Bell Labs, the research and development arm of Lucent Technologies, has been working with Telecom New Zealand over the last two years to trial a new approach to optimizing the coverage and capacity of mobile networks.
This technology, called dynamic optimization, makes adjustments to Telecom New Zealand's mobile network throughout the day in response to changing traffic patterns.
Dynamic optimization is being tested on Telecom New Zealand's commercial 027 network (CDMA) in the Lower Hutt, Taranaki, Bay of Plenty and Mt. Eden regions. Lucent originally built the 027 mobile network and currently manages it for Telecom New Zealand.
"Dynamic optimization gives us an entirely new way of managing network traffic," says Mike Flanagan, Bell Labs technical manager, who is overseeing the testing. "Traditionally, communications networks are designed around a static scenario based on peak usage patterns. But of course traffic patterns are always changing. In the evenings the majority of mobile device use is at the home, during the day it is in the office area, and during rush hours along the major traffic arteries. By creating a network that can optimize itself dynamically to adjust to the changing network needs throughout the day we can enable service providers to improve quality of service related to the delivery of mobile voice and data services."
Dynamic optimization technology leverages Bell Labs' Ocelot software. Bell Labs has combined Ocelot with real-time technologies that measure usage patterns and performance conditions across an entire network at any given time of day. Armed with this information, Ocelot is used to find the best network parameters to achieve optimal capacity and coverage. These network parameters include base station transmit power levels as well as programmable antenna orientation.
The company says currently network parameters are effectively static - they are set up once and then re-tuned on a monthly or yearly basis at best. This current project involves tracking usage patterns throughout the week (weekdays and weekends) and then programming power levels and antennas to regularly modify their settings throughout the day and week to adjust to these changing usage patterns.
During the trial, Bell Labs is working in cooperation with KMW Communications Inc. to use their programmable base station antennas that provide adjustable pan/tilt functionality. By changing the pan, tilt, and power output of each antenna, the system can remotely modify the network coverage topology and shift some of the traffic from congested base stations to underutilized ones by dynamically modifying each base station's coverage area. "We can 'borrow' unused capacity from areas where it isn't currently needed and direct it to other areas where demand is higher at any given time," Flanagan explains. "The result is better service and improved network efficiency."
The next phase of this research is to implement real-time dynamic optimization - where the network will continually monitor network traffic usage and be able to intuitively optimize the network in short time intervals. "Pattern modeling is very useful and a major step, but users don't always conform to those patterns," said Flanagan. "For example, when there is a specific event like a traffic jam or a big rugby game there is a big spike in mobile phone calls in a concentrated area. Traditional network design can't cope with those changes in network usage very well."
Real-time dynamic optimization will also provide subscribers with more reliable service. For example, if a base station were to be damaged in an event like a natural disaster, in a real-time dynamically optimized network, the network would automatically adjust to extend coverage to the affected area and fill the gap.