UMTS network and determine any additional long term planning required for long term success in these areas. Based on the full input of configuration, drive, and KPI data, OptPCS-iCE optimization process will provide this insight to the cluster performance improvement and long term efficiency.

Background: A US Tier 1 operator with limited spectrum and a data driven consumer base recently expanded their LTE network with additional capacity as well as rolled out VoLTE. The network had been optimized considering the existing macro network upgrades along with the oDAS network in place to provide infill coverage and capacity offloading. Even though the area was performing well, Telecom Technology Services, Inc (TTS Inc) was brought in to see if the network could be further improved. Deep dive optimization was performed on a complex cluster to determine what strategies and techniques could be applied to the market in general upon cluster completion.

Study: TTS Inc is an innovator in optimization software, strategies, and implementation for both short and long term improvement of wireless networks. The deep dive study of the network cluster was to determine opportunities in the network for market wide implementation and development of a quality optimization process for market engineering teams to utilize after. Specific focus was made into the footprint of macro and oDAS nodes to determine the proper interaction and overlap. With the fundamentals addressed in regards to proper RF coverage, layer managements and parameter optimization could make the impact desired.

Detailed scan data of up to the top 20 cells of each WCDMA and LTE carrier was collected. This scan data was utilized to determine over-serving, under-serving, misaligned, crossed, and existing serving footprints. It also determined areas of excessive overlap and poor coverage. Consistency checks were run on the configurations and parameter/feature settings to ensure a proper baseline. Parameter Settings were closely looked at through the IMNOS platform. This allowed for distribution and outlier identification. LTE PCI and RSI plan conflict audits were available as well as any plan changes required through IMNOS. The main focus in parameters was limited inter technology handovers and re-selection as well as balancing traffic between layers of the same technology. The balanced traffic approach allowed for the best spreading of users and reduced interference on the previous highly utilized carriers.

Table 1: General Cluster Information

TTS Inc brought in two pieces of software to work in conjunction with existing operator tools. IMNOS and OptPCS-iCE utilized the existing network configuration and parameter setting matched with the detailed scan drive to allow for parameter, configuration, and footprint optimization.

• IMNOS (Integrated Mobile Network Optimization Software): An enterprise web based software focusing on network engineering analytic and effective automated RAN network solutions
• OPTPCS-iCE (Intelligent Cell Engineering) Utilize measured data to reflect simulated changes based on current RF conditions improving cell overlaps, neighbor interaction, and weak RF coverage areas through individual bin measured pathloss.

Throughout the initial data collection phase of the project, observations based on the detailed data collection and network KPIs, the correct settings and adjustments for this RF environment type were determined through engineering experiences and the tools utilized by TTS Inc. Further validation with operator tools were done to validate consistency and work with the operator for an engineering long term process. The purpose of the study was not only to determine areas of network improvement, but to leave behind a process for operator engineering teams to work with.

Figure 1: IMNOS - Coverage Optimizer for Tilt Review

Findings: The network, although operating in an acceptable KPI range, had opportunities in both the macro and oDAS optimization. Initial audits and consistency checks were conducted to resolve basic configuration and datafill errors. These errors were resolved upfront in order to provide a baseline for actual optimization improvements. The configuration audits against detailed scan data showed had various swapped sectors due to planning database mismatches as well as parameter and neighbor errors due to SON implimentation. These items were also resolved prior to implementing any recommendations.

RF Shaping on LTE and 3G network found opportunities to reduce interference caused by overshooting cells, mismatches between carriers, and coverage fixes with tilt and azimuth adjustments. As this was a high traffic area with capacity constraints, the traffic of each cell as well as shared antenna considerations needed to be accounted for before any recommendations were made. 37 physical cell configuration changes were implemented to improve the RF conditions in the cluster. These changes were all identified and modeled through the OptPCS-iCE software to have a complete understanding of the positive and any potential negative impacts of the changes on a cell and composite area view.

Layer Management Corrections involved neighbor adds/deletes and handover parameter adjustments. This included adjusting the SON settings to further account for unusual scenarios where the Inter-Frequency carrier distribution was uneven and also made manual changes where necessary through SON control modifications. Layer Management correction helped to efficiently balance traffic and reduce handover failures and improved accessibility on poor performing cells.

Figure 2: 3G Layer Management Strategy

The oDAS presented opportunities to improve small coverage holes and offload capacity. The detailed analysis found that by adding 4 nodes, the small coverage holes in the cluster could quickly be filled. The network operator needed to work with the 3rd party oDAS company to introduce these node locations. It was also found that some nodes were residing directly under existing macro sectors. Detailed scan data showed these nodes to not be the best server as well as these areas having an elevated noise floor. With no chance of overcoming the macro network power and thus only introducing noise to the network without capacity offload, simply turning off these nodes improved the network quality in the area as well as the overall oDAS performance.

Table 2: Implemented Changes

Upon implementation of the recommended changes, the network subscriber experience KPIs followed with the expected improvement in 3G as well as LTE. The highlights of the improvements are seen here.

• LTE Performance Improvement
  • DL Throughput improved by 4% with over 20% data volume increase
  • Data leakage reduced by 12%, voice leakage by 3%
  • Cluster voice drop call rate improved by 27%
• UMTS Performance Improvement
  • Voice Drops reduced by 600 on daily average basis
  • SHO Success Rate improved by over 33%
  • Voice access failure rate improved by 18%

Table 3: Area Performance Improvements

The large take away was that data and voice performance improved while significantly increasing the network loading.

Conclusion: Network optimization work requires a mix of consistency checks and audits to ensure proper baselines, physical configuration modifications or RF Shaping, and feature/parameter tuning to meet the RAN network performance improvement targets. Each RF environment is different and requires variable approaches to maximize results. As networks get more complex, having the tools to visualize and simulate changes accurately will become even more critical. As LTE networks are utilized to carry the bulk of the network loading, greater emphasis needs to be placed on the SCINR and RSRP levels than early networks that were focused mainly on throughput speeds when served by LTE.

Focused overall area optimization based on fundamental RF spectral efficiency practices tend to result in improved KPIs in addition to a positive customer experience for long term network growth and sustainability. A balanced approach with the correct tools and experience to these items allows for even greater KPI improvements to be made. This improvement is possible even in areas that perform well and have gone through initial optimization by operators and OEMs.