More to quality than pure capacity

11 Apr 2007

Improvements in network quality can be delivered without boosting capex by closely measuring network traffic to detect bottlenecks early.

Like its parent company, Shenzhen Mobile, a branch of Guangdong Mobile within the China Mobile group, has experienced dizzying growth since the late '90s. But as its subscriber base rose to 8.5 million, it found customers were experiencing slow response times when making calls and sending text messages, with a call taking an average of 7 seconds to set up.

The operator, similar to others experiencing rapid subscriber growth, believed its best investment and the fix to slow response times was rolling out new network equipment, which is where it focused its spending. With increased capex Shenzhen Mobile assumed the congestion in the network would be relieved and as a result the quality of service and customer experience would improve.

'It's the usual assumption that if you make the highway a little bit bigger, then it will improve the traffic flow,' said Naveen Bhat, Asia Pacific VP and GM of Agilent Operations Support System.

Finger pointing

But after continued network investments, the company realized there was no correlation between increased capex and improvements in network quality. To address the problem its first step was to approach its two main equipment suppliers and asked that them to fix the quality problem.

Each of the vendors assured Shenzhen Mobile its network equipment was working to spec and that the problem lie somewhere else.

After this finger-pointing exercise, the operator called in Agilent to conduct an independent audit of its equipment manufacturers' network components. It was tasked with diagnosing the cause of the delays and prescribing the fix. It was logical to turn to Agilent, which had been monitoring the operator's network since 2001 and had installed probes that track traffic at various points across its network.

While observing the data is a simple exercise, Bhat noted that after systems get deployed people usually don't pay that much attention to the information, and over time the knowledge is lost and the data becomes meaningless because people don't have the expertise to make sense of it. 'That's why they called us back, and said 'okay you have all this data, can you relate it to a specific problem'.'

As a pure professional services project, Bhat pointed out that the work didn't require any additional measurements in the network or the sale of a product. It took a look at all the information that was coming off the network and found that while Shenzhen Mobile had installed Access 7 signal monitoring systems, it was not fully utilizing the data that was coming out of them.

In analyzing the flow of traffic across the MSCs, the base stations and the transactions across the A interface, it identified that the problem was not with the wireless gear but with the STP configurations, which were the main source of the call delays. Agilent also looked at various parameters that were being reported across the SMS interfaces and was able to get a significant improvement in SMS delivery. Shenzhen Mobile was able to make the fixes in a matter of days.

Related content

No Comments Yet! Be the first to share what you think!