Network Optimization

Cellcom launched service in 1995 and has grown to 2.1 million customers. Its network of 1100 base stations (with 2750 sectors) and 12 switches provides service to the entire nation of Israel using the D-AMPS (IS-136) TDMA air interface technology.

Cellcom’s rapid growth and the dynamics of the Israeli wireless market–Israelis can choose from several nationwide operators and are notoriously picky about quality–have placed a premium on system engineering.

We were skeptical about being able to achieve significant quality improvement with software optimization tools. Nevertheless, over the years Cellcom has tested a number of such tools including the Falcom system from Schema, which provides optimized frequency planning and network management for narrowband technologies like TDMA and GSM.

To accurately gauge Falcom’s effectiveness, we chose to limit its trial to a small, well-defined area where service quality based upon manual frequency planning was already particularly good. We thus selected the southern town of Beer-Sheva, applying Falcom to the 75 sectors in the Cellcom network there. Prior to the trial, the average quality metrics in Beer-Sheva were impressive even by Cellcom standards, with a dropped call rate of 0.88%, downlink BER at 5%, uplink BER at 4%, and busy-hour blockage below 1%.

The Falcom system is unique in that it has the flexibility to take in a wide range of network data from which it creates an Impact Matrix–a mathematical model of network radio system behavior that characterizes the potential for mutual RF interference between any two sectors in the network.

In Beer-Sheva we used extensive signal level drive test data as well as RF propagation predictions from our existing Planetä software tool. The Falcom system readily digested these inputs as well as operational data from the switch that serves Beer-Sheva.

From the available data, Falcom generated an Impact Matrix. Then, Falcom’s optimization algorithm produced a new frequency plan for Beer-Sheva. This plan allocated the required number of channels to each sector while taking into account all constraints such as minimum channel spacing for antenna combiners, the grouping method used and the existing channel allocations in the network surrounding the 75 sectors being planned.

To test its effectiveness, Cellcom implemented the Falcom-generated frequency plan in Beer-Sheva and closely monitored network performance in the 75 target sectors. As shown in the table below, the results were very impressive.

The Beer-Sheva trial clearly demonstrated that a powerful optimization system like Falcom can deliver network performance far superior to that achieved by conventional engineering means. As a result, Cellcom is now considering deploying Falcom throughout its network, expecting gain significant improvements in quality, capacity, and network management efficiency.
Baruch Bar is Radio Group Senior Manager for Cellcom (Israel).