Spectrum dilemmas… where will it come from?
Earlier today I was talking with Yves Bellago of France Telecom about spectrum availability and whether 20MHz channels are possible. It was interesting to hear that he thought that 20MHz channels might be available in the 21GHz band and possibly the 1800MHz band but also believed that operators could get the benefits of LTE with 10MHz channels.
Yves Bellago, who is talking at LTE World Summit in the latter part of May, felt that 10MHz channels will provide improved spectral efficiency, capacity and latency in all bands that LTE services will require and I myself believe these 10MHz channels may also be obtained from the digital dividend. Opinions people??
April 16th, 2008 at 7:16 pm
The channels don’t need to be 10MHz! They can be as small as 1.25KHz depending on the requirements of specific applications!!
Operators can still roll out LTE and provide more basic data services at a much lower cost than they can with hsdpa and can acquire more spectrum at a later stage if they need to
April 17th, 2008 at 4:49 pm
When you are talking about “… 20MHz channels might be available in the 21GHz band … ” it is supposed that it is the 2.1GHz band. Isn´t it?
April 17th, 2008 at 4:51 pm
Won’t indoor coverage at 21GHz be hopelessly spotty? And also, won’t LTE terminals be able to use different bands simultaneously, bonding channels together to get a decent bandwidth even if it’s non-contiguous?
April 17th, 2008 at 11:43 pm
Verizon and AT&T have said they will use the 700Mhz spectrum they won recently for LTE whereas in Europe I would pick 1800Mhz as the most likely frequency as 2.1Ghz is already full of WCDMA / HSDPA / HSPA and 900Mhz will be used for HSPA. The 1800Mhz is mainly used for urban infill GSM currently and is the only one of these frequencies where there is enough spectrum remaining to feasibly deploy. Lets hope the equivalent of the US 700Mhz digital dividend arrives soon enough for that to be an option.
April 18th, 2008 at 5:57 am
20Mhz bandwidth probably ideal for very high speed wireless public data network. In practice, the 1.8G and 2.1G would have all been allocated for 3G, re-harvesting for LTE will naturally put a constraint on service continuity and the idea of fallback service coverage. Ideally a new band should be available like that for 2.3G/2.5G/3.5G being allocated for WiMAX.
May 12th, 2008 at 5:04 pm
LTE brings spectrum efficiency benefits (~4x HSPA / 2.5x HSPA+) whatever the spectrum bandwidth - ranging from 1.4MHz all the way to 20MHz.
Even if deployed in 1.4MHz channel, the subscriber will still benefit from peak rate close to HSPA (because of the much higher spectrum efficiency) and a very low latency of <20ms round trip.
This low bandwidth capabilities gives operators the ability to refarm some of their existing spectrum to deploy LTE in; for example an operator could deploy LTE in 900MHz in 1.4MHz to start with (only having to relinquish 7 GSM carriers or use some of the bandwidth left vacant by subscriber migration to HSPA) and as more subscribers migrate to HSPA or LTE, grow the LTE bandwidth in that spectrum. This way the operator can still keep GSM running in that same spectrum offering an LTE sub layer for when LTE coverage is not available + keeping all existing GSM services.
LTE ability to scale spectrum bandwidth brings operators a fantastic opportunity to refarm existing spectrum in small chunks (where both existing and new LTE services co-habit in that band) and make the most of new virgin spectrum (700MHz/2600MHz) by deploying wide channels (10-20MHz) and get the greatest capacity increase.
Most operators seem to want to deploy LTE in 2 bands – one coverage band (like 700, 900 or even 1800MHz for getting a wide footprint upfront and great indoor coverage) and one capacity band (AWS/2.6GHz for getting max capacity).