Modulation asymmetry is a general technique, applicable to systems operating with either FDD or TDD, and where different modulation schemes may be used on uplink and downlink channels, respectively, in order to obtain different data rates and can provide some capability for asymmetric
traffic. The maximum ratio and direction of asymmetry is basically limited by the equipment design and the modulation formats implemented in practical systems.
However, higher order modulation schemes or coding with reduced overhead require higher S/I than the scheme in the other link. Therefore, such concepts can only be applied as a trade-off between link capacity and coverage for packet services in a link adaptation mode. Using different modulation techniques supports increased peak data rates for users having good radio conditions.
This method could be used to enhance both FDD and TDD systems in future extensions of the IMT 2000 standards and systems beyond IMT 2000. This means that the important regulatory question whether new spectrum should be reserved for TDD or FDD systems does not need to include the consideration of modulation asymmetry.
An additional method to improve asymmetric capacity is the use of adaptive antennas or more advanced detection schemes to increase the link capacity for one link. This would provide some additional possible asymmetry for a given frequency allocation. These techniques will add capabilities either alone or in combination to both FDD and TDD based systems. Similar radio propagation properties for TDD uplink and downlink show some advantages in the application of adaptive antennas for low mobile speed.
Advanced detection schemes to mitigate the impact of co-channel interference can be applied to both FDD and TDD.
2 Comparison of the various methods to provide asymmetric traffic capability
In general, the duplex scheme is one of many factors that determine the overall spectrum efficiency of a system. In terms of efficient support for asymmetric traffic, both FDD and TDD schemes have inherent advantages and disadvantages.
2.1 FDD scheme with symmetrical spectrum allocation
The maximum available user data rate per link is fixed.
This scheme has the following advantages:
– It allows for continuous (non-bursty) transmission in the uplink and downlink. This also allows for faster signalling of feedback information for, e.g. power control, link adaptation, and fast channel-dependent scheduling.
– For wide area coverage the range is primarily limited only by the system margin.
– No additional particular requirements for adjacent channel isolation or co-planning of systems in adjacent channels compared to TDD.
– Multi-operator co-location of BSs is possible depending on system design and frequency reuse independent of overall spectrum asymmetry.
– No inherent relation between the range of available maximum service data rates and the degree of asymmetric capacity, which is the case for TDD.
– Flexibility, to a certain extent, to traffic asymmetry.
Potential disadvantages to be considered are:
– Symmetric paired spectrum with a minimum duplex distance is required.
– The spectrum efficiency of the arrangement depends on the relation between the symmetric spectrum and the actual network traffic asymmetry.
2.2 FDD scheme with asymmetrical spectrum allocation
The maximum available user data rate per link is fixed.
This duplex scheme has the following advantages in addition to § 2.1:
– Flexible pairing of uplink and downlink carriers is possible that allows for asymmetric capacity. The spectrum usage is most efficient if the selected bandwidth ratio for both bands corresponds to the traffic asymmetry. The asymmetric spectrum can be used either as having more carriers in one direction, or as having wider carriers in one direction or a combination thereof.
– The multibandwidth alternative allows for higher peak rates in the direction with the wider band.
– Availability of additional unpaired spectrum is sufficient.
Potential disadvantages to be considered are:
– Asymmetrical paired spectrum requirement.
-
Estimation of future spectrum demand per direction is required and may be difficult in advance. An immediate adaptation may be difficult to implement but indications are that more spectrum in both uplink and downlink are needed and later adaptation is then possible.
-
The requirement for flexible duplex spacing, for the methods that require it, slightly increases the terminal implementation complexity.
– The spectrum efficiency of the arrangement depends on the relation between the degree of actual network traffic asymmetry and the degree of spectrum asymmetry.
– Multirate, multiple bandwidth capability for channels of different width are required if the uplink and downlink carriers are designed to have different bandwidths.
The maximum available service data rate per link depends on the ratio of asymmetry.
This duplex scheme has the following advantages:
– Availability of unpaired spectrum is sufficient. The identification of single blocks of spectrum may be easier than for paired spectrum.
– Flexibility is available with respect to the degree of traffic asymmetry, depending on the co channel and adjacent channel interference conditions. The spectrum usage is independent of the location of the switching point between uplink and downlink transmission.
– The spectrum efficiency of the arrangement is less dependent on the actual network traffic asymmetry since TDD can vary the degree of asymmetry within a specified range.
– If the neighbouring cells/systems agree on the same slot configuration, depending on system design and frequency reuse, the range of asymmetry is given by the number of time slots.
-
Capacity increase by using adaptive antennas can be further improved by using the reciprocity of the radio channel for low mobile speeds.
Potential disadvantages to be considered are:
– Services in adjacent bands must be able to cope with both uplink and downlink interference.
Synchronization and coordination of uplink/downlink of neighbouring cells is required with a small frequency reuse; in the case of a sufficiently large reuse cluster size no coordination is necessary within an operator’s frequency allocation but it is still needed between operators having frequency bands adjacent to each other.
– Multi operator co-location of BSs depends on the system design, the frequency reuse and the frequency separation of co-located operators.
– Isolation between adjacent channels is required.
Share with your friends: |