Providing capacity to keep pace with the explosive growth in wireless data services is a highly complex task due to the time variant and geographically inhomogeneous nature of offered traffic. To ensure a consistent, high quality user experience, both site and carrier density must be tailored to match the offered traffic density, with sufficient margin to accommodate large fluctuations in traffic loading.
A white paper from ISCO International discusses the statistical distributions of traffic and interference from a network perspective, and how poor correlation between measured traffic and uplink channel power can be used to assess interference severity. Interference is a pervasive problem, and as networks strain to handle growth in wireless data traffic, the importance of identifying and resolving performance
degradation due to interference is growing more acute.
Traditional approaches address severely impaired sites on a case-by-case basis, and are often identified by customer complaints of poor service quality. Applying a more network-centric approach, by jointly analyzing RTWP and traffic loading distributions to identify potential trouble sites, represents a more proactive approach to interference management that allows operators to identify RF impairments before they turn into customer complaints.
Figure 1 Total percentage of network traffic carried vs. site percentile.
Figure 1 illustrates the distribution of traffic, ranked vs. site percentile from highest to lowest in five percent bins, for a typical urban 3G wireless network. If site coverage perfectly matched the geographical distribution of offered traffic, each bin would carry the same fraction of the total network traffic and the histogram would appear uniform. The actual distribution, however, is highly skewed: Sites in the first bin (the busiest five percent of sites) carry 19 percent of the total network traffic, and the busiest 25 percent carry half of the total traffic.
Read the complete white paper online at www.mwjournal.com/ISCOtraffic.
