Money to invest in mobile technology is scarce and every resource now needs to be squeezed to deliver its greatest potential. No where is this more obvious than in the data strategies being pursued by many of Europe’s leading operators as they look to GPRS to deliver much more than originally planned. Catherine Haslam reports on the OSS developments that are helping them achieve this.
GPRS is, as we are all aware, a best effort technology. In its basic form, all that anyone can say is that GPRS usually delivers a throughput somewhere around 12–15kbit/s (including overheads) per timeslot. However, those speeds can and will vary dramatically depending on the amount of traffic in each cell. This is where the problems lie because while 30-40kbit/s is good enough for a host of applications, the vast majority require some level of reliability. It is not so much a case of taking GPRS to its maximum theoretical potential but rather delivering an acceptable and consistent transport platform on which applications and business cases can be built.
This is the position we are in now. While not exactly mature, GPRS networks across Europe have reached a certain level of reliability and, as traffic levels are reasonably low, there is little trouble with the quality of service. However, as Russell Crawford, product manager at Mycom readily accepts, if services such as MMS etc really take off then GPRS will offer much more variable throughput rates.
Operators face a choice. They can either deploy solutions capable of monitoring and managing GPRS traffic, or attempt to defer the quality issue by rolling out the greater bandwidth and inherent Quality of Service (QoS) management of UMTS as fast as they can.
In simple terms this is a choice that mirrors that experienced in the fixed world — manage traffic more effectively or throw bandwidth at the situation. The difference is that bandwidth is always going to be a limiting factor for wireless communications in comparison to fixed, even with 3G. The excessive bandwidth associated with fibre is just not applicable in the wireless world and therefore getting the most out of what we have remains the mobile mantra. Indeed, never has this been more true than it is today, as faced with unfavourable market conditions Europe’s mobile operators are increasingly looking at getting everything they can out of GPRS and not just viewing it as a stepping stone to 3G.
It was perhaps no surprise therefore, that GPRS traffic management was a major theme of 3GSM — operators were demanding solutions. And, on the other side of the equation, businesses are recognising the possible benefits. Indeed, the momentum behind mobilising corporate data is such that some believe that failing to deploy it will not just mean that they miss a growth opportunity but will lose business because of it. However, enterprises are used to working with SLAs and therein receiving a certain level of predictability from their data connectivity.
Dynamic shaping
sraeli company CellGlide has taken this as the starting point for its solution which dynamically shapes bandwidth to the requirements of specific traffic loads. “Traffic shaping is a well-known phenomenon in wireline and is based on the ability to deal with bursty traffic, most specifically in IP networks,” says Liron Langer, ceo of CellGlide. He further explained, however, that the fixed experience could not simply be transplanted into the mobile arena, “Most people either concentrate on the optimisation of applications or of the radio transmission but you really need to do both.”
Laure Reiller, senior product manager at WatchMark explains further, “Performance management has been about monitoring and measuring the performance of network elements in order to monitor the performance of the voice services. With data, it’s a different game. It’s not just a pipe service anymore, now there are lots of ingredients — the RAN and the core networks, applications, partnerships with content providers etc — which all go to make up compelling services.
Measure to manage
To manage data, you need to be able to monitor it, identify what sort of traffic is being transmitted across the network and see what network resources are allocated to it — ie you need to monitor the existing and new network elements and those other, now vital application and content layers.
A view of the network can be derived with many of the same tools as are in place today. Performance counters and mobile RF measurements from drive tests etc, together with the configuration parameters and network elements traces, all provide vital information on the way the mobile network is performing. However, GPRS throws in the added complication of data fighting with voice traffic for timeslots. Therefore, at peak times, capacity could quickly become an issue. Ironically, it will be those mobile operators that are successful with their data strategies that are likely to encounter problems first. That is unless they have significant amounts of excess capacity currently lying dormant and that is unlikely.
In many of Europe’s larger cities, capacity is already causing operators problems. Add to this the change in attitude towards 3G that has left many incumbent mobile operators looking to maximise their investment in GPRS rather than rush to deploy UMTS which actually has QoS provision inherent in the Release 4 standard, and the need for more sophisticated OSS solutions to manage data is pressing.
End-to-end view
In terms of what needs to be done, the key is providing an end to end view of what is happening to the service a customer receives and being able to manage the service so that it is delivered effectively as often as is possible. There is little value in knowing that a service is failing if it is impossible or long-winded to identify where the problem is or, conversely, what impact the failure of specific network elements has on services. However, as David Heaps, senior vice president at ADC Metrica states, “Today, most operators don’t have an end-to-end view of the service quality.”
To get that end-to-end view requires a system which looks beyond the basic network structure, or perhaps more accurately, builds on the information derived from the mobile network. At a service level, Key Quality Indicators (KQI), probe systems and SNMP counters add meat to the bones, while a customer view can be gleaned from the call detail records (CDRs), core network records and subscriber traces. Combine all of these with the traditional information sources already mentioned, and you begin to establish a level of understanding about what service level is being delivered.
Exponential rise
As Mycom International’s Crawford explains this is no simple task, “There are a huge number more interfaces to addressed and this has to be done in a multi-technology and multi-vendor environment.” However, not only does all this need to be monitored, it also needs to be managed and due to the massive increase in data which is needed to give an accurate view of the network and the equally large and complicated options for management there is a need to automate and simplify reporting and some actions.
To this end, WatchMark has created an MMS addition to its Prospect product. The rationale behind this is that, “Most 2G operators have performance management systems they are happy with but MMS gives a compelling reason to change,” according to Reiller. What the product gives is the ability to create reports against a set of pre-defined Key Performance Indicators (KPIs) which can either be created by WatchMark or by the operator very quickly to see just how well specific services are working. However, while MMS provides a reason to change in WatchMark’s view, this is not about throwing away existing systems but building on them.
Evolution not revolution
Certainly, the ability to work with the existing systems and drive all information through a central system which delivers the information necessary in an evolving environment is at the heart of Mycom’s solution, NIMS-PrOptima. “Firstly, we have architected the system specifically for 2.5 and 3G,” says Crawford. But, he explains, “Operators want a single solution that has to be flexible enough to fit with what they have and take the expertise that is there.”
The result is a modular solution that builds in a layer of abstraction at the interfaces. This enables NIMS-PrOptima to work with all the layers required as well as any existing systems. The ace in the Mycom hand, however, is that the management side has a level of automation. NIMS-PrOptima learns from the actions taken in response to certain issues and therefore prompts responses and holds much of the expertise in the system, rather than in the minds of a few specialists.
As with every other area of the mobile operator’s infrastructure at the moment, solutions which cannot work with, or build on, what is already installed, will be rejected out of hand. Backward compatible, future-proof technologies which are both scalable and reliable are the order of the day. Fail to meet these basic requirements and there will be no future for the solution. However, at the same time, many operators recognise that this is one area in which doing nothing will not only fail to improve their service delivery capability but will actively lead to its deterioration.