MWC – Highlights & Insights
Dawood Khan, Barcelona
Networks Reaching a Breaking Point?: Improving spectral efficiency; deploying smaller, smarter cells; offloading, but still have challenges
According to Telstra, while LTE increases the spectral efficiency between 30 to 60% over HSPA+, and it opens up the spectrum, reduces latency which all improve customer experience, by itself, it’s not going to resolve the loaming spectrum crunch challenges operators will face worldwide. The traffic demand is growing much faster than spectrum availability. 10 % of cell sites using over 50% of capacity today in Telstra net, over 90% using less then 50%
While governments are working on making spectrum available, it looks like different geographies will have different bands available. Many will use 700 – 800 MHz (Digital dividend spectrum), 1800-1900 MHz, as well as TDD bands, etc. This will cause cost increases in devices that will need to support multiple bands for global roaming. This is why its important to try to harmonize band allocation for LTE globally.
Several operators have suggested that traffic management is going to be critical to controlling network traffic; policy control and QoS will make a significant difference in delivering services to users in a reliable, high quality manner. For instance, a real time VOIP service will only utilize 50% of allocated network resources without QoS, whereas with QoS will have 90% utilization, which allows for much better utilization of the network.
Femtocells: The debate on these is still open, while many have touted the benefits of femtos , it seems that the value proposition is tailored not only to end benefit of offload and coverage, but also the message it delivers to the market. There are two main market messages:
- User deployed femots: Some operators don’t like this as it send a message that the operator hasn’t got a good network; In many such areas, already have wifi indoors so don’t really help with offload.
- Operator deployed femtos: for NW cost reductions in conjunction with macro and pico cells, these make sense in specific areas. Therefore, some carriers aren’t as positive on the promise of femtos.
Offload Strategies: Questions & Answers
Q: How much of a bottle neck exists that needs to be fixed with off-load?
- AT&T saw 3000% growth over last 3 years in data, and has continued to grow wifi network
- Designing for capacity isn’t simple, the concept of typical busy hour has evolved. Need for capacity changes based on location at different times of day – office during day; home during the PM, and the home traffic is rising
o Cisco says that 200% growth p.a. in data growth at home, and by 2012 a large percentage of the traffic will be from homes
- Capacity Crunch will occur in several areas (Cisco), must have an end to end solution:
o Radio: Spectral efficiencies via LTE; offload
o Backhaul: still TDM, ATM and early stage Ethernet, which doesn’t scale economically. Ethernet and IP increasingly growing in this space
o Packet core: not all traffic needs to be in the packet core, consider network offload using intelligence in network to steer away non-montizable traffic
o Data center: as video becomes more prevalent, intelligent content delivery and distribution networks will help
Q: Whats best for offload – wifi or femto?
- Wifi is built into devices pervasively and there is a possibility to use wifi as an offload option
- Femto can be of value as a result of things like the RNC capability in the femto, which allows to offload the network.
- Un-managed offload of traffic to date from cellular is between to 25-35% based on Cisco visualization matrix
Net Neutrality & Spectrum Regulation: Operators have to prepared to consider spectrum and network sharing for LTE
Gabrielle Gauthey, EVP ALU discussed her perspectives on net neutrality. The crux of her message was that while regulators need consider the practicality of requirements such as transparency when it comes to wireless, but that carriers also needed to become more innovative and embracive of business models. Summary of points:
- By 2015 will have 13000 smartphones/sq.km vs. 400 smartphones/sq.km today in urban cores
- Decoupling of traffic and revenue has occurred and will continue to change
- Much of the innovation and growth comes from over the top players who enjoy latitude in business models, global models, and high valuations vs. operators who are usually local
- Net neutrality should not hamper new business models – in the UK as long as there is transparency and non-harmful discrimination, wireless is excused from net neutrality vs. wireline. However, there are challenges with providing full transparency with mobile, advertized vs. random bitrates; laws of physics challenges in throughputs adjusting with location and cell loading; impacts of offloading to wifi, etc.
- Traffic management is the only way to manage traffic congestion and provide good data throughput to subs for critical services
- To increase capacity
- Densification of cells via femto and pico – lack of space for towers, etc. so operators need to consider the incentive to share and sites
- Landline usage via wifi offload
- Spectrum – will be limited, we need to be prepared to share spectrum and revisit our business models
- LTE is the only technology that uses 20 MHz channels, to use the technology most effectively. But difficult to see the duplication of 3 or 4 parallel networks in high urban areas.
- Therefore see spectrum, network sharing across LTE by carriers
How can over the top providers work effectively with operators?
- Even though video consumes the most traffic, its number four on priority for users behind browsing and email. So some users would be looking to have pricing based on services they most value. Net-neutrality – FCC doesn’t want operator to charge of differentiation, people may not pay for differentiation but do pay for access.
- Netflix video represents 20% of traffic, Level3 and Comcast were pointing to each other saying that the other had to cover the cost. We have to revisit our business models