Cover-acity*

Carriers need improved ways to manage network coverage and capacity.

As they confront today's marketplace and economy, today's mobile wireless carriers, much like Charles Dickens, are facing the best of times and the worst of times.

Although the much-hyped next-generation mobile Internet does indeed offer great promise and reward, the risk is high. For one thing, the price for entry is steep. Whether they are investing hefty sums in 3G licenses, taking the evolutionary route to update their networks for new revenue-generating high-speed data services, or both, wireless carriers need to consider areas critical to their success in 3G.

To be leaders in the 3G world, carriers need to get to market quickly with new revenue-generating services, and they need to get better at capacity planning and data management. There is no doubt that a fully optimized network can help carriers add valuable services, gain and retain customers, reduce churn and, of course, minimize expense and satisfy financial markets and investors.

The current practice for adjusting wireless network coverage is labor intensive, expensive and time consuming. Wireless carriers know all too well about the growing pains they experience when they try to take on additional traffic. For years they have been clamoring for a simple, cost-effective way to forecast trouble spots in their networks — well before they occur.

But if managing network coverage and capacity is a difficult process now, one can only imagine what will it be like when carriers' networks experience the increased demand the mobile Internet's high-speed data services will create.

The question is: How can carriers quickly and cost-effectively optimize their networks in start-up and growth phases and easily expand their coverage and capacity to meet traffic demand today and well into the future?

Beyond Drive Testing

Obviously, the industry had to get beyond the current process for optimizing networks, which goes something like this: Radio engineers with testing equipment drive around coverage areas to determine potential problem spots. In many cases, the testing takes place during non-peak hours so it obviously does not accurately reflect usage patterns.

Based on test results, other engineers make adjustments to base-station antennas and other cell parameters to improve coverage. But in the process, fixing one problem can affect the performance of the entire network. This often results in technicians having to make multiple adjustments and tweaks to resolve a single coverage or capacity issue. The process can take weeks.

Given that timing is critical in today's competitive marketplace, a solution to the wireless-network-optimization challenge would have to eliminate the laborious, time-consuming drive tests. At the same time, it should be dynamic, inexpensive and supportive of all wireless standards. If successful, it should alleviate some 3G jitters: A fully optimized network adjustable in real-time from a computer in a central location would ease carriers' time-to-market concerns, lower their costs to enter new markets and simplify migration to the next generation of wireless networks.

Such a software-based optimization tool should direct carriers, in how to configure the cells to obtain the best coverage and capacity. The software actually should create a region-specific mathematical model of a wireless network using a variety of data such as road maps, terrain parameters and base-station locations. Then, using a global objective function that describes overall network performance, the software would perform a true, mathematical optimization.

The procedure ideally would use a selected set of cell parameters, for example, antenna orientations, as the variational parameters. Configuring the cells according to the optimized solution ensures the most efficient allocation of base-station resources for maximum network performance. For the carrier, this translates to fewer dropped calls, higher numbers of voice and data users, and higher average data throughput.

Using the appropriate objective function, it's possible to predict how each antenna in the network should be pointed, tilted and powered for maximum network coverage, maximum network capacity or any desired trade-off between the two. The analysis works for voice, for data or for a mix of the services.

The ideal software-based tool should perform a delicate balancing act and visually depict the results for carriers. Network coverage, for example, would be defined as the fraction of traffic-weighted area where the particular service can be delivered with sufficient link quality.

The coverage measure would capture the influences of thermal noise as well as interference from other cells and users. When optimizing for coverage, the tool's built-in mathematical algorithm should seek to minimize mutual interference. In the presence of multiple services, the tool would derive an overall or composite coverage measure from the individual service-specific coverage values.

Network capacity, on the other hand, would be defined by the amount of traffic the network can handle at a user-specified networkwide quality-of-service (QoS) criterion. The QoS would be selected by the carrier. Cells would be sized so that the spatial distribution of network capacity closely matches the geographic traffic distribution. The result: The load due to traffic hot spots is automatically distributed over multiple sectors or cells to minimize overloading. In the presence of multiple services, the software would find an overall or composite capacity measure consistent with the user-specified mix of service classes and their respective QoS criteria.

Coverage Vs. Capacity

Because concurrent optimization of network coverage and network capacity may lead to conflicting results, the software should seek all of the best potential compromises between both objectives and propose all of the optimum network configurations. The two axes of this plot assign network capacity and network coverage. Each blue point in this plot represents an optimum network configuration for a given balance between capacity and coverage. The green point represents the performance of the initial configuration prior to optimization. With the availability of the full set of performance trade-offs, an individual balance between network capacity and network coverage can be chosen with respect to specific customer demands. Due to the generality of this approach, the software would have a wide range of applications from initially lightly loaded greenfield designs to highly loaded networks.

Once an operating point is chosen, a GUI would allow the user to investigate and understand the performance improvements achieved in the optimization process. Red areas indicate weak coverage due to interference or weak signal strength. The plots show the degree to which optimization improved the coverage of this market. (For this analysis, a road map was used as the basis for the mesh.)

With its ability to examine the trade-off between coverage and capacity and to make rapid adjustments through built-in mathematical calculations, the ideal software-based optimization tool would allow carriers to optimize and re-optimize their networks to handle the increased traffic upcoming 3G wireless applications and services will drive.

*Editor's note: Yeah, we know, it's not a word. However, coverage and capacity are so often used in the same carrier breath, Wireless Review decided to coin a new word. (At least it's not an acronym.)

---

Polakos is Lucent Technologies director of advanced wireless technologies.