Page 196 - From GMS to LTE

182 From GSM to LTE-Advanced Pro and 5G

frame which channels to assign to which users. As shown before, the HS‐SCCH channels
are used to inform the mobile devices which channels to listen on for their data. This
task is called scheduling. To quickly react to changing radio conditions of each
subscriber, the scheduling for HSDPA has not been implemented on the RNC as for
other channels but directly on the Node‐B. This can also be seen in Figure 3.38 as the
HS‐SCCHs originate from the Node‐B. This means that for HSDPA, yet another task
that was previously located in the RNC for DCHs has been moved to the border node
of the network. This way, the scheduler can, for example, react very quickly to deteriorating
radio conditions (fading) of a mobile device. Rather than sending frames to a mobile
device while it is in a deep fading situation and thus most likely unable to receive
the frame correctly, the scheduler can use the frames during this time for other mobile
devices. This helps to increase the total bandwidth available in the cell as less frames
have to be used for retransmission of bad or missing blocks. Studies like [15] and [16]
have shown that a scheduler that takes channel conditions into consideration can
increase overall cell capacity by about 30% for stationary users. As well as the signal
quality of the radio link to the user, scheduling is influenced by other factors, such as the
priority of the user. As with many other functionalities the standard does not say which
factors should influence scheduling in which way, and thus a good scheduling
implementation by a vendor can be an advantage.
As the RNC has no direct influence on the resource assignment for a subscriber, it is
also not aware how quickly data can be sent. Hence, a flow control mechanism is
required on the Iub interface between the RNC and the Node‐B. For this reason, the
Node‐B has a data buffer for each user priority from which the scheduler takes the data
to be transmitted over the air interface. To enable the RNC to find out how much space
is left in those buffers, a Capacity Request message can be sent to the Node‐B, which
reports to the RNC the available buffer sizes using a Capacity Allocation message.
It should be noted that a Node‐B does not administer a data buffer per user but only one
data buffer per user priority.

3.10.4 Adaptive Modulation and Coding, Transmission Rates and Multicarrier
Operation

To reach the highest possible datarate during favorable transmission conditions, several
new modulation schemes have been introduced with HSDPA over several 3GPP
releases, in addition to the already existing QPSK modulation that transfers 2 bits per
transmission step:
16‐QAM, 4 bits per step. The name is derived from the 16 values that can be encoded
●
4
in 4 bits (2 ).
64‐QAM, 6 bits per step.
●
Two simultaneous data streams transmitted on the same frequency with MIMO.
●
To further increase the single‐user peak datarate, dual‐carrier HSDPA (also referred
to as dual‐cell HSDPA) was specified to bundle two adjacent 5 MHz carriers. At the
time of publication, many networks have deployed this functionality. In the subsequent
versions of the standard, aggregation of more than two carriers was specified, as well as
combination of 5 MHz carriers in different bands. However, because of the quick adoption
of LTE, it is unlikely that these features will be seen in practice in the future.