Guest columnist Othmar Kyas, senior technical advisor at Tektronix Communications sees LTE as the winning 4G technology, but with work on its development still to do.
While it took 3G UMTS more than five years of costly deployments before the technology had enough critical mass to demonstrate that it was going to succeed, LTE has had a much easier start.
The volume of broadband mobile data traffic has increased explosively, driven by the launch of the iPhone and other smartphones. By the end of 2010, mobile broadband subscribers are expected to have exceeded fixed broadband subscriber numbers. It became clear to telecoms operators that they would have to upgrade their networks sooner rather than later.
Until a year ago, 4G was still referenced as a set of four competing technologies: LTE, Wimax, UMA and UMB. LTE has since emerged as the winning technology. UMA and Wimax will remain niche technologies. UMB was stopped by Qualcomm in November 2008.
Operators and equipment manufacturers have opted to develop 4G technology only once, which will get us to a single, compatible global communications infrastructure in 10 to 15 years.
Read our full Mobile World Congress coverage
Technical challenges
While the global agreement to move forward to 4G, with LTE as a single technology, is a big opportunity in terms of deployment speed and cost, feasible operational models for LTE are still in the workshop.
Operators face the challenge that mobile broadband demand is exploding, while average revenue per user (ARPU) levels remain flat. Stiff competition has pressed operators to offer flat-fee mobile broadband packages. This forces operators to build a business model that decouples network cost from traffic volume.
The mobile industry is moving towards the same status as fixed broadband today, where ARPU levels can, at best, be kept at the same levels by offering more bandwidth at the same price every 12 months. The only way to get there is to transform the network infrastructure into a highly efficient all-IP packet-switched matrix, which is what LTE was designed towards from the first.
Next-level networks
On the operation and maintenance front, LTE also takes communications networks to the next level.
Key operational aspects such as configuration, optimisation and fault identification are planned to be automated to minimise operating expenses. Using a central configuration server, new network elements will be automatically associated with the network.
Repetitive tasks such as neighbour list optimisation, coverage and capacity optimisation, mobility robustness optimisation and mobility load balancing optimisation will be conducted locally at the eNodeB, using robust, autonomous algorithms.
Automated fault detection
Using a set of well-defined rules and selective KPIs (key performance indicators), frequent faults will be identified automatically and dealt with at a tier-1 response level. Outages of entire cells will be compensated using pre-defined neighbour cell management strategies.
The above autonomous, auto-operational processes will be controlled by tier-2 instances at the MME and OMC, which control and manage local activities in the larger context of a network segment or the entire network.
The traditional diagnostic functions move up to this meta-layer. With these auto-functions (self optimising network functions, or SON) playing a key role in network operations, their operation needs to be assured through well-defined troubleshooting, validation and certification procedures.
While, on paper, LTE has all it needs to get us to 4G, all that remains is to make it work. Key for the operators will be to plan towards an operational model, which for the first time delivers ‘IP-class’ efficiency in mobile telecommunications.
Complex auto-functions, huge amounts of ciphered user-plane traffic, complex QoS and QoE commitments based on countless service combinations, and the complex OFDMA/MIMO radio interface needs to be managed.
Things will continue to go wrong, but operators, with their network test, measurement, optimisation and infrastructure suppliers, will have to resolve the problems in a very different manner. Then, and only then, the way to 4G may be long, but it will be successful.
Author is Othmar Kyas, senior technical advisor at Tektronix Communications