CSL: Setting the Standard for 4G Network Performance

Release Date:2014-01-24 By Li Jingtao and Liu Jianming Click:

CSL is the leading mobile communications operator in Hong Kong, and operates three networks: GSM, WCDMA and FDD LTE. Rapid increase in LTE subscribers has imposed higher requirements on its network performance. In March 2013, CSL launched a network optimization project to boost overall performance of its LTE network. Comparative test results by an independent network testing company showed that the overall network performance increased significantly after optimization, ranking towards the top among all mobile operators in Hong Kong. This monthly comparative test, lasting nearly seven months, used the same test equipment and procedures to collect network performance data of all Hong Kong’s mobile operators via drive tests, call quality tests and MTR tests. Test results showed a substantial improvement in the network KPIs of CSL’s LTE network during the seven months.

CSL’s LTE network started commercial operation in September 2011. Network performance kept on a high level in the beginning when the customer base is small. However, in the later half year of 2012, CSL started to witness a rapid growth of its LTE subscribers. In the highly competitive telecommunications market of Hong Kong, network performance is vital of customer experience.

In March 2013, CSL chose ZTE, amongst other competitors, to undertake the overall optimization project of its LTE network, aiming to improve network performance for metro lines and outdoor sites. Upstream/downstream rates and cell switch performance were in urgent need of improvement for metro lines, while outdoor sites faced more complex challenges: improving benchmark test rates, enhancing load balance between different frequencies and systems, shortening the delay for opening a website, and increasing network performance in areas of weak coverage.

 

Network Optimization for Metro Lines

The cells that cover Hong Kong’s metro lines are simple: 10M bandwidth cells that operate at the 1.8 GHz frequency. The ticket hall, platform and railways are all covered by the same cell. All operators lease the same POI system of MTR Corporation Limited, therefore the wireless coverage of all telecom operators is very close, since signals are transmitted and received through the same single-input and single-output (SISO) antenna.

To achieve best optimization results, a typical metro line was tested, analyzed and optimized, and the optimization plan was then applied to all metro lines. Fine-tuning of physical channel parameters (PUCCH/PHICH parameters) and mobility parameters (same frequency handoff parameters, inter-system handoff reselection parameters, and inter-frequency measurement parameters) helped to significantly increase network speeds. Tests showed that after optimization, the uplink speed increased by 175%, and the downlink speed increased by 50%.

 

Performance Optimization of Outdoor Sites

The optimization involved the following aspects to improve performance of outdoor sites:

●    uplink/downlink speed. Handoff parameters, PA and RS transmission power parameters, and RLC transmission mode parameters were optimized.

●    interoperability strategy. The optimization aimed at ensuring load balance at the 1.8 GHz frequency and 2.6 GHz frequency bands and increasing the chances of a mobile terminal residing in the LTE network. To achieve this goal, parameters of inter-frequency (2.6 GHz and 1.8 GHz) and inter-system (LTE and 3G) reselection, handoff, redirection measurement and event triggering were optimized. Both idle cell reselection parameters and active cell handoff parameters could be optimized.

●    delay for opening a website. The delays for opening a website through CSL’s network and its rivals’ networks were compared and evaluated. Drawing on experience in data analysis, an analytical method was developed for the delay analysis. This helped CSL locate the problems from benchmark test data and eventually solve serious delay issues through software upgrade.

●    random PRB allocation mechanism. The initial position of a PRB was allocated at random. This helped to reduce interference between neighbor cells and improve network performance, especially the performance in areas with weak coverage.

 

After the optimization, the performance of outdoor sites was significantly improved. Statistics from May to October 2013 showed that the delay for opening a Yahoo website was reduced by 34% and 33% for a Apple Daily website, while FTP upload and download speeds increased by 120% and 106% respectively.

 

Improving Network Performance for Better Subscriber Experience

The seven months of efforts paid off: the project was a huge success, and CSL ranked towards the top in LTE performance in Hong Kong, as evidenced by benchmark test results. CSL had a big growth in LTE active users and 120% growth in LTE traffic in the period of project. The improved network performance also significantly improved user experience.

The success of this project is a historic event in LTE performance optimization for both CSL and ZTE. The solution implemented in this project proved effective, and it will lay a solid foundation for further optimizing CSL’s LTE network in the future.