Unpaired spectrum could offer a cost-effective solution for mobile operators that want to meet the growing demand for data. However, several drawbacks of time division duplexing need to be overcome to unlock the capacity potential of this spectrum.
Mobile operators are having to meet growing demand for data, and yet their ARPUs are flat. They therefore need cost-effective solutions to deliver more data. One way of adding capacity to a network is to provide it with more spectrum. Conveniently, a lot of unpaired spectrum is becoming available in higher frequency bands, which is typically less expensive to acquire than paired, sub-1GHz bands.
Unpaired spectrum can be used by transmission technologies using time division duplexing (TDD), which allows both the uplink and downlink to be carried by the same frequency band. In addition, the use of the TDD mode of LTE allows for an asymmetric uplink/downlink ratio (see Figure 1), ideally suited to cater for the increasingly asymmetric data consumption that might be expected in the future. Operators are therefore looking carefully at whether the use of TDD in unpaired spectrum may be a solution to the capacity crunch that they are facing.
Figure 1: Difference in spectrum use between FDD and TDD [Source: Analysys Mason, 2012]
There are a number of drawbacks to TDD that must be overcome. In earlier generations of cellular technology, such as 3G, a major issue was that the integration of TDD with FDD did not emerge in a timely way in most world regions. One reason for this was that the TDD and FDD modes of 3G use different air interface schemes. However, in LTE both the TDD and FDD modes are based on a similar scheme – orthogonal frequency division multiple access (OFDMA) – resulting in greater commonality between the two modes, which should facilitate TDD–FDD integration.
One drawback of TDD is that it requires a guard period between the uplink and downlink transmissions during which the signal can travel between transmitter and receiver before the direction of the communication is reversed (see Figure 1 above). However, this gap is proportional to the distance between the transmitter and the receiver, so the inefficiency is minimised if TDD is used to provide short-range services, as in capacity cells.
Another issue with TDD is that operators using adjacent blocks of spectrum need to synchronise their networks so that both are sending uplink and downlink transmissions at the same time. If there is a misalignment, the base station trying to receive uplink traffic will receive interference from the base station transmitting downlink traffic. The same problem can occur between handsets. Currently, synchronisation requires operators to put in place a separate network element that aligns the two networks’ transmissions. However, the TDD mode of LTE-Advanced may simplify the process through the use of ‘over the air’ synchronisation, thereby making synchronisation more affordable to implement. However, the adoption of this into the standard is yet to be confirmed.
The success of unpaired spectrum depends on the availability of devices that can use it. Traditionally, devices are designed to operate in up to four bands, although more-recent developments suggest that support for at least six bands may be required for LTE. For reasons of backwards compatibility and international roaming, three of these are likely to be taken up by the low-frequency coverage bands of GSM, HSPA and LTE networks used in different world regions. In addition, operators will employ a number of FDD capacity bands, such as the 1800MHz, 2.1GHz and 2.6GHz bands. Therefore the fourth slot (and possibly the fifth and sixth) will be contested by a number of options. There are two solutions to this problem: either manufacturers will find a way to cost-effectively combine more bands in a single device, or they may customise devices for individual operators by allowing them to choose the bands to be supported. The second option would result in significantly higher prices because of the reduced economies of scale possible for customised devices.
This may leave operators with the option of focusing their TDD services on USB modem users, as the bands supported by a USB modem can be customised at little additional cost. However, this would not allow operators to fully exploit newly acquired unpaired spectrum because the volume of USB modem traffic is insufficient to fill the additional capacity – at least so far. Manufacturers will also require a number of years to adapt to demand for devices with more bands or customised bands. Therefore, the capacity potential of unpaired spectrum has yet to be unlocked.
Source: Analysys Mason – http://afridigital.blogspot.nl/2012/08/unpaired-spectrum-and-promise-of.html#!/2012/08/unpaired-spectrum-and-promise-of.html