Today's wireless networks hardly resemble those of only a year or two ago - of course, the applications delivered across them and the devices users use to access them are equally different. This new generation of data-intensive applications has created a surge in network usage, which has, naturally, placed increased strain on wireless networks due to an increased number of device connections. We also know that the number of bandwidth-intensive applications is only going to grow, as network operators and application developers learn how to collaborate more effectively. This will only increase the number of active sessions between devices and networks in any given time period.
Simply put, in order to send and receive data - to access applications, that is - devices must open active connections to wireless networks. In their "normal" state, when they are accessing on-board data or in not using mobile applications, devices typically do not maintain active connections; it's predominantly a question of battery life, which is among the key development issues with mobile devices, so they are designed to conserve power whenever possible.
So, between active sessions - the idle time after which a network shuts down a connection is determined by the operator, but ranges anywhere from 5 to 30 seconds, typically. But, because every active RF connection consumes network resources, whether spectrum, power, or processing resources, these active connections must be carefully monitored and managed by operators to avoid interference or congestion and, ultimately degradation of service.
While the signaling process - the activation of a connection between the device and network - with data applications is much like that of voice communications, the difference lies in the patterns with which devices connect to data applications and services. It's a simple scenario for voice calls - the connection is set up when the call is placed, remains active for the duration of the call, and is terminated when the call is ended.
Data transmissions, however, can act is several different ways, adding significant complexity to resource management and utilization. File transfers behave much the same way voice calls do, with continuous transmission until the transfer is complete. But, other applications, such as security or location-based services, can access the network at regular intervals, requiring set-up and tear-down of the connection each time. In addition, subscribers accessing Web pages or manually retrieving sports scores, for instance, create an inherently unpredictable pattern of access. Then there's also the equally unpredictable scenario where certain external events, like push email, or scores in football games that trigger text messages being sent to countless users at once every time a team scores so that users can follow not only their favorite teams, but their fantasy sports players as well. Both present equally unpredictable connectivity patterns.
The point is that, while some connections are predictable, most today are not, and the variations in user and application behavior can have a significant impact on radio access resources.
Depending on the specific applications and their automatic or manual refresh sequences, and the nework dormancy timer (the maximum time before a connection is automatically deactivated after non-usage), and, of course, download speeds, a recent Alcatel-Lucent white paper describers how different variations on a single application can result in significant differences in transmission loads and the number of signaling events in a one-hour actual RF airtime span.
This, of course, is in addition to the fundamental network bandwidth usage required by the variety of applications, from mobile email and location-based services to VPN access and P2P communications, to, of course, video, the biggest bandwidth-consuming application of all.
What it means is that network operators must be more diligent than ever in monitoring, managing, and effectively allocating network resources to ensure a consistent, acceptable level of service for all users. To understand how Alcatel-Lucent's 9900 Wireless Network Guardian (WNG) provides wireless network operators a management platform that combines traffic analysis, network performance monitoring, and behavioral anomaly detection, helping operators better understand how their users' behavior impacts their networks - and how to adapt accordingly - read the white paper, "Alcatel-Lucent 9900 Wireless Network Guardian - Powerful solution to classify wireless data traffic, understand wireless resource usage and improve network performance."
