This article originally appeared in the October 2010 issue of INTERNET TELEPHONY
Coverage gaps exist in metropolitan, suburban and urban communities in every wireless network. Research suggests that 20 percent of churn is caused by only 6 percent of the networks cell sites. Although coverage gaps can be isolated, they continue to exist despite technical measures, standard procedures, and other failed initiatives.
Chaos theory applied to an understanding of wireless networks and human behavior can improve this scenario by identifying coverage holes with higher accuracy based on real network data.
Chaos theory is the field of studying the behavior of dynamical systems. Chaos means state of disorder. To be classified as chaos the following properties must be present: sensitivity to initial conditions, mixing of topology and density of periodic orbits.
In the wireless world, we see chaos each and every time a call is originated on a wireless network. The complex architecture tracks the call as it progresses through the network: the dialing number, dialed number, trunks used, cell sites and sectors used, duration of calls, handoffs, etc. What appears simple is, in fact, quite chaotic. All this data and much, much more are collected and stored within the systems managing the call process. Add to that the number of calls originated per day and then calculate the amount of data collected, and the volume of data is almost unimaginable.
To capitalize on subscriber-based intelligence, it becomes necessary to mine the terabytes of data collected from cell sites, switches, mobile measurement reports, operations support systems and system performance reports, developing correlations between all the collected data points.
Collected data can be processed through a series of chaos theory algorithms, which sort and compartmentalize the data based on the events that occurred during call. An example is processing data from a certain number of city blocks that experience 10,000 calls in one week. The data from each call is layered, noting the behavior of the calls, how the calls progressed, where they dropped, the routes people walked, the interaction of the network with the buildings (interference), call power levels, which network elements were involved in the call, and finally, the termination of the call.
The correlations appear during the layering of the calls. From the overlay you can see the correlations, which expose the areas of weak signals, interference locations, locations of dropped calls, cell sites related to the calls, complaints, and overall poor system performance – basically the coverage gaps in the wireless network. Output from tools utilizing this technology can be displayed geographically, with resolutions down to building levels, intersections, polygons and specific cell sites. Additional correlations can identify gap size, cell breathing, capacity, and even prioritize the gaps using annualized revenue per unit or complaints, along with a host of other parameters, based on the purpose of the report.
Studying these correlations enables service providers to manage wireless network optimization cost effectively, using subscriber-based intelligence. Analyzing subscriber call data provides service providers with valuable information on usage patterns, demographics, call activities, locations, time of use, data usage, high complaint areas, good coverage, poor data throughput, and much more. This data allows service providers to not only target resolutions for gaps, but also to align their marketing strategies according to customers’ calling patterns.
Once subscriber-based intelligence is analyzed, operators can improve their customers’ experience by surgically targeting problem areas. Advanced technologies such as distributed antenna systems, picocells, and femtocells can provide up to 100mbs uploads, less than 5 millisecond latency, LTE (News - Alert) guideline conformance, further enhancing the customers’ quality of experience.
Applying this technology, coverage gaps can be detected with up to 95 percent accuracy. The business value of this capability is clear: When a service provider in the Middle East tested subscriber-based intelligence derived from using the chaos theory in one region, churn was reduced by .27 percent, equating to a $5.6 million annual savings.
TMCnet publishes expert commentary on various telecommunications, IT, call center, CRM and other technology-related topics. Are you an expert in one of these fields, and interested in having your perspective published on a site that gets several million unique visitors each month? Get in touch.
Edited by Jaclyn Allard
