4G: Glitz, glamour or the grave?The lucrative promise of 4G is luring mobile operators to announce their migration plans in an effort to ensure their brands are viewed as cutting edge, followed by massive investment to overhaul their existing networks while betting on a long-term payback. As operators introduce mobile data services to the mass market today, it is imperative that they are perceived by the customer as “sexy” and of a very high quality level. The problem is that mobile operators currently involved in early data trials over 3G HSPA/EVDO networks are experiencing many problems that must be overcome before a 4G migration can even be considered. Beyond technical network issues, new problems related to service quality and availability are challenging the operators. Because many of these issues have never been seen before, most operators are ill equipped to address them.
Network-level issues
As mobile operators embark on a migration path with the eventual goal of 4G, they will experience a burst in network usage. The impact will be the same as that experienced after the deployment of DSL: the network resource impact from the shift of analog modems to DSL was enormous. The evolutionary path from 3G to 4G will result in a similar phenomenon.
Historically, access has been a bottleneck for data services, with the core network tailored to the access bottleneck. Thus the evolution of access from 3G to 3.75G (HSPA) and then to 4G must proceed in tandem with the migration of the legacy core network to an all-IP core network, while simultaneously supporting legacy voice services and introducing new high bandwidth data services. To achieve this, active network monitoring and ultra-fast reaction times become of paramount importance.
Monitoring the network is not simple, however. There are many signaling exchanges and interfaces involved during data transactions, each giving a different level of visibility into network and service issues. While deploying probes everywhere in the network is unaffordable and inefficient, there are key network points that provide the required visibility.
For example, in a UMTS HSPA network, one of the key interfaces is the Gn, which represents the focal point where user data can be analyzed and where services can be associated to the individual or to groups of subscribers.
By combining information from the control plane and the user plane in the Gn interface, operators can extract several kinds of information such as subscriber ’group’ behavior, QoS, quality of experience (QoE, described below) and Key Performance Indicators (KPIs). By correlating the control and user planes, the operator now has visibility into how users affect the network and how the network affects the users.
Service-level issues
Compared with voice, mobile data services require many new monitoring and troubleshooting techniques. This is mainly due to the fact that for voice, it is generally assumed that simple network metrics (e.g. those associated with classics QoS) can ensure good voice quality from users’ perspective – especially for legacy networks such as SS7 and ISDN.
For data services, this does not always apply. For data services, QoE – that is, understanding how the user experiences the service – now becomes the important metric.
In addition to latency, delay and jitter, which are well-understood by operators, new measurement types must be mastered in order to troubleshoot data services and ensure high QoE, including TCP statistics (such as retransmissions, resets, round-trip time and set-up time provide a window into quality of sessions), packet loss (essentially ‘jitter’ for streaming data services), protocol anomalies (how malformed packets effect TCP and UDP transmissions), throughput per session (the actual data rate the call is occupying) and traffic per session (how much aggregate bandwidth the call used).
Additionally, the ability to proactively address potential problems before the user experiences them becomes important. By correlating real-time analysis on user and group-level behavior with actual network and node behavior, the operator can predict and alarm on emerging service issues such as network congestion, service breakdown and roaming/handoff problems. For example, a monitoring system can generate alarms based upon thresholds to detect degradations of specific traffic indicators grouped by various destinations.
Another important requirement is a fast reaction cycle from the time the front office receives a trouble ticket to the time the back office is able to determine the root cause of the problem. Issue resolution will now be measured in minutes, rather than hours or days. Again, effective use of troubleshooting and monitoring tools are mandatory. For example, the operator should have the ability to rapidly drill down on detail records such as calls, transactions and multimedia services, and also to search by variables such as phone, transaction and video ID.
Further they should have the ability to recover and zoom in on any element of the frame(s) related to that traffic.
High stakes
Before considering a migration path towards a 4G infrastructure, mobile operators need to address the complexities and issues associated with current 3.5G and 3.75G technologies. A new generation of test and monitoring tools exists that can enhance the success of these deployments and accelerate the migration path towards 4G. These tools provide visibility into both network and service issues and correlate them to ensure the highest QoS and QoE achieved for both legacy voice and new mobile data services.
The stakes are high, however. Selection and proper utilization of the right troubleshooting and monitoring tools will help the operator achieve glitz and glamour in the eyes of the customer, while improper selection and utilization will hasten the operators inevitable trip to the grave.
Wireless Mobile Telecom Wireless News
