4G Capacity Gains study: 3.3x gain over today’s 3G networks

Ofcom yesterday published the outcome of our 4G Capacity Gains study developed by Real Wireless. At 120 pages for the main report and a further 90 pages of appendices, we thought it would be useful to give a quick summary of the key findings. In short:

The report was commissioned by Ofcom to assess the realistic scope for capacity improvements in 4G technologies (including LTE, LTE-Advanced, WiMAX and WiMAX 2) relative to earlier technologies, especially 3G (including UMTS, HSPA and HSPA+).

Wireless Network capacity depends on three key factors:

• Technology: the cell spectrum efficiency that can be realised by a given feature set from a given wireless technology
• Spectrum: the bandwidth allocated to a network
• Topology: the mixture of cell sizes and local environments in the network

This study focuses on Technology and Topology as a contribution to the debate on dimensioning of spectrum for 4G networks.Real Wireless found that technology evolution alone will deliver substantial gains in spectrum efficiency, but that these taken alone will not be sufficient to meet forecast traffic growths without also having increases in the available quantity of spectrum and moves towards more numerous and smaller cells.This finding is based on:

1. Their forecast that spectrum efficiency in UK cellular networks in dense urban areas could grow at 18.5% annually (CAGR) between 2010 and 2020 due to technology enhancements introduced via the predicted roll out of 4G networks.

2. The lowest of analyst mobile broadband traffic forecasts which show an annual growth rate of 17% in the same time period. Most traffic forecasts are many times higher this low end forecast, with the highest traffic forecast in 2020 being around 15 times higher than this.

3. Initial 4G deployments will deliver a 1.2 times improvement over high end 3G configurations with considerably larger gains when compared with earlier 3G systems. We estimate a 1.2 times improvement in cell spectrum efficiency between a high end 3G configuration (2×2 HSPA+ 64QAM Release Eight) and typical initial 4G deployment (2×2 LTE release 8). Comparing against a more typical high end 3G device on the market today such as a HSUPA release 6 1×1 handset gives a gain of 3.3 times.

4. Spectrum efficiency in deployed networks will grow by approximately 5.5 times between 2010 and 2020 Our forecast roadmap of 4G feature set roll out and device mix produces a blended spectrum efficiency for 4G over 10 years in an ITU base coverage urban macro environment of 1.5 bps/Hz in 2010, reaching 2.32 bps/Hz/cell by 2020. Combining this with the spectrum efficiency of 2G and 3G devices already deployed we estimate that the spectrum efficiency of UK networks will rise from 0.43 bps/Hz/cell in 2010 to 2.32 bps/Hz/cell by 2020 i.e. 5.5 times. This gives a CAGR of 18.5% which only satisfies the low end traffic forecasts.

5. Their findings on absolute spectrum efficiency challenge assumptions in previous spectrum dimensioning studiesPrevious studies on behalf of Ofcom and the FCC into dimensioning spectrum assume values from 1.3 to 1.5 bps/Hz/cell for LTE. This matches reasonably well with our estimate of cell spectrum efficiency for 2×2 LTE release 8 of 1.32 bps/Hz/cell in an “idealised” ITU base coverage urban scenario. However, this assumption is optimistic for spectrum dimensioning as it does not allow for realistic loading or the impact of mixed traffic types.

6. Capacity gains via small cell topology will gain importance in 4G. They anticipate that network topology will become important to capacity in 4G networks to an even larger extent than in 3G networks. This is based on:

• Capacity gains from technology not being sufficient to meet forecast demand increases so that a combination of spectrum and topology will be needed to address the balance.
• Small cells being built into the 4G standards from an early stage, building on experiences from 3G femtocells.
• Consistent view from stakeholders that small cells and the associated heterogeneous networks will be an important part of delivering sufficient capacity in 4G networks.
• OFDMA, as used by 4G technologies, lending itself better to interference mitigation via dynamic resource scheduling across frequency and time resources than WCDMA as used in 3G.

7. Small cell topology improvements in 4G should mitigate against demand peaks so that dense urban deployments are still the bottleneck for dimensioning spectrum Our study has found that using small cells in three example “hyper dense” scenarios in urban office, shared accommodation and railway stations would be enough to provide the capacity boost needed to mitigate the demand peak. This implies that dimensioning spectrum for dense urban deployments and assuming that topology improvements will serve peak demand in these hyper dense scenarios is still a reasonable assumption for 4G networks. Our findings rely on small cells being cost effective to deploy and that vendors can overcome interference and management issues.

Please find the report below. Happy reading, CVA.