Gain of Dual‐Cell HSDPA in Macrocellular HSPA+
The LTE-Advanced tool set of course comes with LTE specific naming scheme: higher order MIMO, Coordinated Multipoint (CoMP), Small Cells, Heterogeneous Networks (HetNet), and Carrier Aggregation (CA).
Now that the LTE-A networks and devices are hitting the markets with high peak bitrates trumpeted by the industry, I think it’s a good time for a reality check. The whole LTE-Advanced, what does it mean for networks in practice?
By using all bells and whistles – that is by aggregating five 20 MHz carriers to a 100 MHz chunk, deploying 8×8 MIMO with small cell sizes, using terminals with advanced receivers, etc. – LTE-A specifications go up to 3000 Mbps. That kind of speeds will however not come easily, with a reasonable price tag or in a timely manner to any large-scale commercial networks.
LTE-Advanced Carrier Aggregation is currently limited by 3GPP specs, which only allow for combining two 20 MHz LTE carriers instead of maximum five. Using 8×8 MIMO still ensures that 1000 Mbps overwhelming bit rates can be reached.
Despite the cost and implementation challenges, we might eventually see 8 TX antennas per sector in some extreme cases. It is however practically impossible to put eight sufficiently separated receiver antennas even in the most supersized phablets, tablets or PCs. For anything truly mobile it is more feasible to assume 2 antennas. Two antennas is also where we have been the past couple of years – for the obvious reasons.
Now, aggregating two 20 MHz LTE-A channels, the maximum specified spectrum, and using topical 2×2 MIMO configuration we end up peaking 300 Mbps. Still much more than what is currently available.
To be able to provide 300 Mbps bit rates, operators need 40 MHz of bandwidth in suitable bands. In Europe LTE is mainly operated on 800, 1800 and 2600 MHz bands. 2600 band, although with ample spectrum, is limited by poor propagation properties – which means Carrier Aggregation gains in very limited areas. 800 band is only 30 MHz wide, which usually means 10 MHz chunk per operator – Still not enough for 40MHz CA in most cases.
The cornerstone of carrier aggregation will definitely be the 1800 band, which in many cases is still at least partly occupied by GSM. Luckily it can be often re-farmed for LTE. But given the total of about 60 MHz available in 1800 band, it will be challenging for any European operator to allocate 40 MHz for LTE-Advanced in short term. Assuming total of 30 MHz (20MHz @1800 +10MHz @800) would be more realistic in short term. That would then result in peak rate with realistic spectrum just hardly above 200 Mbps. With Network Sharing, e.g. MOCN concept, the situation would be very different though.
Recently launched iPhone 6 indeed supports LTE-Advanced and carrier aggregation, but it is limited to 150 Mbps. The limitation comes from the capability of aggregating only two 10 MHz LTE chunks – a feature obviously tailored for the US market, where operators’ continuous spectrum assets are not so generous. First commercial 300 Mbps capable category 6 smartphones have been already announced – Samsung leading the front with the all new Galaxy S5 LTE-A Broadband and Note Edge. More category 6 devices will surely flood into market in the course of 2015, especially when competition comes with less expensive chipsets.
Actual user bitrates are hindered by many factors such as the radio propagation, network load, subscription limitations, and – sad to say – often also by sub-optimum radio network design. Hence it’s rather obvious that user bitrates will have large instantaneous fluctuations and thus fall far from 150 Mbps. According to our tests, the leading European operators today typically deliver 20-40 Mbps user data rates with the plain vanilla LTE. Knowing the advanced part of the story I suspect that in the course of 2015 and 2016 we are still going to stick below 100 Mbps user data rates in majority of practical networks.
100 Mbps would be nice, but where do I actually need such high bitrates in first place? My point is, killer apps driving the growth do not actually need very high bit rates. Google Maps, the most popular smartphone app, is pretty ok already at 5 Mbps. The same goes for web surfing. Even Netflix FullHD comes quite nicely with 10 Mbps. LTE-A benefits for a typical subscriber are thus probably rather limited in terms of Quality of Experience, which renders higher service fees hard to justify. On the other hand, better food boosts appetite: given good service the subscribers tend to use more data. One more reason to abandon flat rate pricing models.
LTE-Advanced should be seen as a capacity extension to existing networks. Deploying Carrier Aggregation to traffic hot spots could be a true low hanging fruit – notable capacity gain without a massive rollout campaign. Just thinking out loud, given the obvious benefits of Carrier Aggregation and the limited band availability, LTE-A sets a totally new tone of light on Network Sharing. Two operators combining spectrum assets in an oligopoly of three players – should the regulator allow such manoeuvre – could seriously rock the boat and swing market shares with Carrier Aggregation.
The total feasibility and ROI of deploying LTE-Advanced depends heavily on vendors’ pricing models, properties of the existing network and terminal development. As the issue is complex and big investments are included, a solid technology strategy is imperative. Operators should carefully analyse the business case and achievable capacity gains of LTE-Advanced with proper network analytics. LTE-Advanced is surely the right way to go, but for optimum, or positive, ROI the timing is paramount. Failing to plan is planning to fail. Still, if little or no LTE is deployed in your existing network, you want to ensure that LTE-A aspects are captured by any future rollouts.
But there is also a take out for network vendors: Make the pricing reasonable and you might see the desired technology adoption. Most European LTE operators have plenty of capacity in their LTE networks, so they are not keen to spend fortunes just to make their LTE’s a bit more advanced. More business incentives are needed.