Variations
of instantaneous transmission power
- The single biggest challenge of multi-carrier transmission is
the corresponding large variations in transmission power and the related
high peak to average ratio (PAR).
- Higher modulation (16, 64 QAM) also contributes to variations
in transmission power.
- High PAR leads to reduced transmitter power amplifier efficiency
and higher power consumption which increase the complexity and cost of the
power amplifier. This is due to the dynamic range of the power amplifier
which can only be linear in a limited interval.
- This is of no or little in concern in the DL where power supply
issues and cost of power amplifiers are relatively small. However, in the
UL it will have a clearly negative impact on the mobile terminals cost and
battery drainage.
- The reason for including the single carrier component of the
LTE uplink (SC-FDMA) is the lower Peak-to-Average-Ratio (PAR) that it
results in.
Scheduling
- Scheduling decisions in LTE can be made as often as every 1ms
(per TTI) and the frequency granularity is 180 khz (12 sub-carriers).
- There are 3 different types of resource block allocation.
- Resource block allocation type 0&1 both support
non-contiguous frequency allocation but type 2 only supports contiguous allocation
in the frequency domain.
- Type two allocation does not have to use a bitmap and instead
only indicates a start position and a length indicator. In this way the
number of required bits is decreased.
- The basic scheduling unit is a so called resource block which
is a space in the time-frequency domain spanning 180khz (12 sub-carriers)
for 0,5 ms (slot).
- The minimum scheduling resource that can be assigned is two
resource blocks during one sub-frame (2 slots) also called a resource
block pair.
- The terminals are monitoring the PDCCH transmissions for
scheduling decisions and information that is required to demodulate the
transport blocks.
- Scheduling can be made both in the frequency as well as in the
time domain.
Multi
antenna support
- Multiple receive/transmit antennas can be used for diversity,
beam-forming and spatial multiplexing(MIMO).
- The e-nodeB controls the multi-antenna scheme that is used for
each transmission
Terminal
States
In contrast to WCDMA/HSPA, there are only
two RRC_states defined in LTE for the terminal.
RRC_IDLE:
·
A UE specific DRX may be
configured by upper layers.
·
UE controlled mobility.
·
The UE monitors a Paging
channel to detect incoming calls, system information change, ETWS notification,
and CMAS notification.
·
Performs neighboring cell
measurements and cell selection/reselection.
·
Acquires system information.
RRC_CONNECTED:
In this state the UE can be IN_SYNCH and OUT_OF_SYNCH. If the UE is determined to be out of synch a new random access procedure has to be performed.
·
Transfer of unicast data
to/from UE.
·
At lower layers, the UE may be
configured with a UE specific DRX.
·
Network controlled mobility,
i.e. handover and cell change order with optional network assistance (NACC) to
GERAN.
·
The UE monitors a Paging
channel and/or SIB1 contents to detect system information change, for ETWS
capable UEs, ETWS notification, and for CMAS capable UEs, CMAS notification.
·
UE must monitor control
channels associated with the shared data channel to determine if data is
scheduled for it.
·
UE must provide channel quality
and feedback information
·
UE must perform neighboring
cell measurements and measurement reporting
·
UE must acquire system
information.
Multiple
retransmission schemes
- Similar to HSPA, the protocol part of HARQ is handled in MAC while
the actual soft combining is handled in the physical layer.
- HARQ is optimized for an error rate of 10% which means that the
resulting received bitrate (transmission rate - error rate) is maximized
at that point. At any other rate, the HARQ system would be over or
under-utilized.
- RLC on the other hand, should be utilized much less frequently
and can provide an error rate of 10^-5. Resource consumption at this rate
is not an issue due to the much lower rate of transmissions.
- TCP should be configured to receive errors at a rate no higher
than 10^-5 if a high bit-rate transmission is intended. This is due to the
fact that TCP interprets all errors as a result of congestion and will
lower the bit-rate accordingly.
- Seen as a single combined retransmission scheme, HARQ provides
speed and RLC provides reliability. In E-UTRAN cooperation between RLC and
HARQ has been enhanced since they both reside in the e-nodeB.
- An asynchronous HARQ protocol implies that retransmissions can
take place at any time and not only at certain intervals (synchronous
HARQ)
- An adaptive HARQ protocol implies that the frequency location
may change between transmissions
- For LTE, DL is normally asynchronous and adaptive while UL is
normally synchronous and non-adaptive although adaptive is possible.
- The actual timing when a certain ACK/NACK is received is used for determination of which specific HARQ process it belongs to.
Equalization
- Historically, the main method to mitigate the adverse effects
of a frequency selective channel has been to use different kinds of
equalization on the received signal.
- Maximum Ratio Combining (MRC): the filter impulse response has been
chosen to provide channel-matched filtering which is the complex conjugate
of the time reversed channel impulse response. Mathematically speaking,
this is equal to multiplying the signal with one which means that the
impact of the channel has been removed . This is used in the RAKE
receiver.
- MRC maximizes the post filter signal-to-noise-ratio but does
not provide any real equalization.
- The Zero-Forcing (ZF) algorithm provides full equalization but
may also introduce a large increase in the noise level.
- Minimum Mean Square Error (MMSE) equalizing provides a
trade-off between signal corruption due to radio channel frequency
selectivity and noise/interference.
- MMSE provides both equalization and an acceptable SNR.
